• Inside the Synclavier
  • Inside the Synclavier
  • Inside the Synclavier

Magazine Archive

Home -> Magazines -> Issues -> Articles in this issue -> View

Inside the Synclavier (Part 2)

PART 2: David Mellor explores the sampling capabilities of the world’s most expensive musical instrument - the Synclavier.

David Mellor explores the sampling capabilities of the world's most expensive musical instrument - the Synclavier.

Sampling has surely now come out of its 'novelty' phase. Those stuttering 'N-N-N-N-Nineteen' effects and quotes pinched from old Humphrey Bogart films are well and truly past history - I hope! But does sampling have such a thing as a mature phase, where it is used constructively and creatively, and not just as mere titillation to the aural sense? Thanks to advancing technology it certainly does.

When I first encountered sampling back in 1983 (was it really that long ago?), I was not very impressed. The maximum sampling time of the unit I had available seemed just too short. You couldn't even fit a decent drum sound, with natural decay, into the memory without premature cut-off, and playing it back at different pitches from a keyboard - well, that was just a dream. But music itself adapted to the limitations of early sampling devices. Creative people made good use of what early samplers could do, instead of complaining about what they could not.

Today, using a sample that sounds like a sample has become an outdated pastime. The modern sampler has to be a real musical instrument in itself, not just a poor imitator of other instruments. Note that it can still mimic a conventional acoustic, or electronic, instrument. But in 1989 it needs to do it very well and add to the range and freedom of expression available.

Sampling has been available on the Synclavier for around six years now, at first monophonic but now very comprehensively polyphonic. Before that it was billed as a 'digital synthesizer' - and, of course, that was quite something in the pre-DX7 era. Now, considered as a whole system, the Synclavier is probably the world's most advanced sampler. And, what's more, it is reasonably easy to use. There are lessons for other sampler - and sequencer - designers here in abundance, the principal one being that just because a machine has a lot of powerful functions, it should not mean that it is difficult to operate. In fact, no effort should be spared to ensure that the opposite is the case.

Getting down to the basic 'power' spec, the Synclavier 3200 - the baby of the range - is capable of 16-bit sampling at rates up to 100kHz; over twice the CD sampling rate. In a fully expanded 3200, there can be 32 Megabytes of memory and up to 32-voice polyphony. The Synclavier 9600 can have up to 96 Megabytes of memory (16 minutes of audio, sampled at 50kHz) and 96 polyphonic voices. That's more voices than the choir of St. Paul's Cathedral!

But sampling is not all about big numbers (and let us not forget that the price of all this raw power is a big number, too). Sampling is about manipulation and control. And especially in the Synclavier's case, it is also about integration with other system functions such as the onboard sequencer and sound librarian. But we'll cover that next month. First, let's take a detailed look at the sampling process...

Figure 1. Example screen from the Sample-to-Memory Sound Editor module.


Of course, the first step is to get well acquainted with the manual. And when you buy a Synclavier, you don't just get one manual, you get a whole shelf full! There is an inch-thick volume devoted to sampling alone, and as you know from last month's installment, the Synclavier does a whole lot more besides.

Figure 1 shows just one of the screens available in the Sample-to-Memory Sound Editor module. The display on the Synclavier's Macintosh II colour monitor, I should point out, is infinitely more clear than this laser-printed screen dump.

Samples can be recorded either in mono or stereo, to RAM memory or direct to hard disk (an optional software extra). The sampling rate is variable between 1 kHz and 100kHz in 100Hz steps. The default setting of 50kHz is quite adequate for most purposes. The 'industry standard' sampling rates of 44.1 and 48kHz can, of course, be programmed as easily as any others and may be useful with the optional digital output module.

As soon as a sample is recorded, in much the same way as you would do with any sampler, a waveform display is drawn on the screen ready for modification, along with all the relevant information about the sample - which is known in Synclavier language as a Sound File:

Current Filename is the name you give the sample for storage on disk as a Sound File. A caption (which in this example is 'Looped French horn - loud') of up to 128 characters can be added.

Rate is simply the rate at which the sample was recorded.

Crossfade is the crossfade setting for use with Cut, Paste, Delete, Fill and Loop commands.

Length is the length of the sample in seconds.

Cursor gives the cursor position in seconds.

Left and Right give the amplitude of the sample at the cursor position.

To take a good look at the sample, it is necessary to select the Display menu by clicking on 'A) Display' in Figure 1. A selection of the options available on this menu include:

Centre redraws the sound file so that the current cursor position is shifted to the centre of the sound file window.

Search places the cursor at the next occurrence of a specified amplitude. This is done by entering the amplitude in volts.

Zoom can expand or compress the vertical and horizontal scales.

Mark is where Start and End marks can be entered to define the wanted portion of a newly recorded sample. Labels can also be given to different points in a sound file.

The Modify menu, as shown in Figure 1, is where all the interesting things happen to alter the sample. Most of the commands are self-explanatory. REVERSE, for instance, reverses a specified portion of a sound file. CUT removes a section from a sound file and places it in the edit buffer. PASTE takes whatever is in the edit buffer and inserts it into the sound file at a specified location.

CUT, PASTE, DELETE and FILL (over-writing part of the sound file with the contents of another sound file) all work in conjunction with the CROSSFADE command to give either a hard join or a smooth transition, up to half a second, from one part of a sound file to another.

LOOP generates crossfade loops, using the crossfade setting mentioned above, which are given new filenames and stored separately. There is another type of loop, called a Keyboard Loop, which I shall mention shortly.

The Modify II menu was, at the time of writing, empty on the system I was working on and awaiting further software additions. Expected soon is Digital Signal Processing, which will enable time compression and expansion, allowing the length of a sample to be altered to between 50% and 200% of its original value without changing pitch or degrading the sound quality. There will be different algorithms optimised for speech and for music.

Compared with the samplers I have so far encountered, the Synclavier offers a mix of many of the right functions. But most important is the visual display. A good display makes sample editing a pleasure. Roland's S550 and Hybrid Arts' ADAP both come close, but when you see it in real life the Synclavier is tops. Perhaps some computer-based sample editors give it a run for its money, but with the Synclavier, the sample is in the system - you don't waste time shuttling it back and forth along pieces of wire between sampler and computer and back again.


A collection of samples, recorded, modified and allocated to notes on the keyboard, forms a Timbre - the equivalent of a Program or Preset, depending on which sampler dialect you speak.

A Timbre consists of up to four Partial Timbres. In the 3200 system, the Partials will each consist of a number of samples. In the larger 9600 system there may be synthesized and resynthesized Partials, too. For instance, Partial 1 could be a multisampled grand piano. Partial 2 could be a woodwind sound. Partial 3 strings, perhaps, and Partial 4 synthesized brass. You now have a full orchestra at your fingertips.

Figure 2. Timbre Parameter panel.

If the Partial is created from samples, then a Keyboard Patch must first be made to allocate different samples to appropriate areas of the keyboard, as you would do with any sampler. The screen display makes this a simple matter. If the samples are not already looped, then a Keyboard Loop, without crossfade, may be applied which does not alter the sample as stored. There is an automatic looping facility for Keyboard Loops.

When the keyboard has been allocated its full quota of samples, thus creating the Partial, the characteristics of the Partial may be modified in a variety of ways. It is convenient to do this from the Synclavier's button panel - either the real one on the 9600 or the software emulation on the 3200. Figures 2, 3 and 4 show the relevant sections of the panel: Timbre Parameters, Keyboard Control, Timbre Control, and Real Time Effects.

Figure 3. Timbre/Sequence Storage and Keyboard Control panel.

The Volume Envelope needs little explanation because it is an envelope generator similar to the type we are all familiar with. As you can see, its parameters include Delay (the time taken between a key press and the note sounding), Attack, Initial Decay, Final Decay, Peak (the maximum volume level) and Sustain (the level of the sustained part of the note). There is nothing exceptional here, but it's a useful collection of parameters - and many of us are aware that having too many envelope parameters can be more of a curse than a blessing. The Decay Adjust button - lower right of the Keyboard Control section - allows lower pitched notes to have longer Decay times than high notes, as often happens in real acoustic instruments.

The Timbre Control section (Figure 4) includes very versatile control over Vibrato, Stereo (9600 system only) and Portamento. Taking Vibrato as an example of the amount of detail that is presented, the Wave button offers a choice of waveshapes which govern the rise and fall in pitch of the note: Sine, Triangle, Ramp, Inverted Ramp, Square and Random. There are another six possibilities for use with the digital synthesizer section of the 9600.

Of course, the rate and depth of Vibrato can be controlled, but also there is control over Attack and Decay, Bias and Quantisation. There is similar wide ranging control over Tremolo and Portamento.

Figure 4. Timbre Control and Real Time Effects panel.

The section which does most to turn the Synclavier into a musician's instrument is the Real Time Control section (Figure 4), where the characteristics of a Partial or Whole Timbre can be modified on a note-by-note basis. The possible methods of control are: Velocity, Pressure, Pedal 1 and 2, Modulation Wheel, Ribbon and Breath controllers (9600 only), Keyboard Control Voltage (Timbre affected according to whereabouts on the keyboard you are playing).

Those are the controllers, and what you can control with them amounts to quite a lot. In fact, all of the parameters whose buttons have a small white dot next to them are available for Real Time Control (some are only available on the 9600 system):
- Volume Envelope
- Harmonic Envelope (9600)
- Partial Tuning
- Partial Volume
- Track Volume (refers to the integral sequencer)
- Frequency Modulation Ratio (9600)
- Vibrato Rate, Depth and Modulation Depth (Mod Depth 9600)
- Stereo Rate and Pan (9600)
- Portamento Rate
- Repeat/Arpeggiate Rate
- Partial Chorus
- Dynamic Envelope
- Vibrato Attack/Decay

The Dynamic Envelope is an interesting feature, which I touched upon very briefly last month. This allows different Partials to sound at different keyboard velocities. As most of us are aware, it is not only the loudness of an instrument that changes as you play harder. The character of the sound changes, too. Using the Dynamic Envelope feature you could multisample a piano, say, at four levels of loudness and use each of these as a separate Partial. With the appropriate Dynamic Envelope settings, playing the Synclavier keyboard progressively harder will bring in the different sets of samples. I have played a Synclavier with an orchestral Timbre arranged in this way, and I can confirm that it is an absolute delight. You get a feeling of command rather like conducting a real orchestra.

If you take a close look at the three panels shown (of six) - Figures 2, 3, and 4 - you will notice that there are quite a lot of buttons that I haven't mentioned. Never fear, they will all be covered in due course, either in detail or as part of a complete guide that I hope to include later in this series.

In next month's installment, I intend giving a flavour of what it is like to use the Synclavier as an integrated system - and I can say now that until you have tried the Synclavier, you do not know the true meaning of the word 'workstation'.


Harman UK, (Contact Details).


The 9600 version of the Synclavier stands at the top of the range, almost unattainable to ordinary mortals. Its hardware includes the standard Synclavier 76-note velocity and pressure sensitive keyboard and button panel, a Macintosh II-based workstation with 19" colour monitor and track ball, and also one or two 19" rack units, depending on the system configuration.

As far as voices and memory are concerned, the 9600 starts where the 3200 leaves off and is an expandable open-architecture system. There are 32 polyphonic stereo sampling voices (although the 3200 can handle stereo samples, its voices are mono), expandable to 96, and also up to 32 digitally synthesized FM voices. 32 Megabytes of memory come as standard, with 96 Megabytes as the upper limit. Hard disk storage may be expanded up to 960 Megabytes and, like the 3200, optical disk may provide 2 Gigabytes of WORM (Write Once Read Many) storage.

In a nutshell, the basic difference between the two systems - apart from mere size - is the fact that the 3200 is limited in its expansion capability. If you feel that the 3200 is what you want, and you will want to go no further than its limit, then that's fine - go ahead and buy it. But if you want to keep your options open, then the more expensive 9600 would have to be the way to go.

It would be a difficult choice between the two systems. I suppose we ought to be thankful that, at Synclavier levels of finance, it is a choice that few of us will ever have to make!

Technical Data


Number: 2,16-bit linear
Impedance: 200 kOhm balanced, 100 kOhm unbalanced, 20Hz to 20kHz
Gain: Programmable, -3 to +28dB, in 1dB steps, +/-0.05dB accuracy at 1 kHz
Sampling Rate: Programmable, 1 to 100kHz, in 100Hz steps
Total Harmonic Distortion: Less than 0.02% at +4dBm at 44.1,48 and 50kHz
Intermodulation Distortion: Less than 0.02% at 44.1,48 and 50kHz
Signal-to-Noise Ratio: 94dB minimum, gain=0dB, 22kHz bandwidth
Crosstalk: Less than -90dB, channel to channel


Sampling Voices: 32 polyphonic stereo, expandable to 96 in increments of 8; 16-bit MDAC, two per voice
Synthesis Option: Maximum 32 additive/FM voices, increments of 8


Number: Electronically balanced stereo output, 8 assignable individual outputs expandable to 16
Impedance: 100 Ohm balanced, 50 Ohm unbalanced
Gain: -1dB for 1 voice peak amplitude, +3dB per every doubling of voices


RAM: 32 Megabytes, expandable to 96Mb in increments of 4Mb
Hard Disk: One 320Mb drive, maximum of 3
Super Floppy: Built-in 5¼" drive
Optical Disk: 2 Gigabytes WORM (optional)
Kennedy Tape Drive: 15Mb per cartridge (optional)


  • Macintosh II Graphics Workstation package, includes 19" colour monitor, 2Mb internal RAM, 20Mb hard disk, track ball, 256-colour graphics card, workstation furniture.
  • 76-note velocity/pressure sensitive keyboard and button control panel
  • Digital guitar interface (optional)
  • Remote controller/editor locator (optional)
  • 2x8 MIDI module
  • External clock in/out
  • SMPTE reader/generator
  • VITC reader (optional)


  • 200-track sequencer
  • Sound librarian
  • Sound editor
  • MIDI sound/sequence directories
  • Audio event editor
  • Analysis and resynthesis
  • CMX edit list conversion (optional) *
  • Music engraving (optional)
  • Synclavier Timbre library
  • Sound Ideas sound effects library (optional)
  • Denny Jeager Master Violin library (optional)
* CMX is a video editing system. 'Edit list'refers to a list of editing decisions made in a low-cost 'off-line' video editing suite which can be taken to an on-line broadcast quality system and carried out automatically.


DSP/Time compression: Motorola DSP 56000 coprocessor
Direct-to-Disk Multitrack: 4, 8, 12 or 16 tracks


Read the next part in this series:
Inside the Synclavier (Part 3)

Also featuring gear in this article

Previous Article in this issue

Korg M1R

Next article in this issue

A Programmer's Guide to the Roland D110

Sound On Sound - Copyright: SOS Publications Ltd.
The contents of this magazine are re-published here with the kind permission of SOS Publications Ltd.


Sound On Sound - Jul 1989

Donated & scanned by: Mike Gorman


Should be left alone:

You can send us a note about this article, or let us know of a problem - select the type from the menu above.

(Please include your email address if you want to be contacted regarding your note.)


Inside The Synclavier

Part 1 | Part 2 | Part 3 | Part 4

Gear in this article:

Synthesizer > New England Digital > Synclavier

Gear Tags:

Digital Synth

Feature by David Mellor

Previous article in this issue:

> Korg M1R

Next article in this issue:

> A Programmer's Guide to the ...

Help Support The Things You Love

mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.

If you value this resource, you can support this project - it really helps!

If you're enjoying the site, please consider supporting me to help build this archive...

...with a one time Donation, or a recurring Donation of just £2 a month. It really helps - thank you!

Small Print

Terms of usePrivacy