MUNGO GIVEAWAY: WIN A DIGIDESIGN STRIKE VIRTUAL DRUMMER!

[Link]

DEC./JAN. 2007 - VOL. 2 NO. 6

THE WORLD OF SOFTSYNTHS AND SAMPLERS

Akai EWI 4000m

Very Deep Clinic:
Electric Wind
Inside Instrument
Ableton breathes life into
Operator your synths

Reviews:
Steinberg Cubase 4 DAW
Native Instruments FM8 synth
RME Fireface 400 interface
Modartt Pianoteq synth
Big Fish Electro Magnetic Fury and Madahi: African Rhythms

USA $4.95
CANADA $6.50
0 1

0 7 447 0 057 92 5

Breakthrough: 7 Gigabytes of samples loaded on one Mac!

 From the

Editor
ince the last page of this issue is a look forward—what our writers The other big orchestra, Vienna Symphonic Library—probably the
S see happening in 2007 and beyond—it’s only fitting that we begin
with a superficial look back at the year that’s almost past. Quite a few
highest-end music software product in the world—came out with one of
two products that have really advanced the field of sampling: the Vienna
major products came out, and there have been some pretty important Instruments player. Of course, the huge update to their sample library
developments…and the following is by no means a comprehensive list, that goes with it is also no sneezing matter! But this brilliant player
just a few that come to mind. In no particular order: makes it easy to play, program, and manage their gargantuan library by
First, the big libraries. East West Quantum Leap Symphonic Orchestra sensing what you play and in many cases switching articulations auto-
got a big update (the XP version), but for me the company’s show-stop- matically. It’s certain to be influential.
per is EWQL Symphonic Choirs—an absolutely gorgeous library with its The other product to change sampling’s face is Native Instruments
own utility program for making the choir sing phrases you construct out Kontakt 2, specifically its add-on scripting feature. Among many other
of syllables. Now, you’re not going to pull off Mozart’s requiem with it, tricks, of a sudden you can add very good legato to any library. There’s
but it can do a lot, and even vaguely Latin syllables are a big step a lively discussion section about K2 scripting on the VI-Control forum
beyond oohs and aahs (which are also cool, but not the same thing). ([Link]), and it’s well worth stopping by to see what’s
SONiVOX, the company formerly known as Sonic Implants, comple- going on—and to see the latest scripts you can download.
mented their string section with the rest of the orchestra. Sonic Implants On the hardware front, the big news is of course Intel Macs that can
Symphonic Collection is a really nice sounding library, and now there switch between Windows and Mac OS on the same machine. How long
are three major orchestral libraries (soon to be four when Audio can it be before both are happily running different music software at the
Impressions comes out with theirs in January). In addition to the same time?
GigaStudio version, SISC is now available in Kontakt 2 format with its Season’s greetings and Happy New Year from all of us at Virtual
scripting engine (about which more later). Instruments magazine.—NB

Editor/publisher: Nick Batzdorf

Virtual Instruments is published bi-monthly
Art director: Lachlan Westfall/Quiet Earth Design
for $16.95/year, $26/two years by Virtual
Advertising manager: Len Keeler Instruments, Inc., 3849 Ventura Canyon,
Production manager: Laurie Marans Sherman Oaks, CA 91423-4710. 818/905-
Web designer: Denise Young/DMY Studios 9101, 1-877/ViMagzn.
NB@[Link].
Contributors: Jim Aikin, Jason Scott Alexander, Thomas J. Bergersen,
Periodicals Postage Rates are paid at Van
Peter Buick, David Das, Peter Dines, Doyle Donehoo, Gary Eskow, Jerry Gerber, Nuys, CA, and at additional mailing offices
Paul Gilreath, David Govett, Ashif “King Idiot” Hakik, Mattias Henningson, under USPS # 023-464. POSTMASTER:
Mark Jenkins, Michael Marans, Monte McGuire, Orren Merton, Chris Meyer, please send address changes to VIRTUAL
INSTRUMENTS, 3849 VENTURA
Dave Moulton, Zack Price, Frederick Russ, Bruce Richardson, Craig Sharmat,
CANYON, SHERMAN OAKS,
Lee Sherman, Dietz Tinhof, Jesse White. CA 91423-4710.

Advertising contact: Len Keeler 818/590-0018. Len@[Link] Distributor: Rider Circulation Services, 3700
Subscriptions/Address changes: 818/905-9101, 1-877/ViMagzn, Eagle Rock Blvd., Los Angeles, CA 90065.
323/344-1200. Bipad: 05792,
subscribe@[Link]. The best method is to subscribe via our
UPC: 0 744 70 05792 5 05
website: [Link]. Standard disclaimer: Virtual Instruments
Letters to the editor: NB@[Link], or fax: 818/905-5434. Magazine and its staff can’t be held legally
responsible for the magazine’s contents or
Writing for Virtual Instruments Magazine: query
guarantee the return of articles and graphics
NB@[Link] or call 818/905-9101. submitted. Reasonable care is taken to ensure
accuracy. All trademarks belong to their owners.
Everything in here is subject to international
copyright protection, and you may not copy or
imitate anything without permission.

© 2006 Virtual Instruments, Inc.

VIRTUAL INSTRUMENTS 5
 December/January 2006/7 26 7 Gigs Loaded on
One Mac
V2.N6 by Nick Batzdorf
Accessing 7GB of samples on a single Mac.

10 Letters

14 Launch
38 Sequencing Samples
by Thomas J. Bergersen
Introductions, updates, news
One of the world’s true masters at MIDI programming starts
a new series. This issue: balance.

44 Bent Clicks
by Paul Gilreath
One of the most overlooked aspects of a good sequenced
Voltage Control
16 by Michael Marans
performance is the tempo. Here are some pointers.

Let’s Make Some Noise!?Part B in our series on using your

synth’s noise generators to create drums, percussion instru-
ments, and sound effects.

54 MIDI Mockup
Microscope: Nathan
Furst
20 VOperator
ERY DEEP CLINIC:
on the Line
by Frederick Russ
In this installment of our series on composers and how they
did their MIDI programming, talented composer Nathan
by Jim Aikin Furst discusses “Heart of the Baja Part 2” and “Shark Attack.”
The add-on synth in Ableton Live 6 is a lot more than just FM Download these cues at [Link]
lite and follow along.

8 VIRTUAL INSTRUMENTS
 VI
December/January 2006/7
V2.N6

contents
24 Native 48 Modartt
Instruments FM8 Pianoteq
by Lee Sherman modeled
The second coming of the DX-7 gets acoustic piano
a major update.
by Jason Scott Alexander
This surprising piano only uses sam-
ples for things like pedal noise—it’s
actually modeled from the ground
up. As a result, every behavior of a

30 RME Fireface
real piano is accounted for.

VI
reviews
400 FireWire
Audio Interface
by Nick Batzdorf
52 Loop Librarian
This little half-rack box houses a 24- by Chris Meyer
bit/192kHz audio interface with 18
ins and outs. But that’s just the Loops loops loops and libraries.
beginning, because it can also route Reviewed: Big Fish Electro Magnetic
software tracks all around the room. Fury and Mahadhi: African Rhythms.

56 Steinberg
40 Akai EWI 4000s Cubase 4
by Nick Batzdorf by Mark Jenkins
Do you play a wind or brass instru- A look at the update to this massive
ment? The Electric Wind Instrument sequencing package.
is a wonderful controller for playing
synths. Its latest incarnation of this
fabulous instrument now includes a
built-in synth and some new MIDI random
control features—and it comes in at
half the price. tip
61 Importing DX-7 sounds into Native
Instruments FM8

VIRTUAL INSTRUMENTS 9
VI
l e t t e r s

Letters
write to:
nb@[Link] Waveguide synthesis, they’re using a different
variation of the Karplus-Strong plucked string
model, which uses a noise generator followed by
a filter with a delay feeding back into it.
Circling around your original question some
more, some instruments actually do model the
various circuit components of the product being
emulated. Universal Audio does that in their
plug-ins, for example. Whether a given softsynth
actually models the oscillator from the synth it’s
emulating is another question, but to me
absolute purity isn’t what makes a product
valid—it’s the bottom line sound.
Is FM synthesis physical modeling? Who
knows. What I do know is that a softsynth ver-
sion of the VL1 would be fantastic, because I
happening—it seems like they are just “mod- still consider it the pinnacle of synthesis in many
eling” the OUTPUT of oscillators and so forth. ways.
And if that’s the case, then the Yamaha DX7
was doing this kind of thing a long long time
ago.
David Jameson High praise
via email I just recently picked up your magazine
and am very excited about it. I am a police
As your letter points out, it’s not a very pre- officer that has patrolled the [withheld] area
cise term. And let’s face it: it’s very difficult to for many years. I recently was injured and will
market a product just by saying it sounds good no longer be able to do police work. Music
and some rock star—who you’ve just eliminated composition has helped me through my reha-
as a potential customer by making him or her bilitation in many ways, and your magazine
an endorsee—uses it. It’s far more interesting to has given me so much information on the
be able to talk about the technology. “We were emusic world.
very careful to do an excellent job recording this I know your mag is not necessarily aimed
sample library using good mics” isn’t going to at novices such as me, but please know that it
turn a lot of heads. is making a difference for me. Even though I
So yes, we have seen “modeling” used fairly don’t have any of the highspeed soft synths (I
liberally as a selling point. That was especially use Cakewalk Sonar with its stock instru-
true in the mid-’90s, when it was the buzzword. ments), with your help I am learning to get
Having said that, if the definition of PM is the most out of what I have, and that is excit-
using equations and algorithms to simulate the ing. I have a long recovery ahead that will not
Physical modeling behavior of a physical sound source, then what be easy, but emusic is playing a big part in
While I understand completely the notion Yamaha does did in the VL1 Virtual Acoustic keeping me positive and keeping my head on
of physical modeling as it applies to tradition- synth (which may have been the first commer- straight, and your magazine is a big help.
al instruments (i.e. representing the length of cial synth to use PM) absolutely is physical mod- Keep up the good work you are doing and
a trumpet, the materials of which it’s made, eling. The VL1 uses Digital Waveguide synthesis, I can’t wait for the next issue.
differential equations of the wave motion, which actually uses varying delay lines, and Chris Watson
etc), I have never understood what it means Korg used it, and at one point Technics used it via email
to do physical modeling of an analog synthe- in their WSA-1 synth.
sizer (for example, Yamaha claimed such for Applied Acoustic’s Tasman synth and I’m This is one of the nicest letters we’ve received,
the An1X). guessing Apple’s Sculpture (one of the synths Chris. Thanks so much, and we wish you all the
What exactly is being physically modeled? that comes with Apple Logic Pro, although best with your rehab.
It seems to me that to do physical modeling Apple only touts it as “component modeling”) The synths in Cakewalk Sonar are absolutely
of an analog synth, you have to model such are two current PM softsynths that come to nothing to sneeze at, by the way, in fact we
things as the behavior of transistors, resistors, mind. Modartt PianoTeq, reviewed in this issue, have a Very Deep Clinic about them in the
capacitors, i.e. the electronic equivalent of the uses PM to create its pianos, but the actual works And we certainly hope novices will enjoy
length of the trumpet and its materials. technique is a closely held secret. My educated the magazine too. Our “First DAW” series has-
But I don’t get the sense that that’s what’s guess is that if they haven’t licensed Digital n’t gone away!

10 VIRTUAL INSTRUMENTS
VI
l e t t e r s

to the distribution amp’s in, and then from two

of its outs to the synths’ S/PDIF inputs. Put the
synths on digital sync (a.k.a. external sync) and
that’s solved. Be sure to use 75-Ohm cables,
though, or the system may not be reliable.
Your other problem, finding an audio inter-
Virus TI, etc.) and pipe them into my single face with multiple S/PDIF inputs, is a little more
SPDIF I/O on my Firewire 410. I can’t be the difficult. MOTU used to make a 308 box for
only person out there that wants to do this, their PCI-324 and -424 cards, if you can find
can I? one. Lynx AES 16 is another, and in fact it fea-
Any ideas would be a help. Thanks. tures asynchronous I/O, meaning that it can
George Napier resample the digital inputs that aren’t clocked
via email together.
You don’t want to use USB, but adding an
First of all, thanks very much for including the inexpensive USB interface that has S/PDIF I/O
diagram of your studio. It’s actually more might be the most practical solution. There are
Digital puzzle involved than necessary to answer your ques- some very inexpensive ones, actually.
[Please see the fab diagram George included tion, but it’s always interesting to see peoples’ I also see that you’re also using Native
with his question.] rigs. This is technically a hardware question, of Instruments Kore—or maybe not using its con-
What does someone do if they want to use course, but since you’re integrating the hard- troller as an interface? It has S/PDIF out but no
the SPDIF outputs on multiple hardware ware with V.I.s as well… :) in, so that rules it out as a place to stick your
synths simultaneously? Most sound cards and There are two issues here: digital hardware synths’ digital outputs. But you could
interfaces have only one SPDIF I/O, so only clocking/sync and where you’re going to stick make it part of an aggregate device if you want-
one of these things can be connected at a the S/PDIF outputs. As to clocking, every device ed to use it for more analog inputs. (Aggregate
time. in a digital audio system has to be referencing devices are seen as one larger audio interface by
I’ve done a bit of research. There are a few the same digital clock, or you’ll get clicks and Mac OS X software.)
products out there (M-Audio makes one) that pops. Both the VSynth and the Virus have If you go into the Aggregate Device editor in
will let you connect multiple ADAT devices S/PDIF inputs, so the obvious answer would be Audio MIDI Setup, you’ll see a checkbox for
and provide a master clock and connection to to use the clock from your FireWire interface. which device to use as the clock master, and
your DAW/soundcard. I also ran across a But the FW410 only has one S/PDIF out, and then there are options to resample the inputs.
series of product called “digital detanglers” you need two. It so happens that RCA S/PDIF That’s what you want to do with Kore, since it
by Z-Systems Audio Engineering, but these digital signal uses 75-Ohm video cable, so you can’t be clocked externally.
seem to be simply format converters...and can create multiple outputs using a standard
très expensive. video distribution amp. Radio Shack makes one
I’m looking for something in the “pro- for $50 that should work fine, but there’s a wide
sumer” space that will allow me to hook up range of models on the market. Suggestion box
three or four SPDIF sources (e.g., V-Synth XT, Just run S/PDIF cables from the FW410’s out An article for a VI issue that would be very
valuable to me, and to many others as well,
would be a step by step overview of running
V.I.s on an external Mac and connecting it to
a host computer. I know far less about this
than I would like to; it is time consuming to
hunt up the information; and there are many
users, particularly in the Mac community,
who are using second computers as V.I. hosts.
If such an article could address things like
the network feature in Audio MIDI Setup in
OSX, ways to bring a laptop/Traveller combi-
nation or similar rig into the main set-up,
avoiding conflicts with dual installs of the
same plug/V.I., etc, it would be very much
appreciated.
Any chance this might happen soon?
I’m enjoying the publication, thanks for
your continued innovation. Been reading
you since Home & Studio Recording [which
later became Recording] back in the day.
Robert R. Martin, PhD
Victoria B.C., Canada

Absolutely, great suggestion. You shall have

it, Sir, and thanks for the compliments.
This letter is here as an example of what we
welcome from you readers. Please let us know
what you want to see:
NB@[Link]

12 VIRTUAL INSTRUMENTS
VI
l a u n c h

Launch Introductions, updates, news

M-Audio NRV10 Analog Mixer/FireWire

Audio Interface
This $899 unit combines an 8x2 analog mixer with a 10x10 Pro Tools M-Powered-com-
patible audio interface. Among other features, the unit features a built-in effects section
with 16 effects that run on its hardware. It also comes with Mac and Windows control soft-
ware that adds dynamics processing plus two VST effects slots to each channel.
[Link]

Native Instruments
KORE 1.1, Kontakt 2.2
universal binary
This free update to NI’s KORE
Universal Sound Platform can now run
without its controller/audio-MIDI inter-
face hooked up to the computer. Other
new features include a new SingleSound
format; integration with NI’s new
Massive synth and the latest updates to
their other instruments; and a pre-listen
feature for auditioning sounds before
they’re loaded.
The most important new feature in
v.2.2 of their Kontakt sampler is that it’s
now a universal binary, which means it
works on Intel Macs as well as PowerPCs. VSL Saxophones and Elements
[Link] Vienna Instruments libraries, update
Vienna Symphonic Library’s 24-bit/44.1kHz Saxophones
Collection ($475) features 35GB+ of soprano, alto, tenor, bari,
and bass saxes, intended for both legit and jazz use.
Articulations for jazz and big band are included: tongue slaps,
Millennia Media growls, screams, extended harmonics, key clicking, and more.
There are also long and short downward ending phrases trig-
Orchestral Recording gered as release samples.
Discussion #1 Elements ($365, updates $45) is an 18GB expanded version
of the Horizon Series Glass & Stones library, featuring all kinds of
If you’re interested in learning about mallet and percussion instruments such as glass bells, musical
how to improve the production of your glasses, glass harmonica, verrophone, and lithophone. New to
sampled orchestral sequences, this dis- the library are such mainstays as bass waterphone, and flutter-
cussion about large ensemble acoustic tongued blown bottle.
recording is sure to help. Millennia Both come with the Vienna Instruments player, which was just
Media makes some of the best analog updated to v.1.1, which among other things: loads more quickly
recording processors (mic preamps, etc.) and has some user interface improvements. This player features
in the world. Performance Detection algorithms that that switch articulations
[Link] automatically by analyzing the intervals, repeated notes, pat-
terns, and speed of your playing in real time.
[Link]

14 VIRTUAL INSTRUMENTS
 VI
l a u n c h

McDSP Updates
McDSP has been
producing some of
the best Pro Tools-
compatible plug- Digidesign Mbox 2 Mini
ins for a long time.
Their latest updates $329 list now buys you a
to the Native and USB interface that has both
HD product lines phantom powered mic and
include support for line inputs, and line and head-
Intel Macs, phone outputs. The software
improved RTAS effi- bundle starts with Pro Tools
ciency, faster inter- LE, and then it includes the
face updates and Xpand! sample playback/syn-
loading times, and Pro Tools HD Accel thesis workstation, several
card optimizations. The ML4000 ($495) Bomb Factory plug-ins, about
is a high-res brick wall look-ahead limiter 38 DigiRack DSP plug-ins
and a multiband dynamics processor. (dynamics, reverb, EQ, pitch shift, delay, and on and on)…and then the fun starts. It also
Analog Channel LE and ML4000 LE are includes several lite versions: Ableton Live Lite 4, Propellerhead Reason Adapted,
new additions to their $495 Project FXpansion BFD lite drums, IK Multimedia SampleTank SE, IK Amplitube LE, IK T-Racks EQ,
Studio bundle. Celemony Melodyne Uno, and more.
[Link] [Link]

Garritan Stradivari Solo Violin 2.0

The update to this solo violin now comes in a Native
Instruments Kontakt 2 Player, so you don’t need Kontakt to play it.
The Garritan Strad features “Sonic Morphing,” which crossfades
across dynamic layers; lets you control the onset, rate, and speed
of vibrato; and shape vibrato in real time without the phasing
problems you get when crossfading normally. $199, upgrades $29
+ s/h.
[Link]

MPC Lockbox, Big Fish

DrummerPacks for
DrumCore
The SonicEmulations MPC Lockbox loop
library ($59) is a new expansion pack for
Submersible Music’s DrumCore loop librari-
an/player. This new DrummerPack features
Alexander Publishing hip-hop/DJ-type loops and single hit drum and
percussion sounds, ready to play by MIDI.
“How Ravel Big Fish’s “Brain: One Stroke Done” (reviewed
Orchestrated: Mother in our 10-11/06 issue) and “Roots of South
America 2” libraries ($79 each) are also now
Goose Suite” available as DrumCore DrummerPacks. This pack
is also available in loops—which are ready to be
This edition includes a book, score, and drag/dropped into a DAW—as well as individual
CD, designed for learning orchestration. hits to be triggered by MIDI.
The analyses that follow each movement
include electronic scoring insights. [Link],
[Link] [Link],
[Link]

VIRTUAL INSTRUMENTS 15
VI
f e a t u r e
more
online
[Link]

Voltage
Control
Let’s Make Some Noise!—Part II in our series on
using your synth’s noise generators to create
drums, percussion instruments, and sound effects.

Last issue we concentrated on building a standard Let’s open the amplifier by setting the
Decay control to its halfway position. Still
nothing. Why?
trap kit. This time around we’ll look at creating Look at the filter setting. With LP12 select-
ed (a lowpass filter with a 12dB per octave
slope), no sound is passing through the filter,
complementary percussion instruments and immerse so the settings of the amplifier envelope liter-
ally have no effect on the sound. Toggle to
LP24 and play a note; same deal: no sound.
ourselves in some really bad weather. (We would say, Now select HP12æa highpass filter with a
12db per octave slope. Hey! That almost
sounds like a thick hi-hat hit!
“So grab your Macs,” but the pun would be way too If you’ve been following these Voltage
Control articles for a while you’ll immediately
obvious, and we’re not that desperate….) know why. First, a highpass filter allows all
sound higher than the filter cutoff frequency
to pass through and be heard. So with our fil-
Ooh, Baby, I Got an Itch… ter closed down to 0Hz, we’re allowing all fre-
We’ll start off with two versions of the now quencies higher than 0Hz to pass through
ubiquitous “record scratch,” only we’ll be unfiltered.
using virtual vinyl. Second, Oscillator 2, set to generate noise,
In Synth1 (see Sidebar, Downloading Synth1 is the only oscillator sounding (note the fully
and the Example Patches), call up program 45, clockwise position of the Oscillator Mix knob).
Simple Click. Play any note, and all you’ll hear As we pointed out last time, hi-hats and cym-
is a plain old ordinary click?and a rather bals are comprised entirely of noise (all fre-
unmusical one at that. quencies playing at equal amplitude) and the
It’s instructive to analyze the origins of the amplifier envelope is the primary tool used to
by Michael Marans click, so give the Synth1 front panel the once- shape that noise into a recognizable sound. In
over (Fig. 1). You’ll notice that the Attack, this case the amplifier’s Decay setting is creat-
Decay, and Sustain parameters of both the fil- ing the shape of a medium length hi-hat hit.
ter and amplitude envelopes are set to their Now back to our regularly scheduled pro-
minimum values. Since these values essential- gramming: record scratches. Starting with our
ly shut off the filter (i.e. no harmonic content boring click, let’s start molding the sound into
passes through) and the amplifier is “closed,” something resembling the mistreatment of
all we hear is the click of the electronics turn- vinyl. We’ll begin by selecting LP12 for the fil-
ing on for a brief moment when a key is ter. Now turn up the filter envelope’s Attack
pressed. But no “real” sound is actually being time to about one-third and play a note.
generated by the synth engine. You’ll hear that we have a rough approxima-

16 VIRTUAL INSTRUMENTS
 VI
f e a t u r e
Perking up
The high-pitched strike of a Clave, the clas-
sic Latin percussion instrument (Program 48,
Clave F3), can be created using a combina-
tion of noise and filter self-oscillation. A bit on
the latter: when the Resonance parameter of
the filter is set to maximum, the filter cutoff
frequency is so accentuated that it actually
produces a tone (generally a pure sine wave).
The pitch of that tone can then be con-
trolled by adjusting the filter’s cutoff frequen-
cy. (Note that if you want the tone to track
the keyboard, you need to turn the Filter’s
track control [TRK] to its maximum position,
which,on most synthesizers will produce the
proper voltage to play properly intonated 12-
tone scales.) By combining the tone produced
by filter resonance with the tone(s) produced
by the oscillators carefully, you can create
noise-based timbres with distinctive pitches.
In creating our Clave sound, we used two
noise sources. OSC2 is producing pure white
Fig 1: The vinyl “record scratch” patches created in this article began life as a simple, uninspiring electron- noise. OSC1, set to a sine wave, has its FM
ic click. Worse, the click was really nothing more than the sound of Synth1 trying make sound, but the 0 (Frequency Modulation) control set to about
(minimum/off) settings of the envelope generators make playing a purposefully-generated sound next to the 50% mark (64 on the MIDI parameter
impossible. Turning the click into record scratches required selecting the proper filter, adjusting the filter and value scale). The sidebands (harmonic fre-
amplitude envelopes, setting the filter cutoff point and resonance, modulating the sound with LFOs, and quencies)) created by feeding the complex
adding effectsæin other words, using a variety of voltage controls to turn the plain and the boring into the noise waveform of OSC2 into the pure sine
bad and the beautiful. See the main article for details. wave generated by OSC1æwhich is what the
FM control is doingæare wild and “random,”
thereby generating a sound heavily predomi-
tion of a simple scratch, but the sound cuts add a bit of automated panning using an LFO nated by noise. (See Fig. 2.)
off rather abruptly. to animate the sound’s stereo image. To hear OSC1 on its own, turn both the
Time to enlist the filter’s Decay control. (Turn on LFO1 and note that the sound Filter’s RES knob and the OSC MIX control to
Bring it up to about one-third, and you’ll immediately travels across the stereo sound- 0 (fully counter-clockwise). Note that by
notice that you now have the basis of a fairly stage, in accordance with the fairly rapid increasing or decreasing the value of OSC1’s
convincing quick scratchæalbeit one that’s speed of the LFO. Slowing down the speed FM parameter, you can add as little or as
awfully thick. Adjust the filter Attack and allows you to move the sound across the much noise to the sound as desired.
Decay parameters to taste, using Attack to stereo soundstage with each successive key- Reselect Program 48 so that you’re back to
control the speed of the scratch’s onset and on. Click on LFO1’s KEY button to retrigger the original Clave patch. If you listen closely,
Decay to control how long it lasts. You might the sound from the exact same stereo posi- you can hear the sound’s noise characteristics,
find it helpful to think of the word “Wow” tion with each note-on.) though the dominant timbre is the short,
while adjusting these parameters.
Now about that thickness. Please say hello
to the Bandpass filter (BP12 on Synth1). A
bandpass filter allows a specified band of fre-
quencies both above and below the frequen- Let's open the amplifier by setting the Decay
cy cutoff point to pass through and be heard.
This type of filter lets us maintain the core control to its halfway position. Still no sound. Why?
timbre of the sound while simultaneously
attenuating unwanted frequenciesæthat is,
those that are too low and too high.
So select BP12 and play a few notes. Notice
how this thinning out of the sound really helps Our final version of a basic scratch, which pitched strike that’s created by filter resonance.
home in on the tonality of a scratch. utilizes all of the above, is stored in Program Last tweak: toggle between the LP12 and
The next steps are all “tweaks,” and subject 46, Record Scratch 1. Program 47, Record LP24 filter settings. You’ll note that LP12
to personal preference and musical applica- Scratch 2, builds on the core sound to create allows for more low frequency noise content.
tion. First, try increasing the filter cutoff a “double scratch” by adding filter velocity, Our preference (strictly personal) is for the
and/or the resonance to thin the sound fur- and using a stereo delay to split and double “cleaner” LP24 sound, played between C3
ther and give it the perception of having a the sound across the stereo soundstage. and F3. As always, experiment with different
higher pitch. (Remember that pure, unfiltered One of the highlights of the patch: LFO2, keyboard ranges and filter settings.
noise has no pitch, so accentuating higher or assigned to Filter (try assigning it to
lower frequencies will make the sound appear Amplitude as well), can be used to add multi- Give yourselves a hand
to be pitched higher or lower, accordingly.) ple scratches by increasing the value of the The analog handclap, popularized by
Then add compression (Effect > Comp, on speed control and varying the note-on veloci- Simmons way back when (yes, there were
Synth1) to make the sound fatter and more ty. Truly complex rhythms can be created earlier incarnations, but Simmons is credited
consistent in volume from start to finish. through the use of this control and a bit of for creating the handclap that slapped you in
Unison mode and filter SAT (Saturation) can keyboard dexterity. the face), makes use of several key synthesis
thicken the sound considerably, and you can techniques. Programming a basic handclap is

VIRTUAL INSTRUMENTS 17
VI
f e a t u r e
stick has a sound tinguish the “droplets’ ” individuality. LFO2
that contains a num- uses a high-speed random sine wave assigned
ber of the same to amplitude, which further serves to define
spectral compo- individual droplets. Finally we adjust the FRQ,
nents. The primary RES, and SAT controls for the desired tonal
difference between coloration.
it and a hi-hat is in
the shape of the Uh-oh…bad weather ahead
envelopes (both fil- Mother Nature is pretty good at creating
ter and amplitude). impressive noises, and it just so happens that
So even though the two of the most common?lightning strikes and
rainstick is about as thunder?are comprised of ingredients readily
organic an instru- available on basic analog subtractive synthesiz-
ment as one can ers. If you’ve been following this series, you
imagine, the bits know that we always break down sounds into
and bytes of our vir- basic components, i.e. their core tonality
tual analog synthe- (waveform), harmonic content (filter), and
sizer can do a very amplitude (VCA, or Voltage Controlled
convincing job of Amplifier), and then apply various voltages
emulating it. from modulation sources, such as envelopes
Let’s start by call- and LFOs, to manipulate the core ingredients.
ing up Program 51, So with those basic principles in mind, let’s don
Fig. 2: Frequency Modulation (FM) is a type of synthesis (popularized by Medium Hi-Hat. This our raincoats and call in the thunderclouds.
Yamaha’s DX series) in which one wave (a modulator) is fed into another (a carri- provides our basic A lightning strike usually consists of a sharp
er) to generate a timbre that is harmonically rich. When the modulator and carrier timbre. (For those of crack, followed by thunder. The particular
are sine waves, the harmonics produced are predictable. In Synth1, the output of you not using incarnation we created here (Program 53) has
OSC2 is the modulator, and the carrier is OSC1. Since both of these oscillators are Synth1, our example the lightning striking right on top of us. In
using complex waveforms in the Clave patch, the resulting timbre produced when Medium Hi-Hat other words, the thunder happens simultane-
using FM comprises hundreds of frequenciesæwhich, due to their number and sound is created ously with the initial strike.
amplitude, are a form of noise. Try raising and lowering the FM amount to hear using noise The lightning sound starts with OSC2 set
the effect on the “noise” portion of the Clave patch. processed through a to noise, and the filter parameters set to cre-
high pass filter with ate a medium length “crack.” (You should
a bit of resonance experiment with the filter envelope values to
pretty easy; making a knock-you-over-dead added to give it a metallic edge. The filter see how they affect the sound. Hint: Decay
clap is a multi-step process. envelope is used to fine tune the timbre, but and Sustain are critical.) Synth1’s filter SAT
We’ve programmed two claps. Program the amplitude envelope provides the majority parameter has been used to fatten the crack.
49, Small Handclap, relies almost exclusively of the hat’s “sharp hit” characteristic.) LP12 is the filter selection, since it allows both
on basic filter and envelope settings to create We now need to tweak the envelopes to a great amount of high frequency content to
the core sound. A highpass filter is used to eliminate the hi-hat’s sharp attack and create pass through for the initial crack, while at the
nail the timbre, which if you listen closely has the trademark rainstick sound of “grains” same time providing enough filtering for us
many of the spectral characteristics of the hi- (rice, small pebbles, etc.) falling through the to create the rumbling thunder that follows.
hats we constructed last time around. hollow stick. Start by increasing the amplifier’s Now check the settings of the Amplitude
To see for yourself the dramatic effect a fil- Attack time to a bit more than halfway-up.
(CONTINUED ON PAGE 62)
ter has on a sound, toggle through the other Play and hold a note. Notice how this simple
available modes. Once you leave HP12 (the adjustment has a dramatic effect on the
highpass filter) our clap becomes lifeless and sound, and has taken us quite far toward the
sounds more like an unwelcome stranger rainstick emulation.
knocking at our door. Now we’ll adjust the Amplifier Decay Downloading and
Now go back to HP12 and enable Tempo timeæincreasing it to about three-fourths of
Delay. Our settings double the sound and its throwæto create the long, smooth sound installing Synth1
spread it across the stereo spectrum. Next, of the grains continuing their journey through
enable Chorus/Flange, which thickens the the stick.
and the example
sound. For fun you can enable Effect A.D.1, Okay. We’re pretty close to the basic shape patches
which gives you an interesting percussive of the sound, but it’s not definitively a rain-
sound, but out of the realm of the clap. stick. To go that extra mile, we’ll need to As with our past articles, we’re using
Now Select Program 50, Big Handclap. enlist a number of other parameters. the freeware Synth1 for our examples.
This is essentially the same program as Small Call up Program 52, Rainstick, and play a Instructions for downloading and
Handclap, but when we enabled Effect A.D.1, note. You’ll notice that the sound is now installing Synth1 can be found at More
we changed the filter to BP12 (bandpass). much more refined timbrally (it’s not just Online at
Doing that allowed us to retain the basic clap pouring rain anymoreæwe can hear the indi- [Link]. There
soundæby cutting out much of the high fre- vidual “droplets”). We’ve done this by making you’ll also find the Synth1 patch exam-
quency content we heard in the A.D.1 experi- some minor changes to the filter envelope, ples used in this article, as well as those
ment in the previous paragraphæwhile at the but more significantly we’ve enlisted the aid used in our last three installments. You
same time thickening the sound considerably. of the LFOs. can use the synth of your choice, since
LFO1 is set to a noise (random) waveform the controls we refer to are common to
Deep into the (Rain) forest with its output routed to the filter, causing virtually all software synths (at least
While a completely different type of instru- the filter cutoff point to change rapidly over those of the analog subtractive variety).
ment from the hi-hat?and certainly not metal- the course of a held note. This rapid random-
lic in any way?the decidedly primitive rain- ization of the frequency content helps to dis-

18 VIRTUAL INSTRUMENTS
VI
v e r y d e e p c l i n i c

Operator on the Line

The add-on synth in Ableton Live
by Jim Aikin 6 is a lot more than just FM lite

When I first heard that Ableton had released a 4-operator he first part of this clinic is for those who

FM synth called Operator as an optional add-on for Live, I’ll T are new to FM synthesis. It introduces
Operator’s FM features briefly, and shows
how to explore the tone colors available with
admit I glanced at it and yawned politely. As much as I like FM. After shining a spotlight on a few of
Operator’s other features, we’ll tackle some
FM, how could four operators hold their own against instru- expert concepts—both FM programming and
ways to use Operator in conjunction with Live
ments that have six? 6’s new Racks and expanded modulation rout-
ings.
While working on this article, I interviewed
ace remixer and sound designer Francis Preve
When persuaded to take a closer look, though, my opinion ([Link]). Francis designed many of
the factory presets for Operator, and has had
ratcheted upward. The multimode filter and the useful list of tracks on Top Ten dance remix CDs. In addi-
tion, he’s the author/editor of The Remixer’s
waveforms open up a lot of sound design possibilities that Bible (Backbeat Books). You’ll find his insights
(uncredited) scattered throughout this article,
aren’t apparent on the surface. The new version of and a few specific tips in the sidebar. Thanks
Fran!
Operator available with Live 6 contains in itself only a few
FM 101
small refinements—but the more flexible architecture of FM (frequency modulation) synthesis was
first made popular by Yamaha in the 1980s.
Live 6 boosts Operator to a whole new level, as you’ll dis- In FM, complex sounds are created by having
one simple waveform modulate another.
cover in this clinic. If you like synthesizers, this little mon- Often the waveforms being produced by the
ster will truly chop wood and carry water for you. oscillators are sine waves, which are about as
simple as an audio signal can get. The oscilla-
tor producing the sound we hear is called the
carrier, and the oscillator being used to

20 VIRTUAL INSTRUMENTS
change the tone of the carrier is called the FM discovery step-by-step
v e r y d e e p VI
c l i n i c
smaller adjustments, and move the Fine value
modulator. Let’s walk through the process of creating a up to between 3 and 6. You’ll hear the tone
A standard FM synth has at least four oscil- basic sound in Operator. Launch Live and fol- become animated as the two oscillators are
lators, and many have six. These can be con- low along: detuned from one another.
figured in various ways. For instance, with 1. Drag Operator from the Browser into an 6. With operator B still selected, drag the
four oscillators we might have two empty MIDI track. Don’t drag one of the envelope sustain level down to zero. Now
carrier/modulator pairs, or we might have existing presets—we want to start from operator B is producing an attack transient.
Adjust the decay time shorter or longer to
change the length of the attack transient, and
try adjusting the Coarse tuning of B to
change the harmonic character of the tran-
sient. Leave the envelope set at a medium
decay length, perhaps about 3 seconds.
7. Click in the algorithm panel (lower right)
and select algorithm 7. In this algorithm three
modulators all affect one carrier. The sound
won’t have changed yet, because we’ve only
turned up operator B. So turn up operator C
to –15dB, turn its envelope sustain down to
zero, and set the decay time to a much short-
er value, in the 40-60ms range. Turn its
Coarse tuning up to about 7. This will give
each note a very percussive attack. FM syn-
thesis is especially good at tuned percussion
sounds like the one you’ve just created.
Fig. 1: The algorithms in Operator. The box or boxes on the bottom row of each icon are carriers, and 8. Turn up operator D to about –50dB,
those above them are modulators. Algorithms 6 (right end of the top row) and 10 (highlighted) are new in change its Coarse tuning to 8, and give it an
Live 6. At first glance 9 and 10 might seem almost identical, but if you click on 9 and look at the “wires,” envelope with a 4-second attack, a 32-second
you’ll see that in 9 the modulator affects all three carriers, while in 10 it affects only carrier C. decay, and a –15dB sustain. This will produce
a slow swell containing new overtones as the
note sustains.
9. To hear the contribution that each oper-
three modulators, all operating on a single scratch, so select Operator itself in the ator is making to the composite tone, click
carrier. The configuration of oscillators is Browser. their colored boxes to mute them one by
called an algorithm. Operator offers a choice 2. Click the In button in the MIDI track and one. Muting operators is extremely useful
among 11 algorithms, as shown in Fig. 1. play your keyboard to verify that Operator is when you’re developing an FM sound.
Two key factors affect the tone of the car- operating. You should hear a sine wave. 10. Add velocity response to some or all of
rier: the relative frequencies of the carrier 3. Looking at the panel, you’ll see that the the operators to make your new preset more
and modulator, and the amplitude of the default patch uses algorithm 1, in which all playable.
modulator. Cranking up the level of the four operators are stacked. (It’s visible in the
modulator adds more overtones to the lower right area.) But you’re only hearing a Three important features
sound. While the ratio of the frequencies of sine wave, because the levels of operators B, Operator is sometimes compared to the
the carrier and modulator controls which C, and D are set to –inf dB. Turn up the Level Yamaha TX81z, a popular 1980s-era rack-
overtones will be heard, the formula that
determines the exact combination of over-
tones produced by a given frequency ratio is
not simple, and not interesting either. The
best way to work with FM is to try different The formula that determines the exact
tunings for the carrier and modulator, and
listen to the results. combination of overtones produced by a given
To give the user control over these factors,
each oscillator has its own coarse and fine frequency ratio is not simple, and not interesting
tuning knobs and an output amplitude knob.
In addition, each oscillator has its own ampli- either.
tude envelope, which you’ll see in the center
panel when you click on one of the oscillators
in the left column. Unlike an analog-type
synth, where one amplitude envelope is often
called on to contour the entire sound, FM of operator B while playing the keyboard. mount synth that also produced its tones
synths give us separate amplitude control You’ll hear the sound gradually acquire more using 4-operator FM. But Operator has some
over each oscillator. overtones. This is the core of FM synthesis. powerful features that go well beyond what
The combination of an oscillator and an 4. After setting the Level of B to a reason- was possible in those early days.
amplitude envelope is called an operator— able value, such as –15dB, try adjusting the Filter. Operator’s multimode resonant filter
hence the name of the synth. (Memo to Coarse frequencies of A and B. You’ll discover is a vital sound-shaping tool. In Live 6, the fil-
Ableton: when Antares named their filter that each ratio of A to B produces a distinc- ter has been enhanced with a choice between
plug-in Filter, it was not a trend you want to tive combination of overtones. 12dB and 24dB per octave rolloff slopes. The
emulate. Calling Live’s amazing new sampler 5. Return A and B to Coarse settings of 1 24dB options produce a more definite filter-
Sampler... well, maybe all the good names and 1. Then click on the Fine tuning knob of ing action, while the 12dB options are more
have been used up.) B, hold down the computer’s Ctrl key to allow gentle.

VIRTUAL INSTRUMENTS 21
VI
v e r y d e e p c l i n i c

Fig. 2: The Operator panel. The filter is selected, when you do this, the result is the same as if The advanced course
and its Freq<Vel parameter is being edited in the each overtone in the modulator was modulat- A book could be written about sound
center window. ing each overtone in the carrier separately. design in Operator. Here are a few ideas to
Tip: a standard technique in analog synthe- get you started.
sis is to set three oscillators to the same pitch Modulation routings. Live 6 handles
In the filter’s control panel (see Fig. 2), and then detune one slightly flat and another real-time MIDI modulation input much better
check out the Freq<Vel and Freq<Key parame- slightly sharp. Operator’s Fine knobs can be than earlier versions. You can now apply a
ters. Use these to modulate the cutoff fre- cranked only up, not down—but there’s an single source, such as the mod wheel on your
quency from velocity and the MIDI key num- easy solution: for the oscillator that you want keyboard, to any number of destinations, and
ber. Assuming you’re using a lowpass filter to detune flat, set the Coarse knob an octave set the minimum and maximum values that
mode, Freq<Vel will cause notes that are lower (for instance, 0.5 rather than 1, or 2 the MIDI controller will
played harder to sound brighter, emulating rather than 4) and then turn Fine up to 995 have. This means you
the characteristics of an acoustic instrument. or so. can do complex, coor-
When Freq<Key is active, the cutoff will rise as Envelope looping. Operator’s seven dinated sweeps of
you play up the keyboard, ensuring that low envelopes (one for each oscillator, plus pitch, numerous Operator
notes don’t have too many overtones and filter, and LFO amount) can loop. By setting parameters with a sin-
high notes aren’t too muted. the loop modes to Sync and selecting differ- gle physical move-
ment. But your con-
troller routings won’t
be saved with the
Operator preset, which
means if you load the
preset into a different
The sound pallette of FM is deep and wide, song you’ll have to
recreate them from
and as the song says, there's mild and honey on scratch. What to do?
This is where Live
the other side. 6’s new Racks features
come to the rescue:
1. Create an
Instrument Rack and
insert an Operator into
it.
Waveforms. A nice set of waveforms is ent rhythmic subdivisions for various 2. Assign all the
available for Operator’s oscillators (see Fig. 3). envelopes, you can create pulsating, percolat- parameters you want
These include square, sawtooth, and triangle ing rhythms. For the most striking results use to modulate to one of
waves—the types found in analog synthesiz- fairly quick decay times and low sustain levels. the Rack’s Macro
ers. In fact, Operator can be used very effec- Unfortunately, there are still no dotted knobs.
tively as a pseudo-analog synth: dhoose algo- rhythms for the Repeat value, but with a judi- 3. Assign your mod
rithm 11, which has four parallel carriers, and cious combination of 16ths, eighths, quarters, wheel to the Macro
choose a suitable waveform for each of and half-notes, you can generate some ear- knob (see Fig. 4).
them—preferably not a sine wave but some- catching patterns. 4. Save the Rack
thing that has more overtones. Then shape Tip: after setting up a patch in which the preset.
the tone with the filter envelope. envelopes are synced to the beat, start the
Operator lets you use any combination of transport. Play notes in a chord in various
waves in FM synthesis, not just sine waves. rhythms, holding all of the notes down so Fig. 3: The pop-up menu for Operator’s wave-
The results tend to have a lot of overtones. that the looping has a chance to work. By forms. The square and sawtooth waves are most
But by turning down the level of the modula- triggering notes at various points in the meas- often used in emulations of analog synthesizers,
tor and choosing your waves with care, you ure, you can change the pattern in hypnotic but depending on modulator levels and other fac-
can achieve many timbres not possible with ways. tors, they can be used effectively in FM synthesis as
sine waves alone. For the technically minded: well.

22 VIRTUAL INSTRUMENTS
 Now when you load the Rack preset into a
v e r y d e e p VI
c l i n i c
different song, you’ll only have to make one
modulation routing—mod wheel to the
Macro knob.
Francis Preve’s Quick Tips
Also note that individual envelope seg-
ments can be modulated using MIDI or a • When using the filter with FM sounds, program the FM section first, before turning
Macro knob. This lets you do tricks like on the filter. In other words, use the filter to enhance a sound that’s already good,
changing the sound of the note attacks from not to fix problems.
a MIDI slider in subtle and highly controllable • The Spread knob in the pitch section doubles the CPU load of a patch. If your CPU is
ways. hitting the wall, turn the Spread down to zero.
Additive synthesis. Load four
Operators into an Instrument Rack, choose • When starting work on a new FM sound, choose algorithm 8 (two carrier/modulator
algorithm 11 (four carriers) for each of them, pairs). It’s a good basic choice, and you can later change if you need something spe-
and tune the oscillators to Coarse values of 1, cial. When you add the second modulator to a carrier, it’s easy to end up with too
2, 3, 4, 5, 6, and so on up to 16. Leave the many overtones.
waveforms all set to sine waves. Now you
• The pitch envelope can be routed to some oscillators but not others, which makes it
have a 16-sine-wave additive synthesizer.
useful for subtle detuning effects and adding accents in envelope loop patches.
Give some of the operators envelopes that
decay quickly to zero, and give others slow • Unless you know what you’re doing, do not tune a carrier to any Coarse value
attacks so that the tone swells. Adjust the except 0.5, 1, 2, 4, or 8. Those are the octaves of the fundamental. If you tune a car-
Level of each oscillator to get a pleasing rier to 7 and a modulator to 14, for instance, you’ll get a tone that’s difficult to
blend. Try detuning some of your additive retune to concert pitch even with the aid of the Transpose knob.
overtones slightly.
• If you’re using Operator with an arpeggiator, check to make sure the R button is
FM with a low-frequency carri-
switched on. If it’s off, the arpeggio will chew up your polyphony while adding CPU
er. When you click the Fixed button on an
load.
oscillator, it no longer tracks the keyboard.
This feature has at least three important uses. • In an FM patch, apply the mod wheel to one or more modulators’ Coarse tuning
First, if you tune a modulator to a high, fixed knobs to create PPG-style stepped wave sweeps.
frequency (above 2kHz or so) and keep it at a
• Using two Operators in an Instrument Rack will give you 8-operator FM. When
low level, it will mimic the sound of a fixed-
doing this, choose complementary algorithms to make the most of the composite
frequency formant filter. Second, percussive
sound.
attack transients, such as the sound of a mal-
let hitting a xylophone, are pretty much fixed • If you need an extra modulator, use the LFO. Set it to a high frequency range, set its
in frequency no matter what note you play, Rate<Key to 100% so it will track the keyboard, and hold down the Ctrl key to tune
so using a couple of fixed-frequency operators its Rate precisely.
to produce mallet attacks works well.
The third use of fixed frequencies is less
obvious. If you tune a carrier to a low, fixed
frequency, perhaps a Freq setting of between staple of FM synthesis, and very pleasing to sound palette of FM is deep and wide, and as
30 and 70Hz and a Multi (multiplier) knob listen to. If you use Operator’s square waves the song says, there’s milk and honey on the
setting of 0.01, it will add a chorused quality for both the modulator and carrier with this other side.
to the tone of the modulator. This sound is a type of patch, you can get some nasty rasp-
ing tones that sound a lot like pulse width
modulation in an analog synth. Jim Aikin writes about music technology for
various publications. He’s also a cello teacher
Fig. 4: An Instrument Rack with four Operators The operating table and a hobbyist computer programmer. You can
in parallel. Several parameters have been assigned Creating your own sounds with Operator visit him online at [Link]. VI
to Macro knob 1, which I’ve renamed Mod Wheel. (or any other FM synthesizer) takes a little
Note the small pink dot next to the Level knob for more dedication than triggering samples or
operator C; this indicates that it is now under the using an analog-style instrument, because the
control of a Macro knob. operators can interact in so many ways. The

VIRTUAL INSTRUMENTS 23
VI
r e v i e w

Native Instruments FM8

The second coming of the DX-7 gets a
major update

Review by Lee Sherman

ative Instruments’ FM7 has become a

N
Fig. 1: FM8 now includes the intuitive
staple of modern music production. In KoreSounds Browser. This allows you to search for
a way this is the second coming of the sounds and filter them according to instrument,
famous Yamaha DX-7, which dominated the genre, synthesis type, and musical attributes such
world in the early ’80s when it came out. as timbre and articulation
Native Instruments FM8, $339; While the FM synthesis technique these
upgrade from FM7, $119. instruments have in common is capable of
creating some of the cheesiest sounds ever free-form FM matrix, additional waveforms for
recorded, it’s also capable of creating some the Operators (the original only had sine
[Link] wonderful, unique sounds. As a result it’s waves), graphical envelopes with nearly
never really gone out of style. Among other unlimited stages, host-syncable looping, and
things, FM synthesis can produce a metallic chorus and delay effects.
System requirements: Mac OS X sheen that’s fantastic for clanging bells, lush While FM8’s considerably revamped user
pads, glassy electric pianos, and tight basses. interface no longer resembles the panel of a
10.4, G4 1.4 GHz + or Intel Core It’s as welcome as ever in today’s music. DX-7 (it’s now a soothing gray-on-white color
Duo 1.66 GHz +, 512MB RAM; Recreating the sound of the original hard- scheme), the biggest changes are due to the
Windows XP, 1.4GHz Pentium or ware is a lot easier when the original hard- software’s integration with Native
ware was a digital synthesizer. But what’s real- Instruments’ Kore software
Athlon, 512MB RAM. ly important is whether the software has host/controller/interface. FM8’s own patch
improved upon the original. brower is similar to Kore’s, and its sounds use
Indeed, FM7 not only delivered the sound the Kore Single Sound Format.
Formats: Stand-alone, Audio of the DX-7 (you can even import all of the If you do open FM8 within Kore, you can
Units, VST, DXi, RTAS. old patches—see Random Tip: Loading origi- use complex layers of FM8 and Native
nal DX sounds into FM8 [which works the Instruments programs. KoreSounds can take
same way]), it went some way toward advantage of the new morphing technology
Copy protection: online authori- improving the arcane process of program- and multi-effects chains, making for some
ming FM sounds. We’ll explain FM synthesis truly massive sounds.
zation using included utility to man- in the next section, but FM7 also provided an A tabbed navigator along the right side of
age all Native Instruments auths. advanced sound architecture that includes a the software’s main window lets you switch

24 VIRTUAL INSTRUMENTS
 VI
r e v i e w
quickly between the Browser and Attributes glance view of the
view, expert and novice programming modes, key parameters for
access the new arpeggiator, or bring up the all Eight Operators
effects rack. The Attributes view lets you filter all at once.
your sounds according to instrument, genre, Alternatively, you
synthesis type, and musical attributes like tim- can choose to drill
bre and articulation. down into the
Like all NI instruments, FM8 runs stand- more advanced
alone or as a plug-in on both Macs and parameters for
Windows XP machines in all the standard for- individual
mats. It is now a universal binary, meaning it Operators by click-
works on Intel-based Macs. ing on labeled
buttons in the
Smooth operators navigator or the
In FM synthesis, sounds are created by FM matrix. The
modulating one waveform with another, pro- FM Matrix pro-
ducing a more complicated waveform. vides an intuitive
Waveform generators are called Operators; an view of operator
Operator that’s modulating another one is routing.
called a Modulator, and one that’s being pro- If you’re not interested in dealing with the Fig. 3: The new arpeggiator looks and acts like
ducing actual sound is called a Carrier. Any complexities of FM, an Easy/Morph page can an old-school analog sequencer, bringing life to
Operator can act as either a Carrier or a be used to tweak an overall sound quickly. sounds
Modulator, and it is the arrangements of Here you can route LFOs, adjust the ampli-
these waveforms that form the building tude envelope, apply effects, and more.
blocks of FM synthesis. FM7 was limited to a pair of fairly standard sounds for use in the matrix. The 4-way
In FM8, up to eight Operators can be freely effects: chorus and delay; FM8 now has a full matrix is reminiscent of waveshaping synths
mixed in any combination, and you can effects rack that includes 11 high-quality like the Sequential Prophet VS and the Korg
choose from a wide variety of waveforms— effects including talk, wah, phaser and VS software but much easier to program. It
unlike the DX-7, whose Operators produced flanger, tremolo, reverb, pitch-able delay, made me wish for a joystick to control it with.
only sine waves and were arranged into 32 tube amp, overdrive, and EQs seemingly Fortunately all movements can be automated.
set configurations (called Algorithms). Other drawn from NI’s Guitar Rig. Why should gui- Additional movement comes from the pro-
DX-series instruments had four Operators. tarists have all the fun? grammable arpeggiator, with its 32-step
The last two FM8 Operators are special and The smooth workflow and visual approach matrix that’s reminiscent of a classic analog
aren’t found in Yamaha’s implementation of to parameter editing remains a huge advan- step-sequencer. It’s the arpeggiator that
FM synthesis. Operator X adds noise and tage over the old hardware, while the high- accounts for the many on-board patterns
waveshaping, while operator Z is an analog- resolution audio engine in FM8 sounds even you’ll find in the browser. DJs, dance produc-
style multimode filter with a filter envelope. better than its predecessor and far better than ers, and musicians of all kinds will find new
Also unique to NI’s version of FM are multi- a real DX7 (due to better D/A convertors, inspiration from the groove presets.
stage envelopes that can be edited graphical- oversampling, multi-stage EGs, and the built- One of the great things about FM7 has
ly. The next step would be the ability to in effects). always been the wide availability of additional
import your own samples as you can on the presets. You could purchase one of NIs sound
Yamaha SY77/99; perhaps that will come in Motion sickness packages, choose from hundreds of FM7 pre-
FM9? Since it takes some work to master FM pro- sets available online, or import patches from
Instead of only allowing you to view a sin- gramming, FM8’s sound morphing features the DX-7 family (most of which are readily
gle Operator at a time as you did in FM7, are particularly welcome as a serendipitous available on the Internet).
FM8’s Expert window now provides an at-a- means of coming up with new sounds. The FM8 ships with 960 preset sounds for all
sound-morphing feature provides a 4-way styles and genres, including both original
matrix onto which you drag and drop sounds FM7 patches, the complete FM7 Sounds Vol.
Fig. 2: FM8’s new Expert window provides an at- from the browser. 1 and Vol. 2 libraries, as well as 200 new
a-glance view of Operator parameters. By clicking and dragging across the matrix, sounds—enough programs that you may not
you can ever need to know what an Operator is. Plus
morph your there’s the above-mentioned KoreSound for-
current sound mat.
with the tim-
bres of the Is eight enough?
additional At a time when most new virtual instru-
sounds. (In ments seem to play tribute either directly or
order to indirectly to their hardware forbears, its grati-
ensure that fying to see Native Instruments extend the
the original DX-7 legacy. FM8 sounds better, it’s easier to
sound is still program (the interface improvements alone
playable, only are worth the upgrade price), and its sound
the correct morphing and on-board arpeggiator go far
parameters for updating digital FM synthesis for today’s
morph.) A music. VI
Randomize
button gener-
ates new

VIRTUAL INSTRUMENTS 25
VI
f e a t u r e

7 Gigs Loaded
on One Mac
A breakthrough solution to the problem that’s
plagued sample library users for the past few
years, and an explanation for beginners about
why it’s been a problem.

eanwhile, for Mac users it turns out

The great thing about modern sample libraries is
that they’re huge; the problem with modern sample M that an old solution to the problem
that never worked reliably now does
work: running stand-alone software samplers
on the same machine that’s running a DAW.
libraries is that they’re huge. This allows each program to have its own
memory allotment, solving the problem
before all our software gets adapted to work
on 64-bit operating systems.
Being huge means they require lots of computer But first let’s start with some background.
Those of you who just want to find out how
memory to run, even if they stream samples off to access 7GB of RAM on a G5 or Mac Pro
might want to skim over the next couple of
hard disks (because they still need a head-start sections.

memory buffer). We’re on the verge of getting new Why RAM is your friend
Having a healthy percentage of the sam-
64-bit operating systems from Apple and Microsoft ples and programs in these libraries cued up
and ready to play is the name of the game,
and it’s always required way more RAM (ran-
that will allow virtually unlimited memory access— dom access memory) than any single com-
puter can access. For a good percentage of
unlike today’s 32-bit operating systems. our subscribers—about a third of you, to be
precise—the solution is to use multiple com-
puters.
The reason for needing all these samples
by Nick Batzdorf
loaded is simple: you want to them ready to
play even though you won’t actually use all of
them in a given piece of music; it would be

26 VIRTUAL INSTRUMENTS
 VI
f e a t u r e
being able to access it for loading samples.
While Mac OS X will hold and recognize up
to 8GB of RAM in a PowerMac G5 or 16GB in
an Intel-based Mac Pro, its 32-bit operating
system “only” allows each program to access
a real-world maximum of about 3GB.
Part of the reason that only 3GB of the the-
oretical 4GB allowed by 32-bit OS X comes to
work, as we explained in the June/July issue, is
that system frameworks and libraries (extra
code used by programs) take some of it. Then
OS X itself will use about .5GB outside any
running programs, plus it needs some RAM
for caching. Without that RAM for caching
the system will go to virtual memory—mean-
ing it will pretend the hard disk is RAM—and
that’s not good for our real-time applications.

However…
That’s 3GB of RAM access for each
program, though. Unlike Windows, two
loaded-to-the-gills programs in Mac OS X can
theoretically access a combined 6GB…except
that things get unstable when two programs
are that full.
In practice—before the solution offered in
this article—the previous functional loading
record for a G5 with 8GB of RAM installed
was between 5 and 5.5GB. That record was
set running Apple Logic Pro with about 3GB
plus an additional 2GB in the VSL Vienna
Fig. 1: 7GB on one machine, and it’s running reliably. This Activity Monitor screen dump shows 6.96GB of Instruments player plug-in, which behaves
memory access in a G5 with 8GB loaded. The VSL-Server program is listing samples loaded into the plug-in like an external program running outside the
versions of the VSL Vienna Instruments player running inside Logic Pro, which is accessing 1.38GB. (Logic is host DAW and has its own memory space.
also running other programs, including Spectrasonics Stylus RMX.) (Normally instrument plug-ins’ memory
Four stand-alone copies of the Vienna Instruments player (all named differently, which you don’t see) are comes out of the host DAW’s RAM budget.)
running outside Logic. In addition, Native Instruments Kore is running several instruments, including Now there’s a way to push that concept
Toontrack’s EZ-Drummer (a sample-playing instrument). farther. Mac users can take advantage of OS
While the sequence isn’t playing here, you can see from the idling percentages that the CPU hit isn’t going X’ ability to allot each program its own mem-
to be outrageous. That’s what’s changed recently to make this possible. ory space to run—are you ready for this—
about 7GB of RAM in a G5 with 8GB
installed. You can actually load a little more
deadly to have to stop and reload programs tweak more than doubles a single machine’s than that, but again, things become unreli-
every time you wanted to, say, switch violin RAM access—potentially standing in for two able.
bow directions. This is no different from an machines—that’s quite a leap forward. This may well work on Intel Macs that have
artist sitting down to draw with a box of pas- Another limitation is that Windows itself up to 16GB of the (very expensive) RAM they
tels containing dozens of colors; he or she can’t recognize more than 3GB installed in a use installed. We haven’t had a chance to test
certainly won’t use them all in a single pic- computer. However, PowerMac G5s under OS that yet, and not all the software is compati-
ture, but who wants to go fishing in closets X 10.3 or higher can recognize up to 8GB of
for a lighter shade of green. installed RAM, depending on the model
Now, the first logical question is why you (some G5s only have slots for 4GB). The new Fig. 2: It’s necessary to make renamed copies
can’t just install, say, 32 Gigabytes (GB) of Intel Mac Pros can hold up to 16GB of RAM. when you’re running multiple instances of the same
RAM in one computer—more than you’re But note that there’s a difference between program. This is handled automatically by the Mac
likely to need—and be done with it. The recognizing the installed RAM and actually OS.
answer is that there’s a finite number of
memory locations our present 32-bit operat-
ing systems can address; they only have
enough bits to map out 4GB of RAM. That
translates to a real-world maximum of 3GB or
even slightly less.
Getting even more real-world, somewhere
around 1.25GB of RAM is about all you can
access in a stock Windows XP machine with
2GB installed. However, on p. 26 of our 6-
7/05 issue, Mattias Henningson described
how to access almost 3GB of RAM using “The
3GB Switch” tweak. Before the 3GB switch
there was no reason to install more than 2GB
in a Windows machine; considering that the

VIRTUAL INSTRUMENTS 27
VI
f e a t u r e
(which includes Kontakt 2 Player and also
embedded libraries such as East West
Colossus); with the VSL Vienna Instruments
player; and with Synthogy Ivory Grand.
All of these instruments have stand-alone
and plug-in versions, and you can run both of
them simultaneously. What’s more, you can
run multiple instances of a stand-alone sam-
ple player if you make duplicates of the pro-
gram and give each a different name. Just
highlight the program icon in the Finder and
hit command/D as shown in Fig. 2; the copy
will be called “copy of” whatever it is.
You can also run stand-alone hosts outside
the DAW. We were able to run Native
Instruments Kore outside Logic Pro with pret-
ty good stability. The one fly in the ointment
is that not every program is happy when you
run stand-alone and plug-in versions at the
same time, or when you run plug-ins in two
hosts; you’ll have to experiment with what
gets loaded where.
This trick seems to work best if you load
the DAW first, then the stand-alone pro-
grams—although we’re going to stop short of
making that a blanket statement, since we
haven’t tested every combination of software
on the planet. Anecdotally, the system also
feels more reliable with a greater number of
stand-alone program instances, each loaded
with a moderate amount, than to run few
instances all loaded to the gills.

MIDI
The way you send the stand-alones MIDI is
Fig. 3: The IAC Driver in Audio MIDI Setup. Loading 7 to use IAC, the inter-application communica-
Make sure you have enough ports to send MIDI to The 7GB trick isn’t complicated: load your tion feature built into OS X’ Core MIDI. If
all the stand-alone programs running outside your DAW up as usual, then launch stand-alone necessary, go into Audio MIDI Setup program
DAW. versions of the sample-playing programs
you’re going to run outside your DAW. We
tested the trick with Native Instruments Fig. 4: Enabling IAC bus 1 in this Kontakt 2
ble as of this writing anyway, but it should be Kontakt- and Kontakt 2-family instruments Player (running SampleLogic AIR).
possible.
The only caveat is that you also have to
keep an eye on the processing power. If you
run too many programs, you can easily run
out—in fact that was the factor that prevent-
ed this trick from working in the past; some
of the latest programs have quietly become
more efficient when running outside a DAW.
By the time you read this, the 64-bit
Windows Vista will probably be out, and the
full 64-bit Mac OS X will be imminent.
Hopefully it won’t take too long for all the
music software developers to come out with
versions that feature 64-bit memory access.
And hopefully the RAM used in the newest
computers such as the Mac Pro won’t be as
expensive as it is right now (about $250 per
GB as of this writing, vs. $100 per GB for the
PC3200 memory in the G5 in this article—
which is up from $75 at the time we filled it
with 8GB).
Meanwhile, 7GB of RAM access (as shown
in Fig. 1) is far from shabby. What’s more,
running samplers outside the DAW can actu-
ally be advantageous: you don’t have to wait
for them to load when you fire up a new cue
in your sequencer.

28 VIRTUAL INSTRUMENTS
and click on the IAC driver. Make
VI
f e a t u r e
sure it has enough 16-channel ports
enabled, as shown in Fig. 3.
The IAC driver is just another
MIDI interface as far as all parties
are concerned. Simply assign the
appropriate IAC port and channel as
the output of your sequencer track
and the input of your stand-alone
instrument, as shown in Figs. 4 and
4A. Kontakt 2 and its relatives
receive 64 MIDI channels, so you
can just keep creating IAC busses
until you have enough.

Audio
There are a few ways to route the
audio from the stand-alone pro-
grams into your DAW’s mixer.
Which one you choose depends on
the stand-alone program and on
your audio interface(s).
RME’s interfaces have a “loop-
back” mode for their onboard mix-
ers that let you route software out-
puts back to inputs. Apogee is about to intro- consisting of Soundflower and your audio Fig. 5: Creating an Aggregate Device so that
duce the same feature with the Maestro soft- interface(s) in Audio MIDI Setup. Aggregate Soundflower becomes a virtual audio interface. This
ware for their Symphony cards. The hardware devices link multiple audio drivers so they is one way of looping the stand-alone instruments’
does the routing. appear to software as one big interface with audio back into your DAW.
In our last issue we had a Very Deep Clinic lots of inputs and outputs. Fig. 5 shows a
on Soundflower (free download from slightly over the top Aggregate device using
[Link] an inter-applica- MOTU PCI-424 hardware, an RME Fireface Sometimes it’s necessary to use more than
tion audio routing utility. That works very 400, the Mac’s built-in stereo optical digital one driver. For example, the stand-alone
well; simply select a Soundflower input as the I/O, and 16 Soundflower I/Os. Vienna Instruments and Ivory Grand players
output for the stand-alone and the input to You can also just make a physical connec- don’t let you select individual outputs—they
your DAW’s mixer. tion between outputs and inputs on your just uses the default pair for the driver it’s
You’ll want to create an aggregate device audio interface. That can be a sacrifice if you using. So if you’re running more than one
don’t have a lot of I/Os, but it’s a perfectly instance and you want separate outputs into
workable solution if you have extras. your DAW, you’ll need to use more than one
Fig. 4a: Just assign each program its own MIDI Connecting a TOSlink cable between the driver. Soundflower + the built-in audio would
channel as you would with a physical MIDI inter- Mac’s digital output and input would give be one solution.
face. you a stereo path into your DAW’s mixer. On the other hand, the Native Instruments
stand-alones can be directed to any or all out-
puts of the selected sound driver. Note that
all programs can use whatever driver (includ-
ing aggregate ones) they want.

Alignment of the stars

One final disclaimer: we’ve tested a few
combinations of programs and come to the
conclusion that 7GB is a reasonable estimate
of the amount of RAM you’ll be able to
access. Please don’t sic your attorney on us if
it turns out you can only work with 6.75GB
on your rig—although you might also be able
to get better results.
Bear in mind that there are many combina-
tions of hosts and instruments, all of which
may or may not perform well together. Also
bear in mind that these are computers; we’re
dealing with the occult. VI

VIRTUAL INSTRUMENTS 29
VI
r e v i e w

RME Fireface 400

FireWire Audio Interface
This little half-rack box houses a 24-bit/192kHz
audio interface with 18 ins and outs. But that’s
just the beginning, because it can also route
software tracks all around the room.

Review by Nick Batzdorf

here’s not much to say about most ing/routing/metering software to control it.

T audio interfaces. They have a certain

number of ins and outs, hopefully they
sound good and work reliably, and some of
It also has a few features and characteristics
that make it function exceptionally well in
V.I.-based set-ups.
them have no-latency monitoring features.
And some—not the RME Fireface 400 being The ins and outs
reviewed—come with nice light-version soft- RME calls the Fireface 400 a 36-channel
ware bundles, an important a consideration if interface, and indeed it has 18 ins and outs:
you’re first getting started. eight analog I/Os, 8-channel lightpipe in and
The RME Fireface 400, on the other hand, out, and stereo S/PDIF—plus 2 x 2 MIDI. But
RME Fireface 400, $1199 is a higher-end FireWire interface than most calling it an 18 x 18 interface doesn’t tell the
people will start out with, and it has a lot to story of its onboard no-latency mixing and
talk about. There’s some pretty serious analog routing features, because you can also route
[Link], distributed by and digital engineering packed into this little 18 software outputs to any or all physical out-
half-rack box; while it doesn’t have a light puts at the same time the 18 inputs are get-
Synthax ([Link]) software bundle, it comes with some pretty ting routed to the software.
sophisticated Windows and Mac OS X mix- What’s more, it has a Loopback mode that

30 VIRTUAL INSTRUMENTS
 Fig. 1: The Fireface Settings control
VI
r e v i e w
have lower noise and distortion specs.
panel. Note the AutoSync clock mode, Those noise specs come with an asterisk,
which automatically locks to an incom- however, because one of the first things you
ing digital signal. Combined with the notice about the 400’s software is that the
unit’s SteadyClock circuitry, which input level readouts are bouncing around at
removes the jitter from incoming clock about -109dB with nothing connected.
so what goes out is very clean. This Translation for the less nerdy: it’s quiet.
effectively lets you run your whole rig One spec that does make a difference is
from any input without having to worry that the 400’s mic inputs have 65dB of gain
about its quality. vs. the 800’s scant 50dB. That extra 15dB of
gain means you can use the 400 comfortably
with low-output dynamic and especially rib-
bon mics.
drive headphones or line-level out- In case you were wondering whether the
puts. Front panel inputs 3 and 4 bus provides enough power for mic inputs to
can accept line or instrument-level sound good, that’s not a concern. For one, it
inputs; all the other analog I/Os are comes with a switching power supply, so you
line-level jacks on the rear panel. don’t have to use the bus. But FireWire has
no problem delivering upwards of 25 volts,
400 vs. 800 and these mic preamps sound quite transpar-
The 400 has much in common ent and solid; I heard no difference running
with RME’s Fireface 800, which has the unit bus-powered or plugged in. As a
been out for a while. Apart from matter of fact, the Fireface gets enough juice
the 800’s full 1U size, the main dif- from the bus to get warm to the touch.
ferences are that the 800 has an The Fireface 400 is a very nice-sounding
extra 8-channel lightpipe I/O; it interface. You really do hear an obvious sound
isn’t bus-powered like the 400; it quality improvement over garden variety lower-
has more mic inputs (but still eight priced units. The difference is especially notice-
simultaneously active analog ins); it able when you play something like acoustic gui-
can take advantage of the faster tar recordings through it—the sound becomes
FireWire 800 bus; and its line inputs detailed and solid rather than harsh.

lets you route the software signal assigned to

an output back to same-numbered input
(where it can in turn be routed further). This
allows you to send the output of one software
program—such as a stand-alone sampler—
into another one—most likely your main
DAW—without cables. That’s the first feature
that could be very useful for V.I. rigs.
The interface can operate at up to 24-
bits/192kHz…or to be precise it can operate
at anything between 27kHz and 200kHz due
to the way it’s engineered. Quad speed sam-
ple rates (192kHz, etc.) are sent out the
S/PDIF port—which is also an AES/EBU port—
and the Fireface supports the S/MUX protocol
for transferring four channels at 96kHz over
the ADAT connections.
All the Fireface 400’s analog I/Os are bal-
anced 1/4” TRS, with two exceptions: front
panel mic/line/instrument inputs 1 and 2 are
on Neutrik combination XLR/TRS jacks; and
outputs 7 and 8, also on the front panel, can

Fig. 2: RME’s TotalMix software front end to the

box’ onboard digital mixer. The top row is the
physical inputs, the middle row is the software
tracks’ outputs, and the bottom row is the physical
outputs. Outputs 3&4 are in Loopback mode,
which routes their signal to inputs 3&4 so you can
record or monitor them inside your DAW. Among
other applications, this allows you to record and
monitor the output of one software program into
another—for example if you’re running a stand-
alone software sampler outside your DAW.

VIRTUAL INSTRUMENTS 31
VI
r e v i e w
but no digital input to clock
to. It’s really well thought out
and implemented.

Software
You can operate the Fireface
400 stand-alone without a
computer. Its last-saved rout-
ings and functions are active; it
can convert formats, act as a
mic preamp, send digital clock,
and so on. All the settings you
choose in the Fireface Settings
utility (for Mac or Windows;
see Fig. 1) are remembered.
The other included pro-
gram, the Fireface Mixer (Figs.
2 and 3), controls the 400’s
TotalMix onboard digital
mixer/router. When everything
is being displayed, the top row
has faders for all physical
inputs (which you can name)
and the middle row is for up
to 18 software channels head-
ed for the 18 physical outputs.
These two rows can be
assigned to any or all of the 18
outputs, which is what the
bottom row of faders control.
You can group and pair
faders, and they’re scaled. So if
Fig. 3: In addition to being passed directly to the CD player’s optical S/PDIF digital output. If you move one fader in a pair or group, the
host oomputer, any input can be assigned to any the CD player wasn’t jittery to start with, opti- others—which can be at different levels—
or all outputs. That applies to physical inputs and cal connections—whether S/PDIF or ADAT change by the same percentage rather than
“returns” from software tracks. The TotalMix lightpipe—are fine for carrying audio, but the same number of dB. While this mixer
Matrix, with inputs on the horizontal axis and out- notoriously jittery as clock sources. That’s doesn’t have dynamic automation built in, it
puts on the vertical one, shows this (along with the especially true over longer runs, so I used a can be controlled by MIDI using standard
level of each bus). 12’ lightpipe cable. Mackie Control protocol and automated
Lo and behold, I really couldn’t hear a dif- externally. It also has eight factory and eight
ference between the music coming down user preset snapshots, recallable by MIDI pro-
If you run the 400 at 4x sample rates that lightpipe and the hard disk file with the gram changes.
(around 192kHz), the ADAT lightpipe inputs
disappear, since they max out at 2x (around
96kHz). It’s hard to run V.I.-intensive sessions
at anything other than 1x ([Link] or 48kHz)
sample rates anyway, so that’s probably not a The unit should sound the same whether you
consideration for our purposes; regardless,
the word clock output can be set to single clock it with a dedicated diamond-studded word
speed when the interface is running at 2x or
4x speed. clock generator that costs more than the unit itself
That means you can continue to use the
Fireface as a master clock for your whole rig if or from the biggest piece of digital garbage you
you’re running slave computers and other
devices at 44.1 or 48kHz. can find.
Digital
The Fireface 400 has an important feature
called SteadyClock that in addition to locking Fireface running under internal sync. That was TotalMix’ interface is slightly different from
up quickly, removes essentially all the jitter true when monitoring through the Fireface the way most mixers work. Each fader on the
from incoming digital signals. That means the and when monitoring through my usual set- top two rows can only control/display the
unit should actually sound the same whether up, which means that the clock leaving the level going to one output pair at a time,
you clock it with a dedicated diamond-stud- Fireface is equally clean either way. RME takes selected from a pop-up button. So it’s possi-
ded word clock generator that costs more this farther with an AutoSync feature that ble for Input 1 to be showing its level going
than the unit itself or from the biggest piece locks to an incoming digital signal when you to Outputs 3 and 4, Input 2 to be showing its
of digital garbage you can find. turn it on. level to Outputs 1 and 2, and so on.
To see whether SteadyClock really is all This digital clocking scheme can save a lot That’s what you want when inputs are sim-
that, I burned some music onto a CD and of switching around in set-ups with multiple ply routed to different places; once the rout-
locked the Fireface 400 to an old and cheap computers or devices that have digital outs ings are set up, all you’re concerned with is the

32 VIRTUAL INSTRUMENTS
level balance. But this can be confusing when The Windows version of TotalMix supports
VI
r e v i e w
the test dual 2.5GHz G5 registered 35% of
channels are routed to multiple destinations, so ASIO Direct Monitoring, in which the switch- one CPU when playing a stereo audio track.
TotalMix has a Submix mode (see Fig. 4) in ing is handled by “remote control” from the By comparison, the MOTU PCI-424 card in
which all the input/software output fader dis- DAW. Given that lots of interfaces have no this machine registered 30% at the same
plays move to the currently selected output. latency monitoring these days, it would be buffer setting. That’s not a significant differ-
That lets you adjust the total balance going to great if Apple incorporated a similar standard ence.
your main monitors, or headphones in a in its systemwide Core Audio driver. Furthermore, FireWire has the advantage of
recording situation, and so on. (Submix mode The Fireface 400 also comes with a DIGI- being more obsolescence-proof than internal
parallels what MOTU does in CueMix, the soft- check, a Windows-only audio analysis pro- computer cards. With PCs you can swap out
ware that controls its built-in digital mixers.) gram. DIGI-check displays levels, information motherboards for one with the PCI slots you
Now, rather than providing 17 aux send about the phase and spectrum, and bit statis- need, but you can’t do that with Macs. Apple
knobs plus the main fader for each channel, tics and noise. All this is calculated in the Computer has gone through several different
RME created a convenient routing matrix (see unit’s hardware—as is all the metering in PCI varieties just in the past three or four
Fig. 3). This overview screen makes it easy to TotalMix, on Mac as well as Windows. years—and this isn’t the first card change
assign inputs to multiple outputs and see TotalMix’ meters are very useful, showing we’ve seen. While each generation has legiti-
their levels right away. both average and peak levels, plus they have mate technical improvements over the previ-
The big advantage to having a digital a numeric readout. ous one, a lot of musicians express frustration
mixer built into audio hardware is that it pro- over having perfectly good hardware ren-
vides “no latency” monitoring, avoiding the Performance dered “legacy” so quickly.
few milliseconds of delay a signal incurs going PCI slots provide a more direct path to the It also should go without saying that the
through a computer. (“No latency” means computer, so all things being equal, FireWire bus-powered Fireface 400 is intended to work
the delay is defined by the digital converters, is always going to have a little more latency. well as a portable recording unit.
and it’s a couple of milliseconds like on a digi- Furthermore, on its PCI systems, RME is able
tal mixer.) to use hardware to help stream audio in both In your Fireface
That means you’re using the software directions, reducing the CPU load. This is This is a very well-engineered, good sound-
mixer in your DAW and also adding a second explained in the impeccably detailed manual, ing interface with a lot of features that make
one built into the audio hardware to control which has a lot of interesting technical infor- it work especially well in virtual instruments
the routing around your studio. Switching mation. rigs. It costs more than an entry-level inter-
between software and hardware monitoring However, the performance of this FireWire face, but you really do get more. VI
to avoid latency is somewhat of a nuisance on interface is still very good. At a 128 buffer—at
Macs—although the TotalMix preset snap- which the latency is subjectively not an issue
shots make this much easier than it could be. when you’re playing V.I.s (at least for me)—

VIRTUAL INSTRUMENTS 33
 Especially
Mungo
Enter to win a Digidesign Strike
Digidesign Strike is a revolutionary instrument More than just a high-defini-
tion drum module, Strike
plug-in that makes it easy to create professional drum
uses real drum performances
performances in Pro Tools with uncanny realism and and provides unprecedented
unbelievable human feel. control over every aspect of
your virtual drummer’s sound
and playing. Select a kit, set
the mic placement and leak-
age, choose a room, and tell
your drummer how to play as
you adjust the intensity, com-
plexity, timing, groove,
dynamics, and more—in real
time. This is one instrument
you have to see to believe.

With Strike, getting the

groove you want is as simple
as choosing one of the high-
definition kits and telling the
“virtual” drummer how to
play: Adjust the playing
intensity, complexity, timing,
groove, dynamics, and much
more—all in real time.

36 VIRTUAL INSTRUMENTS
Giveaway
ultimate virtual drummer

You’d know all about Strike if you’d read

our last issue. Can anyone really afford
not to subscribe?

Why would anybody want to miss the best writers in the industry,
helping you get the absolute most out of your softsynths and samplers?
You’ll pay only $16.95 a year or $26 for two years, compared to $59.40
on the newsstand.

Not only that, you’ll be able to download the magazine before it even
gets printed, plus you can access PDF files of our back issues. Your
support will help us print more issues and more pages, which will
benefit all of humanity.

But most importantly, you won’t miss a single important issue of

Virtual Instruments Magazine.

Just go to [Link]/giveaway to enter.

VIRTUAL INSTRUMENTS 37
VI
f e a t u r e

Sequencing
Samples
Part 1 in our new series on the art of MIDI
programming: balance

T
he problem often starts at the sample 2. Now it’s time to find proper balance
Without doubt the most library level (lack of proper/natural between the various instrument groups. At
dynamic relations between the various this stage I like to have a classical recording
common beginner layers, particularly often in the case of cross- handy, just to benchmark against. The record-
fading patches) and works its way into com- ing you’re benchmarking against has to be
positions through peoples’ careless handling recorded with a simple A/B mic setup (no
mistake is lack of of MIDI CC#11 (expression) or CC#1 (mod spot mics—simple stereo pair over the con-
wheel) data. ductor) and the piece has to be dynamic. It
In this article I’ll offer a few pointers that needs a triple forte tutti passage, since this is
attention to natural can help you out if you’re fighting the how you find the maximum sound level for
headaches of improper volume relations each instrument group. I recommend any
balance among the between your orchestral instrument samples. recording of Gustav Mahler’s 5th from the
1990s.
Basics Begin with the first instrument in your
instruments at play. This All commercially available libraries have score, the piccolo flute. Find a part in your
normalized sample pools (meaning that every benchmark recording where it’s playing at full
instrument, from solo bass flute to four horns blast. Approximate the level in your
is a problem even in in unison, is set so its loudest point is as loud sequencer/orchestral template and make sure
as possible). So the first step is to make sure all the articulations of this instrument retain
experienced peoples’ your template is balanced properly. This is a their relative balance.
rather long and tedious process, since you Configure any pan data too at this point if
need to rebuild the natural volume relations you want to. Move on to the next. Repeat
mockups, and it’s the between the patches. until you’ve done this for all the instruments
I’ve split this process into two phases: in your template. This could take a few days,
1. Find the proper relative balance between but it is well worth it.
result of an improperly each articulation of an instrument in your If you’re meticulous in your approach, the
template. Essentially this means you should end result will be an almost properly balanced
balanced orchestral find the proper volume relations between orchestral template that is the basic starting
your staccs, sustains, runs, tremolos, trills, etc. point.
Start with the loudest articulation (top layer
template. of aggressive staccs, or sustains for strings) Advanced
and measure the volume of your other articu- Adding to the long list of what I consider to
lations against this. Flautando patches, sul be problems with commercial sample libraries,
ponticelli, sul tasto etc. will all be relatively we have the normalization issue, which effec-
quiet in the real world, so make sure these tively kills the natural relation between dynam-
patches are lower in volume. Just how much ic layers. A good example is the flute, which in
by Thomas J. Bergersen you need to lower them is hard to say, but the real world is an instrument of great dynam-
use your ears— the mids/low mids are gener- ic variation through its register; you can’t play
ally good indications. If you go too loud, the loud in its lower register and you can’t play soft
mids become too prominent and balance in its higher register.
between the other articulations is lost. Do this Due to ruthless normalization, the flute’s
for each instrument separately. low range is as loud (in dB) as its high range,

38 VIRTUAL INSTRUMENTS
the horns’ lower range is as powerful as its like 36 violins and a 4-part chord will sound
VI
f e a t u r e
mid/high range, and so on. It’s not possible like 48 violins.
to reconstruct the proper natural relations Or at least this is the general consensus.
unless you know all the instruments very well. Theoretically speaking this practice is obvious-
So if you really want to create a properly ly wrong, especially when you consider the 1 voice: CC#11 - value 127
balanced orchestral template, you’ll have to improper balance that is obtained throughout
hire a musician for each instrument and the orchestra once you accumulate all these 2 voices: CC#11 - value 114
record them all performing a controlled chro- instrument groups.
matic scale at ppp, mf and fff! Then you ana- In reality, however, the result is a bit differ- 3 voices: CC#11 - value 110
lyze the sound wave and find the approxi- ent. While the effect certainly is something to
mate dB differentials between notes in each be on the lookout for, it is usually not some- 4 voices: CC#11 - value 105
part of the instrument’s various registers. thing people will perceive as “a bad thing.” If
Finally, you transfer these relations to your overdone, it can lead to the dreaded “organ” 5 voices: CC#11 - value 99
orchestral library—not exactly a cakewalk, but effect” so much of the frequency range is
truth is without proper balance, even within occupied with sounds of similar timbre and
an instrument itself, you can never go by tra- character that the sense of space between the
ditional orchestration rules or guidelines. colors is lost. This
There are too many discrepancies between is akin to the effect a painter experiences
real orchestras and today’s orchestral libraries. when he or she mixes colors carelessly, result- Fig. 1: As you add voices, lower the volume
Another technique I use when dealing with ing in a brown mess. This is a typical result using MIDI CC#11. This chart was put together for
samples of ensembles in homogenously when layering different samples from various TJ’s custom library, but it’s pretty close for all the
instrumented harmonies (such as a string libraries. Vibrato is partially to completely lost, commercially available ones.
triad comprised of three notes with a 12-vio- and it becomes akin to a thick synth pad.
lin ensemble sustains patch) in order to avoid There are ways to avoid this problem if
a buildup in volume is to use CC#11 to atten- you’re using sample libraries that offer wide complex harmonies. It’s not uncommon to
uate the output as more and more voices are selections of solo instruments, or sampled see two horns per note in a triad (comprising
added. divisi strings. In the case of solo instrument a 6-horn section), although the traditional use
Fig. 1 is a chart based on my own custom samples, it involves layering multiple tracks of of the horn section (consisting of four horns)
samples (your results may vary), and these various similar articulations to form the is one horn on each note in a 4-part chord.
have proven themselves pretty accurate with desired section size. Some people prefer a large horn section,
TASCAM GigaStudio’s linear handling of The upside is that you have complete con- others like the space that a more modest
CC#11. Again, your results may differ trol over each individual musician’s perform- horn section gives. Experiment with chords
using both section and solo samples, but
keep in mind that there are different rules in
All commercially available libraries the digital orchestration domain.
Sampled woodwind instruments are diffi-
have normalized sample pools, so the first cult to deal with. If you have too many instru-
ments of the same timbre playing in unison,
step is to rebuild the natural relationships the personal character of that instrument is
transformed into a more authoritative but less
between the patches. charming by-product. I really recommend
solo woodwind samples in place of ensemble
samples, especially in the digital domain
depending on your library and the instrument ance; the downside is that this is a lot of where balance can be corrected very easily
in question. work. It becomes more difficult to keep track using volume control.
But in general I would lower the volume of your work, and the playing techniques, Experiment both with solo and ensemble
from 127 to around 100-110 when playing intonation, recording techniques, and reso- samples of the same instrument and find out
triads with a sustain patch. That’ll make sure nance issues all become very apparent very what you like the most. A solo instrument will
you retain some of the balance in the orches- quickly. In the case of pre-sampled divisi sam- provide more nuances and a higher level of
tra. ples you’re obviously going to enjoy a more detail to your orchestration, while ensembles
comfortable work procedure. provide a slightly more sterile yet smooth
Orchestra sizes and MIDI Brass instrument sizes in libraries are usually edge—flute ensemble doubling violins for
There are distinct disadvantages and within reason: 3-4 trumpets, 3-4 trombones, example.
advantages to working with orchestral sam- and 4-8 horns. It is common that the library Woodwinds have a tendency to blend bet-
ples. The obvious advantages are the cost, also offers solo instruments for each section, ter in the real world than with samples. This is
perfection (tuning and intonation among and obviously a tuba. Some specialized sam- often because the intonation and tuning in
other things), and flexibility. ple libraries offer a piccolo trumpet, Wagner orchestral samples are too perfect, resulting in
When you’re working with sample libraries tuben, euphoniums, Bb and C trumpets, a battle for the same exact frequencies.
that have large sections of instruments, such Viennese horns, bass trombones, and even Consequently the distinct woodwind colors in
as 12 violins or 10 celli, you are stuck with contrabass trombones. There are certain traditional orchestration are easily lost.
these sizes. The act of dividing these groups libraries that even offer horns in different sec- Finally, don’t be afraid to crank up the vol-
into smaller sections (called divisi, or div. (It.)) tion sizes from pairs of 2 to 6 in unison. ume of your instrument sections to let them
is impossible, and thus you are stuck with this With brass I advise you to use solo instru- be heard, especially when using woodwind
size on single notes. ments in harmonies of three notes or more. A ensemble samples. Because of their smooth
Here the advantage is that you can easily certain synthetic sound becomes increasingly nature they tend to disappear in between the
achieve a huge sound without much effort. A audible with larger harmonies consisting of strings and the brass if the orchestration does
huge soaring string line in 1st, 2nd violins, two or more instruments in unison per sample. not offer them their rightful space.
and violas is easy to achieve with single-note The exception is horns: their mellow tone
sustains for each section. A triad will sound lends itself better towards large sections in (CONTINUED ON PAGE 63)

VIRTUAL INSTRUMENTS 39
VI
r e v i e w

Akai EWI 4000s Electric

Wind Instrument
Do you play a wind or brass instrument? The Electric Wind
Instrument is a wonderful controller for playing synths. and the
latest incarnation of this fabulous instrument now includes a
built-in synth and some new MIDI control features.

t’s hard to believe that the EWI (Electric

I
Fig. 1: The EWI 4000s is now self-contained with
Wind Instrument) has been around for a no rack unit. It has a built-in modeled analog
quarter of a century, and its EVI (Electric synth, and MIDI out for controlling other instru-
Valve Instrument) brass counterpart even ments. While it’s considerably thicker than its
longer. What’s more, it’s hard to believe—and model 3020 predecessor, cutouts on both sides (like
really great—that Akai has come out with a the one above where it says EWI 4000s) make it
new EWI model about eleven years after the feel just the same when you play it. The basic
introduction of the previous 3020 model. design has been around for 25 years, and it’s a
The big thing about this new EWI, the mature, solid instrument. Anyone who plays a wind
4000s, is that it’s self-contained—there’s no instrument can just about pick it up and play, brass
separate rack unit. It has its own built-in mod- players can adapt pretty easily, and it’s not a diffi-
eled analog digital synth, effects processing, cult instrument for beginners either.
Review by Nick Batzdorf and MIDI I/O. You can use an optional exter-
nal power transformer, but it’s normally pow-
ered by four AA batteries. It even has a built- behind, and like everything else in music it
in 1/8” headphone output so you can just pays back ten times what you put in. The
walk around and play. most mundane synth patches can come to
The 4000s maintains the identical high life when you play them with a wind con-
quality feel of the previous model, but it adds troller.
a lot of features and has a street price not
much more than half the price. While there History and overview
were reportedly 15,000 EWIs and EVIs floating For those of you not familiar with this won-
Akai Pro EWI 400s Electric around as of 1998, hopefully everything will derful instrument, we’ll start with some back-
Wind Instrument, $1000 list converge with this new model to bring wind ground.
controllers more to the forefront. They cer- The original EVIs and then EWIs were built
(street price about $700) tainly deserve it. by inventor (and virtuoso trumpet player)
It would also be great if more people Nyle Steiner starting in the pre-MIDI mid-
learned to play the EWI, because it’s really not ’70s—Steinerphones. Early versions of these
[Link] a difficult instrument to pick up. Anyone who instruments became popular in sessions
plays a wind instrument can pretty much play around the early ’80s, in fact Steiner himself
it already, brass players aren’t all that far used to play EVI in the studios. If you remem-

40 VIRTUAL INSTRUMENTS
 years ago, and sometime around the same
VI
r e v i e w
in the same breath without tonguing them
time Yamaha came out with the first of their and they’re automatically connected; play
more sax-centric WX series of wind con- successive notes and they all sound different
trollers; other companies have introduced because your breath pressure and attack will
trumpet-like controllers more recently. There be different.
have been several iterations of the EWI, and I Put another way, wind controllers are
understand that Nyle Steiner was still involved extremely expressive instruments. That’s what
with the design of the 4000s. always overwhelms anyone who tries one for
Steiner himself makes a custom instrument the first time.
called the MIDI EVI, but as a widespread
instrument it seems to be in limbo. However, Fingers
a lot of EVI players are reportedly flocking to Because of its maturity, it’s hardly surprising
the 4000s. that the basics of the EWI 4000s haven’t
So the EWI has been a mature, solid instru- changed. While a lot is new, its essential feel
ment for a long time, and I personally have is pretty much identical to its predecessor.
had a longstanding and passionate love affair Rather than the mechanical keys most
with it. woodwind instruments have, the EWI uses
capacitance-sensitive buttons like the ones
Why wind control? you find on some elevators—you just touch
The obvious answer to that question is them and they’re “on.” The instrument
Because it allows someone who plays a wind comes with a neck strap, and you also sup-
Fig. 2: Fingerings are the same in every one of or brass instrument to use his or her tech- port it by resting its grounding strip on your
the instrument’s seven octaves, so once you know nique to play synths and samplers. But there’s right thumb.
one you know them all. Then you select which more to it than that. Some wind players find the lack of keys to
octave you’re playing in by putting your thumb When you trigger a note on a keyboard depress a difficult adjustment, so much so
between two of these rollers—a very natural and controller, it follows its programmed enve- that at one point Akai offered add-on faux
intuitive interface. The bar at the tip of the thumb lope; you use a slider or other physical con- keys. Coming from a background playing
is the Glide strip, one of the instrument’s continu- troller like the mod wheel to modify the recorder, I didn’t even notice it, in fact after a
ous controllers. Normally the Glide strip turns port- sound after the initial attack—a playing/pro- couple of days I actually felt more comfort-
mento on and off, but the more thumb it feels—the gramming technique that’s obviously 100% able on EWI than on recorder.
harder you squeeze—the higher the value it sends valid. Wind controllers augment keyboards in The instrument can use the Boehm fingering
out. V.I. studios rather than replacing them, for system used by clarinets and saxes, or you can
many reasons, most notably that the vast use a simplified recorder fingering in which
majority of sample libraries are set up for key- each finger is only responsible for one key—it
ber the St. Elsewhere TV series, for example, board control… never mind that only key- never moves to cover other ones. This finger-
Nyle’s EVI playing was a big part of the sound boards let you play independent notes with ing is really easy, for example all fingers on
of JAC Redford’s terrific scores. Since then each of your ten fingers. gives you a C, and then you step up the major
many big-name jazz musicians have adopted However, wind controllers have some great scale by lifting successive fingers in order.
the instrument, including Michael Brecker. advantages. When you play a note with a Then you determine the octave you’re
The original EWI/EVI sent out control volt- wind controller, it follows your breath (if pro- playing by placing your left thumb between
ages to an analog synth that was part and grammed to do so, of course). Innately that’s any two of a series of eight rollers (see Fig. 2),
parcel of the system, in fact even the 3020m a very physical, human connection, and it giving the instrument a 7-octave range (plus
synth, which was part of the rack brain of the gives you continuous and total control over pitch-shifting). Every octave is fingered the
EWI 3020, is digitally-controlled analog. (Akai the sound without even thinking about it—
later came out with a 3030m module that just like an acoustic instrument.
used samples.) Originally, MIDI’s 128-step res- Blow progressively harder and the sound Fig. 3: The EWI 4000s next to the 3020, the
olution seemed to be a problem for synthesiz- shadows your performance, getting louder, previous model. (Normally the 3020 is white—this
ers under wind control, but that got sorted brighter, and possibly more distorted with particular one has been painted black.) Not shown:
out and MIDI was added pretty early on so louder overtones. Tongue the note (“ta”) and the 16” deep 2U rack unit you need to operate the
you could use it as a controller. it has a sharper attack; play connected notes 3020. Cool-looking instrument, eh?
A lot of professional
wind players would use
the EWI with an Oberheim
Matrix-1000 synthesizer in
the early days—an experi-
ence we can now dupli-
cate using softsynths
(arguments about soft-
ware models aside). Later
(around 1994) the
Yamaha VL1 Acoustic
Modeling synth came
along, and it and its lesser
relatives became very
popular as sound sources
for wind controllers.
Akai took over manu-
facturing and distributing
the EWI and EVI about 20

VIRTUAL INSTRUMENTS 41
VI
r e v i e w
same, so once you know one you know them sure-sensitive transducer. By definition any- notes…well, glide from one to another.
all. Some musicians coming from other wind thing you blow into has an embouchure, but The Glide strip senses how much of your
instruments have a hard time avoiding glitch- the EWI doesn’t require a special one—it’s just thumb is making contact with it, so it sends
es when they go over an EWI octave break, a matter of blowing into it. That’s another higher values when you feel like you’re
but again, I personally didn’t even think characteristic that makes it an easy instrument squeezing it harder. In V. 2, both the pitch
about it when I first learned the instrument. to learn. (Embouchure = the way you hold and Glide controllers can be assigned to MIDI
In practice, people new to the instrument your mouth when playing a wind or brass cc numbers (cc = continuous controller; the
are likely to start with the simplified recorder instrument.) MIDI spec has room for a lot of them).
fingering and then gradually incorporate Early EWIs were closed without any air The mouthpiece incorporates a bite sensor,
alternative fingerings and most likely some venting, so you had to let air out the sides of which sends a pitch or breath “blip” to trig-
extra keys to make given passages less awk- your mouth; like the previous model, the ger vibrato. Unfortunately the opportunity to
ward. Using alternative fingerings for trills is 4000s’ air system has a small vent, but letting separate this control from the breath or pitch
routine on wind instruments; the difference is air out is still part of the playing technique. information was passed by in the new 4000s
that on the EWI they’re in tune. I’ve read complaints from some wind players EWI; while pitch blip vibrato is a very good
What makes the EWI 4000s accessible to that this is hard to get used to, but here yet feature, this really begs to be a discrete con-
brass players is a new valve mode designed again, I personally have never even been con- trol in my opinion. (Windworks Design’s
by Matt Traum of Patchman music scious that I was doing it. BrainBox and upcoming CV-Midi Pro control
([Link], where you’ll find The breath pressure sensitivity is adjustable interfaces for the 3020 do allow you to sepa-
lots of info and wind-oriented soundware) with great precision over a very wide range. rate the blip and the continuous control,
that makes the fingerings brass friendly and You can set it so light that just the air pres- although you can also combine them the tra-
reverses the octave rollers. The EVI uses some- sure in the room makes it play—i.e. it’s on all ditional way.)
thing like a can that you rotate to switch the time, like bagpipes—or at the other The four controls we’ve mentioned so far
octaves, and the reversed roller direction extreme you have to blow so hard that your are identical to the ones in the 3020. If it
makes more sense to someone used to the eyes bulge and your saliva glands hurt. sounds like using them requires a lot of coor-
EVI. Very clever. dination, bear in mind that you’re only using
Since human beings don’t coordinate their Controllers them while you’re sustaining notes; they’re
fingers and breath with 100.000% precision, In addition to the breath sensor, which in actually quite easy and intuitive. The only
the instrument has an adjustable key delay the version 2 software can send out MIDI possible exception is that it can be hard to
that produces fewer glitches, i.e. it waits a lit- Breath (cc2), Volume (cc7), Expression (cc11), avoid triggering the Glide strip by mistake.
tle while before determining which keys Aftertouch, and Velocity—any or all of For that reason I always set its sensitivity very
you’re touching. A setting of maybe 25 mil- those—the EWI has several physical controls low so it only gets activated in response to a
liseconds lets you play as fast as you want that affect the 4000s’ internal synth and also very deliberate squeeze.
without generating false triggers. go out over MIDI. (Unlike previous models, There are two additional controls on the
the 4000s’ firmware can be updated via MIDI, 4000s: Octave and Hold. These on/off but-
Lungs and Akai has already added a considerable tons also function as up/down switches when
A lot of kids start off on recorder as their number of features in version 2.) you’re tweaking EWI 4000s settings and
first instrument because it’s so quick to learn. First, there’s pitch bend up and down, patch change up/down buttons when you
The EWI is even easier. That’s partly because which you trigger by sliding your right thumb move your right thumb onto a metal screw
you can play it with an even simpler finger- up or down from its normal resting point on on top of the Pitch Up plate.
ing, but mainly because you don’t need to the grounding plate onto Pitch Up and Pitch But while you’re playing, they’re lighted
develop much breath control to keep the Down plates. Then there’s a Glide strip on the on/off toggles that you engage with your
instrument in tune and sounding even. far side of the pitch rollers (you can see it in right middle and index fingers. The bottom
You make noise by blowing into the Fig. 2). This is normally set to control port- one, Octave, causes the instrument to double
mouthpiece, which directs the air to a pres- mento—press it with your left thumb and the note you’re playing an octave (or any
other interval you set) below. A
red light makes it easy to see
when either button is engaged.
Normally, the Hold button sus-
tains the first note you play sort of
like a sostenuto pedal on the
piano, so you can then other
notes over it. You can set it to sus-
tain until you attack another note
with your breath (the default) or
until you disengage the button,
regardless of what you do with
your breath. The Hold button can
also send cc64 (sustain pedal) info
over MIDI.

Fig. 4: The UniQuest editor for the

4000s’ built-in virtual analog synth.
Notice that breath control takes the
place of envelopes for the oscillators
and the multimode filters for the oscilla-
tor and noise generator. But what
makes this synth come alive is that it’s
so responsive to the EWI.

42 VIRTUAL INSTRUMENTS
 It doesn’t take long to get comfortable the Setup button at the top of the instru-
VI
r e v i e w
painted, but these adjustments are pretty
engaging and disengaging the Octave and ment, use the Hold and Octave buttons to much set and forget.
Hold buttons on the fly. The Hold button scroll through parameters, release the Setup You can use a wireless MIDI transmitter
especially is a lot of fun, opening up all kinds button on the parameter you want, and and/or wireless guitar system with the EWI,
of possibilities. You can jam over notes, play then use the Hold and Octave buttons to since it’s self-contained and runs on batteries.
double stops on string patches, and so on. select the value. That’s a great feature for performance, and a lot
There’s a dedicated button to transpose the of players will gravitate to it just for that alone.
The new body instrument to any key, which you program But the other side of the sword for studio
To accommodate all the extra circuitry, at the same way: hold, and use the Octave and use is that you can now end up with as many
26-1/4” the EWI 4000s is 4-1/2” longer than Hold buttons to scroll up in half-step incre- as four cables attached to the EWI: MIDI in
the 3020. (See Fig. 3) It also has a larger ments to Eb or down to E. Two other buttons and out, audio out, and power (although the
perimeter, but due to cutouts in the body the let you set the global and individual program four 9V batteries do last a long time). The sin-
keys feel to your fingers like they’re in the output levels; and the level of each of the gle wire clip on the 3020 only has to hold a
same locations. Actually, the right pinky keys three effects: reverb, delay, and chorus—also single multi-pin cable. On the 4000s, that
are the same distance apart from one anoth- globally and for each program. same clip can only hold one cable with three
er, but it turns out that they’re 1/4” closer to What you’re doing is reflected in of those others tied to it—not the most elegant solu-
the adjacent one on the 4000s, a subtle but early ‘80s-style 7-segment LED displays (two tion. Furthermore, you need to loop the wires
noticeable difference that I personally find of them). You know, when everything is lit up to avoid strain if they don’t have right angle
more comfortable. it forms a boxy number 8 or letter B, if the plugs.
Though black rather than white, the styling middle bar is off it’s a 0 or letter O, etc. This instrument really wants a dedicated
is essentially the same as the 3020’s, with a Initially I was concerned that all the writing cable to carry all that. If you don’t ground the
high quality plastic body and metal sides. At on the black body is molded, so you can only MIDI connections (which makes sticklers mad
just under 2 lbs. the 4000s is quite a bit heav- see the button labels when there’s a light but in practice works fine), you’d only need
ier, but it’s still very comfortable, and it feels source from the side, but there are only four to wire up eight conductors (4 x MIDI, 2 x
reassuringly solid. of them. It would be good if the Glide, Pitch, audio, 2 x power). And you’d need to locate
Getting around the EWI 4000s’ functions Breath, etc. knobs (accessible under a remov-
(CONTINUED ON PAGE 62)
is straightforward. You simply hold down able plate near the to of the instrument) were
VI
f e a t u r e

Bent Clicks
Using the tempo track to enhance performances

by Paul Gilreath

It’s curious how someone can spend countless owever, playing without the click isn’t

hours choosing articulations and tweaking individual H always possible or necessarily even
desirable with the sample libraries we
have today. Especially the larger ones are set
up to be “programmed” as much as played
notes, while ignoring one of the most essential completely in real time.
So we’re going to analyze why, when, and
elements of a living, exciting musical performance: how to program these tempo changes into
MIDI compositions after the fact. We’ll
variations in tempo. Sometimes it can take a little time assume the music was recorded to a static
click track or step-entered with the transport
stopped.
to introduce these tempo variations, but it’s very
Where
easy and well worth it. Slowing down (ritardando) or speeding up
(accelerando) can make a phrase more musi-
cal or expressive. These deviations from the
piece’s set tempo can fit in many places with-
When any musician plays without a click track, slight in a phrase, such as: where the melody is par-
ticularly touching; where the most tension
variations in tempo occur naturally throughout the occurs, often right before it resolves; and just
before an emphatic downbeat (where the
tempo usually resumes. Ritards often mark
course of the piece. If your performance skills are up the ends of phrases, sections, and pieces as
well.
to it, you may be able to simply turn off the click and Obviously some pieces want more variation
than others—an orchestral march will have a
play; those variations will be there by default. After more rigid tempo than perhaps a waltz or a
more delicate piece; it’s usually just a matter
of interpretation.
that, different sequencers have different ways of lining
Tempo tracks
up their beats and barlines to coincide with your Most MIDI sequencers have two playback
modes: using a fixed tempo setting, or fol-
freely-recorded performance. lowing a variable tempo track that you’ve
programmed. The fixed tempo setting may
be a working mode that you only use while
recording the parts. It’s easy to evaluate vari-
ous tempi, or perhaps slow down the piece to
play a difficult line.

44 VIRTUAL INSTRUMENTS
 The tempo track has various names within
VI
f e a t u r e
different sequencing programs—it’s the
Conductor Track in Digital Performer and Pro
Tools, the Tempo Track in Cubase, Nuendo
and Logic, and the Tempo View in Sonar. In
all four programs, changes can be incorporat-
ed using a graphic editor (where the tempo
changes are shown linearly and can be drawn
directly on the track) or an event editor
(where the changes are shown numerically in
a list form just like MIDI note data).
The following examples use Steinberg
Nuendo, but the technique is exactly the
same no matter which software package you
use.

How it’s done

Different musicians work different ways,
but programming changes is one of the last
things I do before mixing. Working one sec-
tion at a time, I go through the piece to
“spot” where the changes should go, drop- Fig 1. Graphical tempo editor with no tempo change and tempo at 60 BPM
ping markers as I go. Then I work phrase by
phrase.
My preference is to use the graphical editor
to make changes. This editor plots tempo on
the y-axis against time on the x-axis (in either
seconds and minutes or bars and beats), so a
piece with no tempo changes would be rep-
resented as a single straight line. (Fig. 1)
To implement a change, you simply draw a
new point on the timeline. If you draw a sin-
gle point, you will get an abrupt change; if
you want a gradual change (as in a ritard or
accelerando), you will need to insert several
points.
Let’s look at a couple of examples. First,
let’s apply an abrupt change. Decide on the
new tempo (in this case 90 BPM), insert a
point with that value, and you have a com-
pleted tempo change from 60 BPM to 90
BPM. (Fig. 2)
Now let’s look at the more common
change, a ritard. Usually you’re going along
at the set tempo, the music slows down, and Fig 2. Graphical tempo editor with change from 60 BPM to 90 BPM
then the originally tempo is resumed. We
need to insert points on the tempo editor
that correspond to these three events.
First, we define where the original tempo
ends, going into the ritard. In this example,
there’s a tempo insert point at bar 14 beat 3
(Fig. 3a) at the original tempo. Next define the
end point of the ritard where the original
tempo will resume. This is done by placing a
second tempo insert point at bar 15, beat 1
(Fig. 3b), again at the original tempo.
Finally, we place a tempo insert point
somewhere between these points to program
the ritard. This point will represent the last
two instructions: how much change happens
and how quickly it occurs.
Usually the tempo gets slower and slower
during the ritard until just before the event is
over and the first (or primo) tempo resumes.
Consequently, you’ll see the point entered
near the end of the ritard as shown in Fig. 3c.
Notice that I’ve changed the curve option to
ramp instead of jump. This allows the tempo Fig 3. Graphical tempo editor with end point of first tempo bar 14, beat 3 (15.4b)
to change gradually between the points

VIRTUAL INSTRUMENTS 45
VI
f e a t u r e

Fig 3b. Fig 3c.

one at beat 4.1 and ending of the present ritard at the point just
one at beat 4.3. This before where the 16ths should slow more. I’ll
makes it possible to insert this on the existing curve just before the
create a ritard that second 16th note of beat 4 (Fig. 5c).
slows slightly from Next, I can either insert another point just
beat 3 to 4.1, slows before my point at beat 4.4.045 or I can just
more from beat 4.1 to move that point to a slower tempo, which is
4.3, and then slows the option I choose: lowering the tempo to
the most from 4.3.0 to 15 BPM (Fig. 5d). This results in the perfect
4.4.87. When plotted ritard.
against time (Fig. 4b), For the final example, let’s enter some
the start of bar 15 events that represent unintentional tempo
Fig 3d. now occurs at 56.724” changes. Very few human beings have perfect
instead of 57.446” machine-like time. To simulate that, the goal
instead of changing immediately (like it does from above. That means the ritard occurred is to change the tempo slightly every so often.
in Fig. 2). Also, when I change the x-axis to slightly faster than the one in example 3. Subtlety is mandatory for this to work\
display time rather than beats, the last point Thus this technique can be used to perfect The easiest way to do this is just to insert a
of the ritard occurs at 57.446” (Fig 3d). your ritards and it can be used to alter hit tempo change every 6-10 bars. It’s not neces-
The next step is to listen to see if the ritard points when working with media. sary to place these changes at the beginning
works with the music. In general, the faster In our next example, let’s insert a final ritard of the bar, but they certainly can go there.
the tempo, the more successful the single- for a big finish. To create a dramatic flare, I The values of these points need to be close to
point ritard will be; conversely, this type of want the second to the last measure (#104), the original tempo, perhaps within a range of
ritard does not work as well in slower pieces. which includes sixteen 16th notes in the +/- 1.5% (total 3%) of the main tempo. And
In slower pieces, success usually depends accompaniment, to really slow down starting the changes need to be in increments of only

Fig 4. Fig 4b.

on the accompaniment and in particular the on beat 1. As before I draw a point that repre- about 1% of the main tempo.
rhythm of the accompaniment. If the meas- sents the end of the primo tempo and a point So if the main tempo is at 90 BPM, then
ure contains only slow moving notes, a sin- that represents the end of the ritard and begin- the overall range of the tempo changes can
gle-point ritard may work fine. However, if ning of the first tempo (although this step be about 1.5% faster and 1.5% slower, result-
the measure is filled with a lot of activity, you might not be necessary since this is the end of ing in a range from 88.65 to 91.35 BPM (but
might find that the music really drags toward the piece) (Fig. 5a). Finally, I insert three points you don’t need to get out the calculator—
the end. This can make eighths or 16ths as shown in (Fig. 5b). we’re just being precise for this article). The
sound strange and out of context, like a The result is a very dramatic ritard, but let’s changes shouldn’t be greater than about .09
music box running out of power. When this assume that the last few 16th notes sound BPM. Insert the changes slowly and you can
happens you need to add more points to the too fast. Up to that point the ritard sounds go back and forth between values—like 90 to
track to obtain the desired result. good, so I just need to insert a few more 89.1 to 90 to 89.5 to 88.65 to 89.5 90.2 to
In Fig. 4a I have added two more points— points. First I draw a point that represents the 90.9, etc. (Fig 6.)

46 VIRTUAL INSTRUMENTS
 VI
f e a t u r e

Fig 5a. Fig 5b.

Fig 5c. Fig 5d.

Clicking heels tions using sampling

Tempo changes don’t have the immediate technology. Gilreath
wow factor of great sounds or a well-written lives in Atlanta with
piece, but they can add an incredible amount his wife Channie and
of realism to your sequences. his two children,
Quintin and Birdie.
Paul Gilreath has composed music for feature More information
film, television, documentarie,s and video about Gilreath and
games. He is the author of the best selling book, the book can be
The Guide to MIDI Orchestration, 3rd edition, a found at [Link]-
700-page book detailing how to achieve maxi- [Link].
mum realism when producing orchestral emula- VI
Fig 6.

When the Rit Won’t Groove

f you just can’t program a ritard to sound natural, this is a simple Typically I do this manually. The way to do this is to write down the

I technique that help. It works best when the accompaniment has a

lot of movement rather than just consisting of sustained note:
Record a piano performance (MIDI only) of the problem area with-
times in your new performance where each quarter or even eighth note
falls. Then mute the piano line and go to the tempo track, viewed as
tempo plotted against beats and measures instead of linear time.
out using a click. Analyze the MIDI events to see how you play the Enter two tempo points, one for the beginning of the ritard and
notes naturally. Translate that timing into your main sequence. one for the beginning of the measure that follows it if there is one;
You’ll need to mute everything you’ve recorded from the point these points will of course be at the original tempo. Now enter a
where the ritard starts on, which you can do by automating the piano point at the first eighth note after beat one.
track’s solo button to mute everything else or by cutting all your Lower the tempo until this eighth begins at about the same time it
tracks temporarily at that point and sliding them forward out of the did in the piano performance. Insert a point at beat 2 and pull it
way (you’ll put them back afterwards, of course). Be sure to play down like you did for the eighth, then the next eighth note and so
along with the recorded tracks before they drop out, and if the music on until you reach the end of the ritard.
resumes its tempo after the ritard, keep going to ensure a natural The result should sound almost exactly like your piano perform-
musical transition. ance. You can use this technique for problem phrases or for the
At this stage you can look at the graphical or event editor to see whole piece.
how your performance lined up against the bar lines. Or you can use Another tool that can be useful is your DAW’s Tap Tempo function,
your software’s audio to tempo conversion function (Cubase’s Beat which lets you “conduct” the tempo by playing a note on your key-
Calculator, Digital Performer’s Adjust Beats, Sonar’s Extract Timing/Set board, which you record and use to build a tempo track. You can
Measure/Beat At Now, Logic’s Beat Mapping) to help automate the then go in and tweak the map using the techniques we’ve discussed
process. in this article.

VIRTUAL INSTRUMENTS 47
VI
r e v i e w

Modartt Pianoteq
modeled acoustic piano
Look Ma, no samples

Review by
Jason Scott Alexander

he idea of physically modeling acoustic

T
Fig. 1: Modartt Pianoteq as it appears when first
Modartt Pianoteq $249.00 ?
piano is nothing new. We first saw proof opened. The photos hide “advanced” parameters
(about $315 at press time) of concept in the mid 1990s with for tweaking the model and changing the sound
Yamaha’s Virtual Acoustic technology, togeth- radically.
er with promises of polyphonic hammered
Formats: Windows VST; string models that, unfortunately, were never
Mac VST, Audio Unit. released. Many terrific electro-acoustic models puted from discrete models without the use
have emerged in the ten years that have fol- of samples whatsoever. (The exception is that
lowed, but still no practical solution for a truly it uses small samples for acoustic noises such
modeled acoustic piano has been found—up as the pedal mechanism.)
[Link]
to now. Creating a realistic sounding math-based
In what its French creators at Modartt call acoustic piano emulation is no small task,
the “fourth generation of the piano” (first: though, what with all the psycho, acoustical,
Copy protection: online serial Cristofori’s pianoforte in 1698; second: elec- physical, and mechanical variables present.
number tro-acoustic in 1929; third: sampled in 1984), Pianoteq tackles each of these variables head
Pianoteq is the very first commercial piano on.
application where the sound is entirely com-

48 VIRTUAL INSTRUMENTS
The grand scheme
VI
r e v i e w
All the complex factors that make piano a
truly vivid instrument, such as the interaction
between strings, use of pedals, cabinet reso-
nance, position of the hammers, time and
space within a given room, etc. are taken into
account. In fact Pianoteq is the first virtual
piano factory, if you will, in that it can pro-
duce entirely new “brands” as well as copies
of historically poignant instruments. It does
that all in real time.
The plug-in opens to a clean, compact,
and rather attractive user interface that’s
divided into six main sections (Fig. 1).
Physical modeling is in the top half, where
you click on photo depictions (Fig. 2) to edit
Tuning, Voicing, and Instrument Design.
In the lower half of the display you find vir-
tual pedal board controls, along with an
equalizer with a graphical display and a
mediocre reverb unit (which you can bypass if
you prefer to use higher quality plug-in
effects instead). You can also adjust the veloc-
ity curve in quite some detail to suit your
playing style and controller keyboard.
When conjuring up a new piano sound, the
Voicing section will likely be your first destina-
tion. This is where you find the Spectrum
Profile editor, with sliders similar to organ
drawbars for adjusting the first eight over-
tones to define your instrument’s timbre, tone
color, and ultimately character. You can adjust
the “ringiness” of a piano very subtly; many
early piano manufacturers used to avoid a
strong seventh overtone, for example. More myriad of parameters available. Fig. 2: Absolutely every the characteristic of a
experimental synthesists can easily turn The Design portion of Pianoteq features real piano is accounted for.
Pianoteq into a harpsi-organ or piano-clavinet three soundboard controls that can also affect
with just a few mouse strokes. the character of overtones directly: imped-
Hammer hardness, another extremely ance, cutoff, and Q. Though impedance can octave higher when play harder. Both this
important parameter, determines how bril- drastically alter the length of the piano sound, and the Global Resonance are really well
liant the piano sounds. You set the hardness adjusting the piano Size (or harp/string thought out and implemented features.
for p, mf, and f dynamic levels, and also the length) parameter naturally affects both vibra-
Sitting on the bench
I have to admit, when I first played
Creating a realistic sounding math-based Pianoteq I wasn’t immediately sold on its
sound. There seemed to be no getting past
acoustic piano emulation is no small task, what the fact that its presets sounded boxy and less
present, and certainly lacking the same stereo
with all the psycho, acoustical, physical, and edge that live sample recordings have.
While it was certainly impressive that this
was all coming from a model and not sam-
mechanical variables. ples, I wasn’t going to trade in my gold stan-
dard GigaStudio collections Granted, I’m a
amount of hammer noise and soft pedal. tion time as well as timbre. A short string pretty harsh critic of piano recordings, piano
These parameters together are timbral and length produces greater inharmonicity, which sample libraries, and now virtual pianos;
percussive adjustments rather than changes means a very bell-like sound. piano is the instrument I grew up on.
to a layer of samples. The result is very realis- A Global Resonance control governs the As it turns out, comparing the presets side-
tic—not only do you feel how close the piano whole instrument, including strings, sound- by-side to my favorite Steinway, Bosendorfer,
is to you, but the mechanical behavior of the board and cabinet, while Sympathetic and Yamaha libraries was actually the down-
hammer assembly itself. Then there’s the Resonance adjusts the weight of the string fall in my initial evaluation. You see, Pianoteq
Character parameter, which essentially con- resonance between individual notes. If you doesn’t come with “branded” presets, and it
trols the irregularity of overtone intensity; press and hold a few notes (slowly so they doesn’t directly attempt to emulate any of
increasing this may surprise you with a dis- don’t make a sound) and then play some your favorite piano flavors.
tinctly organ-like sound. short notes with the other hand, you can Instead it comes out of the box fairly neu-
Modartt even points out in their literature actually hear the held notes ringing. Anyone tral sounding, or hybrid as it were. Particularly
that this is pretty heady stuff, and under- who has played a piano will appreciate how in the middle registers, Pianoteq has a ten-
standing of basic piano behavior will help you realistic this. dency to sound a little “honky” to me, not
adjust Pianoteq. Thankfully they’ve included Likewise, the Quad Effect slider adjusts the unlike the way some digital pianos sound
excellent pop-up help floaters for each in the level of that distinctive overtone ping an when played softly.

VIRTUAL INSTRUMENTS 49
VI
r e v i e w
It wasn’t until I began exploring the voic- The amazing tuning section gives you all Pianoteq to catch a flurry of staccato notes
ing and design provisions in greater detail the possibilities typically done only by piano with intermittent presses of the sustain pedal
that I began appreciating Pianoteq for its true tuners. For example, you can set diapason and treat them accurately.
strengths flexibility. After increasing hammer (415-466 Hz), different kinds of tempera- Because the plug-in operates with 32-bit
hardness for more overtones on the attack, ments (from equal to well tempered), unison internal resolution at rates up to 192kHz,
adjusting the hammer hardness and pedal tuning (for changing the timbre or color of there’s absolutely no funny digital noise, plus
response at each velocity level to suit my the sound), octave stretching, and direct the instrument produces a real progressive
Korg Triton Pro X controller, tweaking the sound duration. Pianoteq also features a pro- variation of the timbre at 127 MIDI velocities
tone color with the spectrum profile, and
adding presence through string response and
some creative tuning adjustments, I was able
to get very close to my favorite sample
libraries.
The interaction between strings, use of pedals,
In fact I was able to get the entire key-
board range to sound absolutely stunning, cabinet resonance, position of the hammers, time
and I don’t use that term lightly. The dynam-
ics are extremely convincing. and space within a given room, etc. are all taken
You can also get creative. For instance, the
ability to increase the length of the piano into account.
incrementally means that you can take it from
A size to D size, or even create a 33-foot
grand if you like. Yet because with the har-
monics change as the piano is lengthened, gressive sustain pedal, allowing partial pedal per note. A phase/repetition variation scheme
the bass notes remain incredibly solid and effects such as quarter and half-pedaling; a ensures that the hammer strikes the strings
believable. Conversely, I boxed up my perfect Sostenuto pedal; harmonic pedal; and una slightly differently with each note, and thanks
Steinway D emulation into a respectable corda (“one string,” i.e. soft) pedal. to the wonderful math going on behind the
sounding apartment-sized Heintzman piano, Pedal play on glissandos sounds incredibly sympathetic vibrations algorithms, chords
detuned Coldplay-esque stage piano, detuned lifelike—it’s as though you have your head large and small truly sound like chords played
honky-tonk, and even a tinkering toy piano. hanging over the harp. Pedaling in general is on a real piano as all the individual strings
But creative sound design is only half of spot-on. And it doesn’t just consist of the arti- interact.
Pianoteq’s bag of tricks. Its playability is sec- ficial thunking in a lot of sampled pianos. I
ond to none. was especially impressed by the ability of Left me tickled
I really like Pianoteq. Its sound is pure,
vivid, alive with responsiveness, and without
question the most flexible of any “acoustic”
piano instrument on the planet today. Thanks
to its rather modest system requirements,
Pianoteq will run on any reasonably current
desktop computer with next to no impact on
the processor, and the extremely small 8MB
size and negligible RAM requirements means
it can run on any modern laptop. (A stand-
alone version with studio-grade compression,
EQ, and reverb would be useful for “straight
out of the box” stage use—and in fact the
developers are working on a stand-alone ver-
sion; they also plan a free update this spring.)
With more likely to come, there are cur-
rently two free add-on presets available for
registered Pianoteq users at the Modartt web-
site. These include dynamic model construc-
tions of an 1812 Pianoforte Schöffstoss
recorded at Schloss Kremsegg in
Kremsmünster, Austria and a Johann
Evangelist Schmidt pianoforte, circa 1790,
recorded at Handelhaus in Halle, Germany.
The presets are meticulously modeled with
authentic tuning (well tempered) along with
all its charms and artifacts.
Though it may not capture one hundred
percent of the image that a stereo sampled
piano will, it’s the subtle real-time details that
make Pianoteq sound leaps and bounds more
realistic for solo work—and win you over
every time. VI

50 VIRTUAL INSTRUMENTS
VI
r e v i e w

Loop Librarian
Loops loops loops loops and libraries

Review by
Chris Meyer

Big Fish Electro

Magnetic Fury
sound library

like the idea of industrial music, but rarely contain a 30+ second demo plus from three

I do I find an industrial sample library that

lives up to the promise of the genre. It
seems that many try too hard to be abrasive,
to 16 component loops, ranging from two to
12 bars in length (four being the most com-
mon). Typically, one part is presented in two
clever, or obscure, while providing isolated or more variations, and there is usually an
Electro Magnetic Fury $99.95 loops that give you too little variation to build ambient loop among the components.
an entire, cohesive song around. The promo Although these kits are intended primarily
materials for this collection—promising it was to be rhythmic backing tracks, there are often
Big Fish Audio “downright dangerous” and would “blow tonal (dare I say melodic) elements, including
through your speakers” with “punishing pitched drums, vocoded rhythms, processed
([Link]) drums and beats” and “ear-shattering FX”— traditional instruments, and the occasional
made me assume we had another entrant in sequencer. Tempos range from 67 to 200
the too-abrasive category. bpm, focused around the low 100s. In addi-
platform: 16-bit 44.1 kHz Therefore, I was pleasantly surprised to tion to the kits, there is also a folder of 17
AIFF/Apple Loops, WAV, & REX2 hear that this library hews much closer to the “atmospheres”: lovely 30 second to two
intelligent industrial mindset, combining minute droning, ambient, mostly tonal beds
files
ambient and exotic elements with a pulsing with an occasional rhythmic pulse that would
groove. make great intros or bridges
As is the case with most Big Fish loop As I hinted in the intro, what struck me
license: May be used in the libraries, Electro Magnetic Fury repeats its most about this library was that the atmos-
licensee’s own “derivative” live content in AIFF (Apple Loops), WAV, and phere wasn’t nearly as harsh and monodi-
performances or recorded com- REX2 flavors. The AIFF version weighs in at mensional as I expected. There is indeed an
just over 1.5 gigs of 44.1kHz 16-bit stereo abundance of strong grooves and distortion,
positions, but not in a sample or files. but some pieces are downright ambient,
music library. There are 95 construction kits listed by while others wander into house or even
tempo and occasionally key. The kits each lounge territory. There is also more of a feel-

52 VIRTUAL INSTRUMENTS
ing of understated mystery, dread, or power result, quite often you will need to create
VI
r e v i e w
than pure chainsaw mayhem. Many kits movement by switching elements on and off,
throw in surprising elements such as African rather than being able to pull up evolutions
chants, Middle Eastern percussion, Far Eastern of the existing parts. Many who create music
strings, military snares and toms, ambient in this style are already used to working this
hits, and a lot of bandpass-filtered beats. way. For those outside of this genre looking
The only thing I can hold against this col- for additional spices and flavors, I think you’ll
lection is a lack of variation, as most compo- find a surprising number of useful layers and
nents are only presented in one version. As a elements to include in your own work. VI

Big Fish Mahadhi: African Rhythms

loop library

et’s set the record straight from the proximity effect by being miked too closely,

L start: This is not a pristine set of careful-

ly recorded solos and grooves per-
formed in isolation to a click track by well-
while the talking drum sounds a bit muffled.
This can make you cringe while listening to
loops in isolation, but it becomes far less of
paid session musicians in high-end studios, an issue when mixed into a track.
preserved in 24-bit 96 kHz fidelity. The vast Someone used to a steady diet of Western
majority of the loops were recorded in mono, grooves might listen to real African rhythms
close-miked with varying degrees of fidelity in and hear just a cacophony of competing
live ensemble performance situations, com- instruments, assuming there’s no way that the
plete with some mic bleed and the occasional timing is tight or would line up against a
grunt. steady Western pulse—hey, even I did at first.
And you know something? It’s great. My In reality, all those mad percussionists are
left brain says it shouldn’t be, while my right cleverly hitting in-between beats and placing
brain says shut up and dance. emphasis on notes other than The One.
All loops are available in AIFF, WAV, and As a result, these loops work surprisingly
REX2 formats; I tested the AIFF versions, well as layers mixed in with straight-ahead
Mahadhi: African Rhythms which weigh in at just over a gig. There are Western drum kit grooves, as they hit the
$99.95 27 construction kit folders labeled by tempo. spaces in-between, rather than fighting over
These include from five to eleven subfolders the accents. Although the occasional groove
that break out each song by instruments, plus needs some timing correction to bring it back
Big Fish Audio a minute-plus 24-bit stereo demo that illus- to a metronomic beat, the vast majority are
trates various combination of instruments. actually in the pocket.
([Link]) (Don’t be fooled, though, as most compo- Another nice thing is that most of the
nents are 16-bit mono.) loops are closer to backing grooves with vari-
Each instrument folder contains anywhere ations rather than solos, again making them
platform: 16-bit 44.1 kHz from one to 47 mostly 2-bar looped phrases work well as layers supporting other rhythms.
AIFF/Apple Loops, WAV, & REX2 in 4/4, tallying up to over 1800 loops total. Stripped of their melodic instruments, the
With the exception of the occasional kalimba “songs” come across more as moods or fla-
files
folder, you get just the percussive instru- vors, also resulting in them being more flexi-
ments, although you can hear through the ble than anticipated.
mic bleed that a full ensemble including Honestly, my initial impression of this
license: May be used in the singers was actually performing at the time library was not good, but now I expect to use
licensee’s own “derivative” live these were recorded. it quite a bit for texture and spice. While
performances or recorded As noted, sound quality varies: the conga recording live like this might have caused the
and djembe have particularly nice transients, fidelity to suffer, the flip side is that the per-
compositions, but not in a sample plus the bembe and ogenne have appropri- formances are exuberant, and authentic emo-
library. ately present rattles and buzzes, but the shak- tion is all too hard to find when it comes to
ers and kalimba suffer from exaggerated samples. VI

VIRTUAL INSTRUMENTS 53
VI
i n t e r v i e w more
online
[Link]

MIDI MOCKUP MICROSCOPE

In this installment of our series on composers and how they did their
MIDI programming, talented composer Nathan Furst discusses “Heart of
the Baja Part 2” and “Shark Attack.” Download these cues at
[Link] and follow along.

At only twenty-eight years old, Nathan Furst has Tell us about your rig and your sam-
ple libraries.
That’s a crazy thing. It’s always in a state of
already begun to carve out a niche for himself as one flux. I envy the people who run out and buy,
say, four or eight Vision computers [custom
of Hollywood’s elite young composers. Nathan has machines from [Link]] and
then don’t touch their systems for another
developed a reputation for blending world instruments four years. I wish I could live like that. I
can’t—it drives me insane.
and electronica soundscapes seamlessly with the So it’s constantly changing. Currently I
have just taken four computers out of rotation
strength and elegance of a symphony orchestra. He because they have been causing too many
problems, or they don’t hold enough memo-
has composed themes and underscores for over 20 ry, or whatever it is. I just built three comput-
ers—myself actually. I went to Fry’s, bought a
motherboard, bought a CPU, bought one of
film and television projects. This signature style has those metal rack chassis that look like the
Vision computers, and just built it. So I have
earned Nathan three nominations for his outstanding these three new machines.
I’m currently starting a project, and I’m
original film and television compositions. going to be using four PC farms, with a fifth
that kind of floats in there if I need it. I have
two G5 dual 2.0s. One of them I don’t use
that much; it basically sits in the background
by Frederick Russ as a potential node and a potential [Native
Instruments] Kontakt 2 secondary.

54 VIRTUAL INSTRUMENTS
 i n t e r v i e w VI

For the most part, it’s just sitting there as a Fig. 1: Nathan Furst has come up with a very clever way of navigating his huge orchestral template (34
back-up computer. I try to keep it as near to tracks just for violins alone), as shown in this Logic screenset. See the shrunken window with the colored bars
my original G5 as possible, so if my first G5 on the left? That’s a second Arrange window saved as part of this screenset, and the bars are folders.
goes down I take the first one to the shop All his string tracks are packed inside the dark grey Strings folder, the Woodwind tracks are packed inside
and slide the second one over and keep work- the sage green folder, and so on. Clicking on (in this case) the Strings folder opens up the main Arrange win-
ing. I got bit once where I had a Version 1 G5 dow to all the string tracks.
and I was in the middle of a deadline, and The reason this works is that the main Arrange window is Linked to the same level, meaning that it jumps
the computer went down and I didn’t know to whatever is being selected in any other window. That’s the little yellow chain link icon at the upper left; to
what to do. So I literally bought a new com- prevent the shrunken Arrange window with the folders from scrolling to the current location, it isn’t linked.
puter. Also note the locked markers corresponding to different events in the show he’s scoring.
I kept going, but of course there was
another hump to get over: installing the soft-
ware. Meantime, my deadline is coming up Logic sends MIDI Time Code to Pro Tools so I audio (by soundproofing it).
on me. Ever since, I try to keep a secondary can go to bar 40 and the picture will follow There’s nothing that enters the real world
sequencing machine that is near to my first along—as if the video were in the sequencer. At except for what’s coming out of my MOTU
one in case it goes down. the very last stage when all my MIDI tracks are 2408 interface to my monitors, which are cur-
So right now I am using four PCs, which is playing simultaneously, I just bust the masters rently Mackie HR824s. I like them. A lot of
pared down—I originally had six. And then I out to Pro Tools, using it as a tape machine. It hardcore engineers call them the “composers’
have a dual 800 Mac that does nothing but locks perfectly in sync, and I go into Record monitors” because they argue that it is not a
run Pro Tools. mode, and that’s how I have my sends that get true flat sound and that it makes everything
delivered to the stage. It’s instant—it’s so fast, I sound good. They claim there is a lot of
Which Pro Tools systems is that you can almost mix in real time. sweetening in the monitors—which is proba-
are running? Many times on a film they want the per- bly true, but I don’t care because I like it.
It’s just a basic Pro Tools Digi 001. While cussion track separated or they want the choir
I’m writing, Pro Tools runs all my video. I like separated. Having the Pro Tools system and Are you using a mixing controller?
to keep video off my main machine. working it this way allows me to do that. I have a short-lived Logic Control. It’s a
Whatever sequencer I’m writing in, I great little box, but I have to be honest, I
inevitably bring it to its knees, because I’m What kind of monitoring system do don’t go to it that much. I’ve been using
writing these gigantic orchestral things. I you use? Logic for 12 to 15 years.
might have Space Designer [the convolution Probably not the best one! I just do what
reverb in Apple Logic Pro] and a couple of works best for me. I’m a big fan of keeping Let’s talk about “Shark Attack.” From
[Spectrasonics] Stylus instances on there— the room as lean as possible, and I like that I 0:01 to 0:05, how did you get the brass
plus I try to work at a low latency. have successfully driven out any real world to sound that way?

VIRTUAL INSTRUMENTS 55
VI
i n t e r v i e w
playing, and then in the
middle of the cue walk to
the booth, it sounds very
different.
So I don’t really think in
terms of Sonically how it
would sound if I was physi-
cally standing there, I think
in terms of production.
How do I produce this cue
using my favorite com-
posers and my favorites
scores as a standard?

So you’re listening to
the orchestra and not
necessarily trying to
emulate it?
Oh absolutely emulating
it, absolutely. Some of those
guys have a great sound. I
was absolutely trying to
emulate the production
sound. Not necessarily any
one in its entirety—not only
a John Williams sound, or
only a Don Davis sound,
but sort of taking the things
Fig. 2: This is the part of the same screenset shown in Fig. 1 that’s displayed on a second monitor (as you I love from all of them—then take what you
can see behind Furst in the photo with this article). One of the things it shows is the wonder of remote access believe you do, and incorporate all of that.
over ethernet.
At the same time that clicking on the Strings folder brings up the string tracks shown in Fig. 1, on the sec- You have a lot of clarity in this cue.
ond monitor it brings up an “adaptive mixer” with channel strips for every track that’s being used in the For me anyway, some of that is the writing.
Arrange window. I try to write—orchestrate—in the sense of
Note that there’s room in the screenset for two remote computers to be displayed above the mixer. The thinking of it as a spectrum of light.
machine at the left is a G4 Mac running Pro Tools LE, which Furst uses as a mixdown machine for recording Everybody sort of gets a space. And absolute-
stems (since Pro Tools is usually the delivery format); the one at the right is a Windows XP machine running ly, somebody can cross over to the other per-
Native Instruments’ Kontakt 2 sampler (you can see the ethernet address [Link] at the top of the son’s space, but I try not to do it at the same
window). time. That helps create a sense of separation.
Clicking in either remote computer window activates it for keyboard and mouse control just as if it were
running on the local machine. The response over a network is noticeably more sluggish than a wired moni- You like panning wide, I’ve noticed.
tor, but it’s perfectly fine for recording stems or just loading sounds into a soft sampler running on a slave Like with the custom library you’re
machine. working on, everything is panned
wide—the violins are way to the left.
Definitely, and maybe a little too much—
Actually I wrote that in, I think, the sum- playing a lot with reverb tones and rooms maybe. I’ll say I naturally go a little wide. For
mer of 2002. At the time I think I had one and sort of just discovering how I was going me it helps me to hear the orchestration a lit-
machine that had the Sonic Implants stuff on to fit these things into the same space. tle bit better, and also when things are really
it and SAM Horns. That’s the way I started, which is essentially wide, inevitably in almost every score I do
the way I work today—basically having one there is some sort of, shall we say, a specialty
And those cool trombones? “pre-delay” sort of room reverb on a buss, and element.
That would have been me custom re-pro- then everybody gets that buss send. It’s pre- Lately I have been hired to do a lot of
gramming the SAM stuff, layered with fader. So essentially every single instrument has world scores—big orchestra, but there is
Miroslav Vitous. its own mix to do, it has its own wet fader and always a sort of a Mideastern or world ele-
its own dry fader. Obviously the trumpets ment. That stuff tends to hang in the center a
How did you do that? would get a lot more wet and almost none of little more, and the orchestra is off harder
I grabbed the original files and I basically the dry signal—that kind of a thing. panned. I feel I’m able to make sense of what
built, like a 20-key keyswitch program. is happening a little bit more, as opposed to if
Everybody thought I was crazy, but now it’s So you’re approaching it like you’re everything is generally hanging around the
like the norm. At the time, EXS [the sampler standing there—the trumpets are far- middle.
built into Logic Pro] was really the only instru- ther back, so you’re kind of thinking I’m not really re-inventing the wheel, I’m
ment that allowed you to go that deep. I about how the reflections go? emulating my favorite scores and my favorite
would just go in, grab the original files, and I was basically thinking What should it productions. I guess I am re-inventing the
remap them. sound like? I don’t know if this is “right” or wheel for myself in the sense of making it
Usually I would have a starting point to not; I wasn’t thinking of how it would sound happen completely in the room with no
successfully convert, say, an E-mu patch, just if I was standing there, because that is sort of money and no recording schedule, or what-
so I would get the key mapping done correct- irrelevant. If you’ve ever heard a cue or an ever. You’re basically faking it, but absolute-
ly. And then from there, just start making lit- orchestra in a recording studio, if you walk ly—create real estate and create space.
tle tweaks with the filters, adding buzz, and back and forth from the stage while they’re

56 VIRTUAL INSTRUMENTS
 Do you remember what you were 2”—what is that, a movie?
i n t e r v i e wVI
and keep incorporating it. The theme helps
using back then for ambience? Yes, it was originally a TV movie, like a spe- me to be able to write four minutes a day.
Yeah, it’s terrible. Now I can’t stand it, it’s cial event. It was aired on NBC around the 23 On top of that, some of my favorite scores
way too much ambience. It’s way too washy. I of December. If I were to guess, it’s about 80 work that way. There’s a sense of consistency.
was basically using the Waves Gold stuff minutes of music—maybe 90. From spot to
because that was what I had available at the delivery was about four weeks. You have some interesting ideas
time— the TrueVerb plugins for the room and about creating realistic performances.
using a separate hall verb for the hall. [Audio Oh my goodness. So you were averag- When I am playing something like fast stac-
Ease] Altiverb was out, but I didn’t have it ing... cato stuff, I also like to layer it with half note
working then. …a few minutes a day. I find two minutes a trills and bury them—anything that will help
I had made my own TrueVerb preset for day to be very comfortable—I can hang out, I me create bad intonation. Because if you
each section. The strings had one to represent can write the music I want to write. I can were to have string players playing those pas-
them being in the front of the orchestra, orchestrate it and experiment with the sages, there is no way in hell they are going
woodwinds a little behind that, and the horns orchestration if I am on a two minute a day to hit the tone. It’s going to be a mess.
were way in the back, as were the trumpets. schedule. If I am on something worse, which So even though it’s a fast staccato passage,
It was an experiment, shall we say. tends to be happening more and more—like I’ll double it with the trill pass and bury it, so
if I have to come up with four minutes a day you don’t hear perfect intonation. If you are
Now you’re using convolution reverb. or more, then I am just writing it the best I trying to create a sense of realism, I think it’s
I just use Space Designer [the convolution can. So pretty much when it’s done, it’s done. a common mistake to try and write runs in
reverb built into Logic]. I like Space Designer staccato, or to write runs with a short note
a lot. You don’t have time to second guess patch of some sort. I prefer to write mine
This is sort of my little rule. To me there are yourself. with legato, doubled with trills, doubled with
way too many options. So if it requires extra No, there is no time to second guess your- trem. I find that for me, those are the most
things that are unnecessary, like hardware self, let alone experiment. On a creative level, successful runs.
dongles or if you load one instance and it and also on a practical level, I work a lot with I had rebuilt some of the runs in that cue
soaks up half your CPU, we’re done. There themes—I try to create themes because I find from the old Roland libraries. For things like
are too many amazing options happening out that is an appropriate tool as well as a useful the woodwind runs that are happening in
there to have to tolerate things that are little one to be able to recall characters’ themes con- there, those are probably not performed—by
annoyances. stantly and consistently throughout the film. me. They are probably just runs I grabbed
So when I’m writing music, I can create a from a library and reprogrammed in EXS. But
I realize that it is an older piece, but mood with my orchestration, and what is I will play with the runs.
it still sounds cool to me. So “Shark happening in certain sections rhythmically— So if it is a woodwind run that just goes
Attack” is from “Christmas Vacation sort of sneak around and play with the theme up, I will double it myself by, say, playing a
piccolo with it. Inevitably I’m not fast
enough, and that produces the sound I want.
Fig. 3: While note-on velocity has been recorded in the sequencer (as evidenced by the orange and red I strive for imperfection.
notes in Logic’s graphical Matrix editor), Furst turns off the velocity response in all his sample programs,
instead relying on MIDI continuous control changes to create dynamics. The blue pane shows the fairly radi- I want to talk about breathing
cal mod wheel riding that shapes each note in this passage. strings and breathing life into strings.
This screenset also has a notation editor open; Logic’s windows are interactive, so changes in one are You have a small cue here—“Heart of
immediately reflected in all the others. the Baja, Part Two.” You’re really emot-
ing.
That would have been
a combination of many
things, including another
custom library. But for me
the real trick is that with
the exception of staccato
or short notes—even with
marcato stuff—I never use
a velocity-sensitive patch.
To me that is not what it
would sound like. The first
thing I tend to do when I
load a string sample, I dis-
able any response to
velocity.
Then I rebuild it with
crossfades only on the
mod wheel to help me
get in and out at the very
tail end of it, almost like
you can’t really discern
when it went away.
Always, everything going
in and coming out is a

(CONTINUED ON
PAGE 62)

VIRTUAL INSTRUMENTS 57
VI
r e v i e w

Steinberg Cubase 4
A look at the update to this massive
sequencing package

Review by Mark Jenkins

teinberg’s Cubase digital audio

S
Fig.1: The main Project window in Cubase 4
sequencer is especially popular in its handles audio, external MIDI, and internal virtual
native Europe, but it also has a lot of instrument tracks with absolute equality. It can be
Steinberg Cubase 4, $999.99 users in the US. The program has undergone made as plain or as multi-colored as you wish.
constant development for a lot of years, and
that shows in the depth of its features and
Cubase Studio 4, $499.99 refinement. revision. The same large Project window (Fig.
The last version of the massive Cubase 1) handles audio, external MIDI, and internal
sequencer was SX3, but now the SX prefix virtual instrument tracks with absolute equali-
[Link], distributed seems to have been quietly dropped, so the ty, and can be made as plain or as multi-col-
new revision is referred to simply as Cubase 4. ored as you wish. A large Transport window
by [Link]
There’s also a Studio version with a stripped- can be positioned anywhere, helping to make
down feature set, a much lower price point, the package ideal for use with dual monitor
and an upgrade path to the full version. set-ups.
Formats: Mac OSX 10.4 (G4 or The fact that Cubase (as we’ll simply call But beneath the hood Steinberg claims 50
better) including Intel Macs, and Cubase 4 from now on) runs on both Mac or more major revisions, and there are cer-
Windows XP Home/Professional and PC offers some flexibility, and a Steinberg tainly many new virtual instruments and
USB key—omitted from some of the compa- effects included, offering thousands of new
ny’s simpler packages—is included. You’ll sounds. Let’s take a look at some of these
need a DVD drive to install, followed by an additions and revisions in more detail.
Copy protection: USB dongle
internet licensing procedure.
(comes pre-licensed; upgrade As always with major revisions of complex Mixing and editing
requires online update) sequencers, the manufacturer lists scores of Cubase was always pretty flexible as
updates (as well as updates to the way it han- regards final mixdown, but the addition of a
dles scores), but the basic look and feel of new Control Room Mixer window (Fig. 2) will
Cubase has not changed very much in this help interface the package better to the out-

58 VIRTUAL INSTRUMENTS
 simply using drag and
VI
r e v i e w
introduction of this new track type is a huge
drop. The individual pre- improvement.
set name of the plug-in
effect is also now visible MIDI and VST
(like “BigFlange6” rather There’s little change to the way the new
than just the name of the Cubase handles MIDI except in the area of
effect). scoring. The settings dialog has been simpli-
Some editing functions fied and the score editor now has clearer
are also improved. In ear- icons. There are two new score fonts—”Jazz”
lier versions of Cubase it and “Classic”—and song lyrics can now be
was difficult to edit con- imported from a MIDI file.
troller data alongside the In fact there are several improvements in
related notes, so for the area of file and preset handling. MP3
example if you moved Surround (which is compatible with MP3
some notes you may stereo) can now be imported and exported,
have lost contact with and a new media management system called
the vibrato that accom- MediaBay makes it much easier to search for
panied them. Now notes and modify files and preset sounds within VST
and controllers can be plug-in instruments via the “SoundFrame”
edited together, which is Universal Sound Manager.
a huge improvement.
There’s a new Trim
Tool to trim start and end
points for multiple Beneath the hood
events, which is going to
be useful for complex Steinberg claims 50 or
cues. A new Preferences
setting allows event more major revisions,
Fig. 2: The new Control Room Mixer features selection to select the relevant Track automat-
Talkback and Headphone busses, External Inputs, ically. and there are certainly
and Control Room Monitor busses, all without hav- In fact there’s a whole new type of track
ing to leave the software. too: the Instrument Track, combining a MIDI many new virtual
input and an audio output for VST instru-
ments into a single track and mixer channel. instruments and effects
side world. Now there’s a Talkback and a One of the most confusing points of the old
Headphone buss, External Inputs and Control Cubase up to SX3 was that a lot of the mixer included.
Room Monitor busses, assuming of course channels were doubled—for a virtual instru-
you have the external hardware to take ment you’d have one for its MIDI activity and
advantage of all those. effects, and one (often far removed from it)
With up to four Studio Send busses per for its audio output and audio effects. In other words, Cubase now accesses and
audio channel, the whole cosmetic appear- This made it hard to pin down exactly searches the preset list of any virtual instru-
ance of the mixer (Fig. 3) has been sharpened where an instrumental sound was coming ments in use, and also adds a preview func-
up so it’s easier to read, and insert plug-ins from, and made for big mixer displays too tion that if you select a looped sound, plays it
can be copied from one channel to another wide to fit on one or even two monitors. The at the current project tempo. The MediaBay
browser can help search for audio loops and
clips, MIDI Files, video files and entire Cubase
project files on your system too.
The greatest excitement though probably
lies in the new Cubase VST (Virtual Studio
Technology) instruments and effects. In fact
the standard has now been revised to VST3,
and some 17 original plug-ins having been
updated accordingly. VST effects now include
mono, stereo, and ping-pong delays, nine dif-
ferent dynamics controllers including com-
pressor and expander plus vintage and multi-
band compressors, nine modulation effects
including chorus and autopan, an octaver and
tuner, plus mono-to-stereo and stereo
enhancer routines.

Fig. 3: With up to four Studio Send busses per

audio channel, the whole cosmetic appearance of
the mixer has been sharpened up so it’s easier to
read, and insert plug-ins can be copied from one
channel to another simply using drag and drop.
The individual preset name of the plug-in effect is
also now visible, rather than just the effect.

VIRTUAL INSTRUMENTS 59
VI
r e v i e w

Fig. 4: The Prologue analog-style synth. erful EXS24 sample player built in. Fig. 5: Spector is a spectrum filter synth, one of
These new VST instruments have more the new instruments included with Cubase 4.
than 1,000 instrument sounds, drum kits, and
While most of these are familiiar, the VST effects to offer, so it’s becoming increasingly
instruments are completely new. Prologue possible to compose entirely using internal compliant, meaning that it runs on PowerPC
(Fig. 4) is an analog-style synthesizer, Spector instruments rather than having to spend and Intel Macs.
(Fig. 5) is a spectrum filter synthesizer, and more money adding plug-in instruments. New additions such as the Control Room
Mystic is an impulse/comb filter synthesizer. Steinberg obviously still has other soft instru- Window and the Media Bay/SoundFrame
These are all based on individual pages of the ments available—their high-quality Steinberg browser (Fig. 6) combination for searching and
flexible Steinberg D’Cota softsynth, and offer Grand piano for example—but particularly at accessing the presets of VST instruments and
various numbers of oscillators and sound syn- the price of a package like Cubase 4, it’s only effects running within Cubase should speed up
thesis techniques together with hundreds of reasonable to expect to find a wide range of the work process considerably. On the subject
preset sounds, promising much greater syn- powerful instrumentation included. of speeding workflow, for those who like to use
thesis power than the simpler A1 and other key commands as much as possible and avoid
analog-style synths built into earlier version of Progression the use of the mouse, Cubase 4 has a com-
Cubase. Steinberg has made a lot of progress with pletely revised set of key commands;
Halion One in contrast is a stripped-down Cubase 4. A full printed handbook is now Deactivate All Solo and Unmute All are going
version of Steinberg’s Halion sample playback supplied, not just a set of PDF files on disk, to be particularly useful, as are being able to
software. This is a very simple instrument with and between this clearly written 600-page select altenative mixer views and open the
envelope attack and decay, filter cutoff and monster and the 75 pages of Getting Started Control Room, MediaBay, SoundFrame, and
resonance settings, however its “ROM” info, Cubase 4 is not going to be difficult to Loop Browser windows quickly.
sounds come from the Yamah Motif key- understand. The cosmetic appearance has And if you want to limit the budget a little?
boards. This helps put Cubase 4 more on a changed, but not all that much (the new VST Cubase Studio for around half the cost offers
par with Apple Logic 7, which offers the pow- instruments look pretty sleek though, and many of the new improvements. It comes
rather similar to their Apple Logic equivalents) without the Spector and Mystic instruments,
and some major annoyances such as the split- Surround facilities, or full MediaBay searching,
Fig. 6: The Soundframe browser, which is func- ting of mixer channels between the MIDI and and it has fewer VST effects. But it includes
tionally much like Native Instruments’ Kore, makes audio facilities of virtual instrument tracks the Halion One sample player and Prologue
it easy to manage and categorize large libraries of have been eliminated. Most importantly, analog-style synth, and it has similar multi-
sound by name. Cubase 4 is Universal Binary/Intel Mac OSX- tracking and scoring abilities.

Score?
Cubase works smoothly and reliably with-
out any needing any settings in the back-
ground, and many of the layout aspects that
were less than perfect—such as the channel
split between MIDI and audio facilities of vir-
tual instrument tracks—have now been
removed. The new VST instruments in Cubase
4 are powerful and flexible, and having a
built-in sample-playing instrument is a mas-
sive bonus.
Users on a limited budget can readily
upgrade from Cubase Studio, while users on
an unlimited budget can one day upgrade to
the even more audio/visually-oriented
Steinberg Nuendo. Cubase 4 is a strong can-
didate for either Mac or PC users as the cen-
ter of any recording studio. VI

60 VIRTUAL INSTRUMENTS
 random
tip
Importing
Yamaha
DX-series
patches into
Native
Instruments 1. First choose “Import SysEx” from the File menu

FM8
t’s hard to imagine anyone who worked with

I music technology in the 1980s and didn’t own

a Yamaha DX-series instrument at one time or
another, most likely a DX-7. That means many
of us have old projects that use those sounds.
It also means that billions of people were
programming sounds for DX instruments, and a
lot of those sounds are still very good (never
mind that a lot of them are also still very bad).
Native Instruments FM7 and FM8 have a great
feature: the ability to import DX-series sounds
and play them back at 2006 sound quality.
Fortunately getting those sounds into FM8
(or FM7) is a lot easier than choosing which
ones to import. FM8 can accept System
Exclusive (SysEx) Data from Yamaha’s DX7,
DX7II and DX200 synthesizers and convert
them into its own parameter format. 2. FM8 converts a DX7 SysEx bank to the .ksd format
There are two approaches; which one you
choose will depend on whether or not you
already own a DX series synthesizer.
A. Connect your DX-series synthesizer to your
computer via a standard MIDI interface. In
stand-alone mode (rather than using it as a
plug-in) FM8 will automatically receive any
MIDI SysEx data, in other words the sounds just
go in over MIDI. Single Presets go into the edit
buffer and you must save them manually; Banks
are converted into .ksd files and are saved auto-
matically. It’s not possible to receive SysEx data
via MIDI when running FM8 as a plug-in due to
the current limitations of the plug-in standards.
B. Load SysEx data from a file. To do this,
simply click on the Import SysEx button in the
File menu, navigate to a compatible SysEx file,
and open it. Note that SysEx files will have a
.syx extension under Windows. On the Mac, it’s
usually a good idea to add this extension to the
file before attempting to import it. VI
3. The converted DX7 SysEx bank appears in the KoreSounds browser for use in FM8.

VIRTUAL INSTRUMENTS 61
VI
i n t e r v i e w
NATHAN FURST two parts at a time. Say, I’ll play a couple of What about brass?
(CONTINUED FROM PAGE 57) parts on the cello and just move the bottom Same thing. I always, always, use con-
note that I played down to the basses. Most trollers 1 [mod wheel] and 11 for virtually
combination of sample layer crossfading on of the samples at that point are going to be everything, all the time. Sometimes this pro-
the mod wheel and controller 11 to sort of in combination of VSL [Vienna Symphonic duces a very noticeable, dramatic effect, and
duck in and out. To me that’s the best chance Library], a different custom library, and Sonic sometimes very subtle. I find that the instru-
you have to making your strings sound real. Implants. ments sound the most real when they have
And I love, love, love Sonic Implants. It’s small fluctuations in dynamics—which trans-
“Though None Go With Me”—that’s a gorgeous and it’s interesting because it’s actu- lates into layers as well as the timbre of an
more recent piece, isn’t it? ally very “under-programmed.” Most of the instrument when going from ƒƒ to mƒ and
Yes, that was another TV movie I did prob- patches have no velocity layers, it’s a single then back again.
ably ten or eleven months ago. I can’t say it’s thing. If there are multiple velocity layers it’s I mostly do this as I go. Very rarely do I go
the best example of my MIDI production only one or not very many—it’s not heavily back and add more CC info the next day or
work. That score is about 100 minutes of programmed. They got the timbre right when something. I can’t move on to the next bar if
music, and I scored that in two weeks. they recorded it. To me it’s a gorgeous the legato line doesn’t have the delicate tail
I play in the line I am using in the violins, sounding library. out that I want to hear. OCD I guess! VI
and just quickly flesh it out. Because I think in I also use a lot of SAM. I love everything
terms of intervals, sometimes I have to play that Maarten Spruijt does.

don’t see in the screen dump is that all the And that right there is precisely the point:
EWI
(CONTINUED FROM PAGE 43) parameter settings are shown when you attached to an EWI, it doesn’t take much
mouse over them, so for example what looks more than a sine wave to sound really good.
right-angle MIDI plugs, which aren’t all that like a totally rudimentary delay can be set in We have at least one article in the works on
easy to find. 10 millisecond increments up to 1.27 sec- programming instruments for wind control.
onds. It’s the way to go.
The synth But the reverb is what it is—you control Also, it’s worth pointing out that the built-
The EWI 4000s contains your basic mod- the time, density, and high frequency damp- in synths in previous EWIs were also quite
eled analog synth. As you can see in Fig. 4, ing, and of course the level, but nothing as conventional. About the only thing the 4000s
the screen dump from the included UniQuest sophisticated as, say, the type of room. You’ll doesn’t have is an external input for control-
editor for Mac and PC, it has two identical want to use a studio-grade reverb with less ling other synths alongside the built-in one.
oscillators, each with faders for sawtooth, tri- grain for recording applications; this is a very
angle, and square waves. It also has a noise mono reverb for a mono instrument, Conclusion
generator, which you can use for breath noise designed for live use. After 25 years and several generations with
or just for noise in the traditional way. While the sound of this synthesizer is good, all the glitches worked out, the EWI is a
Notice that in place of envelopes, the oscil- there are some nice factory presets, it’s a lot tremendously satisfying instrument to play.
lators and filters are controlled by breath. more versatile than you might gather from This latest one has a lot of new things going
There’s also a formant control for the oscilla- looking at the screen, and it’s great having for it, not the least of which is that it’s now
tor filter with woodwind and string settings. good sounds built right into the instru- down to a more accessible street price.
The delay/chorus/reverb (in that order, not ment…and the last thing I want to do is sell it The EWI is an instrument that opens up a
the way it appears) effects section is quite short…this is really a pretty traditional synth whole new world of synthesizer control. I
basic, but it’s fine for what seems to be its pri- that wouldn’t turn too many heads if it can’t recommend it more highly. VI
mary application: live performance. What you weren’t attached to an EWI.

VOLTAGE CONTROL The Program 54, Rolling Thunder, uses the drum kits to ethnic percussion to wild and
(CONTINUED FROM PAGE 18) same basic approach as Lightning Strike, but wacky sound effects can be generated using
the Attack time of the Amplitude Envelope that annoying stuff we spend so much time
Envelope; they’re essential in the creation of has been increased to eliminate the crack and trying to eliminate.
this sound. Though the filter envelope con- the LP24 filter is used to deepen the sound. In So the next time someone asks you about
trols the timbre of the sound, it is the amplifi- addition, the two LFOs are enabled, each the signal-to-noise ratio of your system, puff
er envelope that truly shapes the sound, routed to the filter, but programmed with dif- out your chest, crack a big smile, and tell them
allowing the crack to happen immediately at ferent speeds and amounts. This is what cre- you’ve got the noisiest system on the plan-
significant volume, while the thunder gradual- ates the “rolling” effect as the thunder plays etæand a host of great sounds to prove it.
ly fades to silence. through.
The settings of the Release segments of One final note: Filter Tracking, which raises Michael Marans, often accused of being over-
both envelopes allow the sound to be played the filter cutoff level as you play higher up on ly noisy, wishes to inform VI readers that no ani-
in two different ways. Strike a key and imme- the keyboard, is enabled. This allows you to mals were harmed in the creation of this article.
diately let go, and you’ll get a crack followed play deep, low rumbling thunder at the low However, several windows were shattered and
by thunder. Hold the note and you’ll get a end of the keyboard, and thunder that his earthquake insurance premiums were raised
crack that gradually turns into rain. evolves in howling wind at the top. dramatically. VI
Also note that the Compressor is enabled,
which helps fatten the sound. Stereo Delay is Stop the noise!
also used; the setting doubles the sound and Two full-blown articles later, we’ve only just
spreads it out over the stereo soundstage. If begun to explore the possibilities noise offers
you’re looking for something less dramatic, us for sound design. Everything from gun-
simply turn off the delay. shots to flutes to jet planes to A-bombs to

62 VIRTUAL INSTRUMENTS
SEQUENCING SAMPLES high frequency range of that instrument/
VI
r e v i e w
Thomas J. Bergersen (contact@thomasberg-
(CONTINUED FROM PAGE 39) instrument group, and pull down some of the [Link]) Bergersen (25) still lives in Norway.
mids in order to make it appear slightly closer He is a composer working in the media music
Naturally, an intelligent orchestration will (without saturating the mix) as well. industry. His credits include hundreds of produc-
offer the best results, but it is not “wrong” to Join us next time for some exclusive tips on tions in film, TV, trailers, radio, and video
give particular boosts to certain instruments dynamics, including how to sequence a games, as a composer, orchestrator, and
or instrument groups. When doing so it’s crescendo properly. arranger. VI
important to give a slight EQ boost in the

TRENDS David Das: to quench our thirst for power. We’ll start see-
(CONTINUED FROM PAGE 64) I think 2007 will shape up to be the Year of ing mobos with rows of processor sockets
Unlimited Processing. similar to the way we currently see PCI-X or
The strides that have been made in this RAM module sockets. Then you’ll simply add
Peter Dines: direction are just starting points compared to processing power either in series or in paral-
1. Linux will make greater inroads into what they will evolve to be. Things to consid- lel.
computer music with the release of Energy XT er: Another cool concept that’s already in use
for Linux. This will spur more developers— • Apple Logic Pro’s nodes (which offload at the enterprise system level (mainframes,
smaller, more agile ones at least—to try cross processing to remote computers on the net- servers, etc.) is hardware virtualization: taking
compiling their products. However, Linux sup- word). They only work for Logic plug-ins at a single physical computer system and,
port from sound card and MIDI controller the moment, but wait until Apple evolves through its OS, breaking it up into discretely
vendors will continue to be abysmal. them and/or other DAW companies go on operating and appearing systems. A single
board with stuff like this. I bet other compa- super duper high-speed processor could oper-
Peter Buick: nies will come along and do it better than ate discretely as a very high-end signal
1. Computers will continue the trend of Apple. Soon the only limit to your plug-in processor, with drivers to work specifically
being more powerful for less money. But I see count will be how many computers you can with its own memory and a certain piece of
64 bit as a (fairly poor) interim to the real deal. afford—and they’ll all work perfectly integrat- hardware I/O, while another “part” of that
To gain/retain a unique edge in music produc- ed. (Okay, okay, honestly, this will probably same processor could be told to operate as a
tion, the elite will invest more in DSP card not see ultimate fruition in 2007; it’ll be discrete high-end sampler, again addressing
power and tactile hardware controllers. The longer than that. But it makes for nice copy.) its own memory and I/O.
mass will remain unimpressed with the deluge • The popularity of powered plug-ins like Just like old studios with lots of hardware
of “me too” alternatives out there already, but UAD, TC, Waves’ new hardware boxes, and boxes, this is highly reliable (if one crashes,
still refuse/be reluctant to learn how to make (half-finished but always in the race) Metric the others don’t) and highly powerful (no
more out of what they already have. Halo [also SSL Duende]. Prices of these boxes sharing of resources and much lower system
Apple Garage Band and the like will are coming down too. Now anyone can add overhead for each one). The effects processor
become even more powerful (and include boutique plug-in power to a native worksta- doesn’t need all the crap that Windows runs
video) and ultimately allow end user listeners tion. And they (mostly) sound really good. in the background, and it could run on a very
to be interactive, instead of passive, as finally • Apple’s switch to Intel chips, which are a lean and mean OS that is different from the
a multitrack mp3 format (probably including more universal platform and more hackable. needs of the sampler.
video) standard is defined. People will be able We’re already seeing isolated reports of hack- A final really cool idea to chew on (and this
to do Acid-style re-mixes on their mobile ers swapping 8-core chips into a Mac. If this eludes to your everything-down-one-pipe
phone (okay, by 2008 then). goes further—like seriously mainstream and concept) is that of companies designing
By the way, my New Year’s resolutions: I people start hot-rodding real DAWs that can breakout DSP that streams extremely high
will index my sample library (likely to be bro- actually take advantage of it—there’s another bandwidth data directly off the PCI bus to
ken). I will listen to more other artists’ music speed limit gone right there. Intel chips are outside hardware. Need more? Just add
(including socio/eco sites like garageband and readily available to end users. another DSP box. Latency and conflicts won’t
myspace), especially music from other musical I remember back when the first version of be a problem anymore.
styles and cultures. I will limit myself to make Altiverb was released and it was famous for
more out of what I already have by having being the first plug-in (nothing else came Frederick Russ:
one half day programming session per week close) that would bring a then-state-of-the-art Someone will develop a virtual instrument
where I will not start a song, no matter how G4/400 to its knees with one instance... player à la Vienna Instruments that it will be
inspired I am by my tweak. I won’t eat too All this to say that the bottleneck—CPU universal and work with all libraries—Sonic
many mince pies at xmas (likely to be bro- power—for plug-in fans is being abolished, Implants, Project SAM, etc.
ken). I won’t use presets 1-10 any more. and I bet 2007 will be a banner year for this. Virtual Instruments Magazine will quadru-
2. Technology will go full circle and we’ll ple in size and scope, will be available in all
return to “people” interacting —on-line, with Jason Scott Alexander: stores, and go monthly. VI
more virtual studios, rehearsal rooms and 1. Definitely! 64-bit is going to be a big
gigs. Several real-time sharing sites have thing—I’m just not so sure if it will hit full
emerged recently, and more are about to. It’s stride any time imminent (on the PC that is).
time people stopped playing with themselves First we must get MS to roll out Vista in a big
in their bedrooms. Even the porn industry lets way. Third party softs won’t be far behind, I
punters interact with models (i.e. web cam, don’t think.
text instructions) in real time. 2. The whole concept of “chip farms,” or
IMHO I think the single-wire studio is a fal- motherboards with many (not just two)
lacy—what people want is one single box processor sockets. According to the chipmak-
that’s powerful enough to do all. We only tol- ers, they will only be able to cram so many
erate multiple boxes because that promise is “cores” onto a single processor before they
nowhere near true yet—nor is it with 64 bit. have to resort to multiple physical processors

VIRTUAL INSTRUMENTS 63
VI t r e n d s

Trends
VI writers stir the tea leaves

Thomas J. Bergersen:
Jim Aikin: I. have to agree. :)
1. 64-bit computing—who really cares? 2. More famous hardware units turned into
Not musicians, that’s for sure. Manufacturers, software plug-ins (more a wish than a predic-
you betcha. The more people they can sucker tion), even more bloated software and sample
into upgrading, the more money they make. libraries, new cross-platform plug-in standard,
The trend I’m seeing in synthesis—and I new higher resolution MIDI interface standard
expect this to gather steam in 2007—is a shift (more wishful thinking rather than a predic-
in emphasis away from the filter and toward tion, though :), tougher copy protection fol-
the oscillator. For many years, the sound qual- lowing the Vista OS.
ity of the filter has been considered a make-
or-break deal for a synth, but by now the Orren Merton:
problem of how to make good-sounding digi- 1. Slowing down of exact hardware emula-
tal filters seems to be pretty much solved. tion V.I.s. I think we’ll start to see fewer and
Meanwhile, advances in computing power fewer models of Moogs, Arps, Prophets, and
have made it practical to do waveshaping other classics, and more VIs that try to cap-
within the oscillator itself in complex and ture the zeitgeist of many different analog
great-sounding ways. Three instruments that I synths, put in a new interface without the
know of illustrate this: Cakewalk Z3ta+, NI limitations of hardware.
Massive, and u-he Zebra2. I expect others More hardware-based copy protection.
have appeared that I don’t know about, and Piracy is absolutely ravaging the smaller play-
ust for fun we sent a very simple note to still more will show up before too long. ers, and we’re going to continue seeing more

J our writers, asking them two questions

(with leading answers just to get the
crystal ball rolling):
I predict that Steinberg’s announcement of
VST 3 will get a lukewarm response, and that
support from other manufacturers for the
softsynths using dongles, integrated con-
trollers, and other related hardware.
2. Hardware synths will all offer VI front
1. What predictions do you have for the VI new spec will be spotty, at least until Q2 of ends. We’re already there with the Virus TI
world in 2007? [NB’s example: 64-bit com- 2008. and the Roland SH-201, but as time goes on,
puting—which you are free to agree with or 2. Wire? Why do you need a wire? The fully all hardware synths will need to offer a VI
not.] wireless studio is only a matter of time. front end for integration into the DAW.
2. How about for the future in general? (Thanks to Dave Smith for suggesting this Lee Sherman:
[NB’s answer: everything will be connected utopia to me.) Walk into the studio, plug in a 1. The few remaining classic instruments
by a single wire. We’re almost there with pro- power cord, and your hardware can talk to that have yet to be emulated will find their
grams that send audio and/or MIDI over anybody else’s hardware wirelessly. Audio way into software form.
Gigabit ethernet, but in a couple of years channels, MIDI, all configured automagically. 2. We’ll see entirely new virtual instruments
Fibre Channel or something else should make You know what I’d like to see, but don’t that don’t just mimic classic hardware but
this the norm.] expect to see anytime soon: a resurgence of instead go off in exciting directions made
enthusiasm among the general public for possible by today’s user interface paradigms
Chris Meyer: music made with electrons. It seems to me and processing power. NI’s Massive is a prom-
1. Sound libaries will continue to grow to that the problem of how to make amazing ising first step in this direction.
the point that once we've installed them, sounds with computers is totally solved. If I Gary Eskow:
there won't be room on our internal drives to never installed another plug-in, it would take Obviously, 64 bit computing, although the
install the next major upgrades to MacOS or me a couple of hundred years to explore the only place it seems necessary to me is with
Windows next year. musical potential of the stuff that’s on my my VSL Vienna Instrument samples... I’m per-
(Which really brings the need for licensing hard drive today. But the career opportunities fectly happy with the processing I’m getting
schemes to easily accomodate them being lag very far behind the technology. What we on my dual Opteron pre-64 bit right now.
installed on external drives that we can move need is... I don’t know, a Keith Emerson for More and more fantastic classic synth mod-
between our studio towers and live perform- the new millenium, I suppose. Somebody els.
ance laptops—not—currently the case.) who captures and ignites the public imagina-
2. Um, see parenthetical aside above. tion by jamming with a laptop. (CONTINUED ON PAGE 62)

64 VIRTUAL INSTRUMENTS