Fairlight-CMI

HI there : 

I glad to hear youre getting closer. Unfortunaly, you might be at the point where you might 
have to recognise the limitations of how close you can ultimately get. This is a post I made 
to the KVRAUDIO list describing why the fairlight is unique, and why you cant emulate in 
software with the current state of play in software design : it refers mostly to the Series III, 
but im sure some of the series I and II owners will have their own thoughts :

--------

The series III specification wise still holds its own against everything else Ive got. Sampling 
up to 100 Khz theoretical maximum at 16 bits still sounds very nice. 

But is it still unique ? 

The series III comes with 3 sequencers : Page R, which is basically a tracker, CAPS which is 
like an old version of cubase, and MCL which is a text based programatic sequencer. 

What makes it special is that for each step on the grid, you can have a different sound 
loaded, like an old amiga tracker, so for doing those old tracker tricks, its great, especially 
page-R. CAPS is clearly outdated, and not something thats very usefull anymore. MCL is 
great, but ive never got the hang of it, and might have been great for achedemic electronic 
music, if more fairlights had found their way into that environment. 

What gives the fairlights a unique sound is due to its brute force engineering architecture : 

You have the digital side : with a digital voice card which controls the data buffers etc for 
each 16 voices, which routes to an analogue voice card per voice, which has a sample 
playback occilator, and the rest of each voice is an analogue subtractive structure. 

So, you get all the nice sound of analogue filters etc. And of course, misalignment between 
the settings on the cards ensures each voice sounds a bit different. 

Also, each playback occilator transposes not by multiplexing, but by altering the clock 
playback speed. This means a note slowed down because it is played lower sounds very 
different than it does on a multiplexed system such as an akai S6000. 

Also if you play a note quiter, the analogue VCA levels are altered, not the sample 
playback depth, so even a quiet not is 16 bit, giving a lot more dynamic range. 

There is no multiplexed output either : each voice is exposed as an XLR output, and you 
have to mix them on an 16 channel analogue desk, so you get into that "Analogue 
summing" vs digital multiplexing argument. Also it ensures there is absolutely no 
dithereing going on anywhere in the system. 

The upshot of all this, is that sounds a million miles away from a s6000, or softsynth. 

In terms of interface : its very simple, and very focused. Its easy to do what you need in 
terms of sample editing, and is great for zooming in and doing "micro" editing and mixing 
different samples and crossfading to create new sounds : mixing a tr808 bassdrum with a 
bass guitar sample to produce a playable hybrid etc. The sort of thing people are too lazy 
to do these days. 

Lastly there is the additive and resynthesis functionality. Load in a sample, click analyse, 
and go in and manually redraw some of the FFT frames, interpolate between them, etc etc. 
Additive is the same thing, but starting with a blank waveform. I dont know of any current 
program that does that, maybe cameleon ? I know it does additive, but i dont know about 
resynthesis. And then you have all the standard analogue subtractive on the outputs. 

It has 4 midi ins and outs, which are rock solid in timing, and a funky graphics pad for 
drawing stuff. It has a keyboard with keys dedicated to rests, sharps and steps to make 
programming much quicker than using generic keyboard shortcuts. 

It comes with a beutifull fully weighted keyboard, with a built in remote keypad, and you 
can get a extended MFX controller with even has a secondary display, and some of the 
alphanumeric keys are velocity sensative so you can triggers samples without having to 
reach over to the main keyboard. You can plug in a mouse to use instead of the graphics 
tablet if you prefer. 

There is a retrofit board to give you colour VGA output, and you can also get a dedicated 
DSP timestretch card. 

So its a pretty funky and unique beast ! And the additive and re-synthesis subsystems 
where never really tapped at the time, as sampling was the thing, so there's still plenty of 
room for original sounds beyond SARR1.VC 

Cheers, 

A 

/// edited to add : ive heard the IIX library, and as may possibly be obviosus from the 
above, playing a fairlight sample back on another sampler sounds nothing at all like a real 
fairlight, for all the reasons above. If yove heard the IIX sample library, and think it sounds 
nothing special : I agree with you. If youve heard a IIX or series III playing back the same 
samples, its a very different experiance.