AUTO-ADMIN: Non-MP3, WAV, OGG, SoundCloud, YouTube, Vimeo, Twitter and Facebook links in this post have been protected automatically. Once the member reaches 5 posts the links will function as normal.
Hi,Thanks for the kind words.
Given the present situation I'm not really going to comment much more on the "Pro Rec" version of the disk, other than I've ordered a couple of copies and they're winging their way for me to take a closer look at.However I can point out that WRB = William Robert Brady (My initials, I go by my second name, and that 181172 = My date of birth (UK format) and 01 because it was the first disk I'd produced.
Yes, I know that makes me all of about 11 when the Series II came out, but I've had a long facination with the CMI ever since I got to breifly play with one the very first time that year. As a kid I used to tag along with my (Much) older brother when he went to visit his friends' houses in and around Bath and Frome in the UK. No, I'm not naming names (except to say it wasn't Mr Gabriel's), but my love affair with music technology started right there. As soon as I could I got my own about 10 years later, even though it was antique even then.
There was originally a small booklet that came with the CD, most of the contents of which are up on the Pro-Rec website.
In the listings the first entry on each line is the name of the sample of the 8" Fairlight floppy, if a pitch is given then that's the note on the keyboard that was pressed to generate the sound. If not pitch is given the sample was played back at its recorded pitch. Then there is a short comment about the sound itself.
There were 2 main versions of the library, "Blue" and "Pink" identified by the color of the labels on each disk. The disk represents the contents of the later "Pink" library, which was a superset of the "Blue" one. This was the factory collection of sounds to get you started. It really was the first sample library ever, and so many of the concepts and tools that we use today (and take for granted) have their roots in this.
What's really amusing is if you listen closely to music from the period is just how many of the tracks produced using a Fairlight are simply playing back presets, presets which are on the Sample CD. Ooh, lazy cheaters the lot!
There's an interesting if I feel somewhat overly gushy yet incomplete essay about the history of one sample from the library that can be found here.
http://homepage.mac.com/WebObjects/File ... 0PM%20C00A (http://homepage.mac.com/WebObjects/FileSharing.woa/wa/downloadFile?user=rfink1913&path=.Public/PDFs%20for%20submission%20to%20PM/The%20Story%20of%20ORCH5%20%28for%20PM%20C00A)
Only a handful of people ever truly got to grips with the Series II and as such were hired as "Fairlight Programmers" acting as session musicians on countless records. The "Programmer" title is not undeserved, as outside of the drum machine like Page R there is a whole PROGRAMMING LANGUAGE (MCL) that can be used to control every nuance of a sequence.
Let's put this in perspective: Multi-timbral, polyphonic sequencing with (some) automation in the late 70s and early 80s, long before anyone had dreamed of hooking up a Personal Computer (The concept of which bearly existed) to control instruments. With a CV analog board or the IIx MIDI expansion the CMI could even conrol external Synths in the same manner we do now. Utterly amazing.
JJ's "Art of Sampling" contains many other sounds that are strongly associated with the "Fairlight sound", and between the two disks that's pretty much everything that there is to grab outside of the personal libraries of the likes of JMJarre, Peter Gabriel etc. Which is a shame as I'll bet a nickle that those sounds are all but lost to time now.
Given the limitations of the voice architecture of the CMI when the Emu EMUlator was released at a fraction of the cost many folks snapped them up, including those that owned Fairlights. Why? It wasn't really multi-timbral, had no sequencing or the cool lightpen, but it had by comparison massive memory for not much (Relatively!) money. As a result, several of the sounds that people think of as Fairlight actually are EMU!
Also, given how stupidly expensive the CMI was, and is still if you can find someone who would be willing to part with theirs, many early tracks that contained samples were actually a delay unit playing back in one shot mode. Paul Hardcastle's Nineteeen is the biggest example of this that springs to mind.
Anyway, back to the sounds and why I feel that a straight emulation of a CMI II(x) would really be a waste of time. If you listen to the library it's really obvious that by today's standards the quality is pretty shocking and the sounds themselves are mostly not that interesting other than for their, admittedly huge, historical interest. What made the Series II what it is and how it has it the mythical status it does is because it was very very rare (I heard only 300 were ever made), it was the first and radically different, had a ton of limitations that lead to creative workarounds and that it was exorbiantly expensive.
Yes, they're built like tanks and used the groundbreaking, cutting edge hardware of the time, but in reality the limitations of the system are what drove creativity with it.
If you know what you're listening for you can hear all kinds of twisted uses of the samples in the library in music of the time, but once you poke around in the actual library it's apparent just how much "Studio Fairy Dust" has been applied to the samples to get from the raw output from the unit to what you hear on the record.
The 8 voice short sample limitations also make you realize just how talented Trevor Horn, Anne Dudley and JJ Jeczalik actually are. The Art of Noise sessions featured multitracked CMI and a lot of planning.
Synching sequences to tape involves recording a sine tone to a track and playing it back to the CMI. It would start playback at a speed defined by the frequency of the waveform. Ingenious if you think about it, as the CMI will track the wow and flutter in the tape by varying its playback speed in response to the changes in pitch of the signal.
There was no concept of Song Position or timecode, if you wanted to replay a section you had to start from the very beginning! Again, knowing the flaws of the CMI makes the results that it produced only more spectactular when in the hands of someone who really knows it inside out.
The voice structure and control of the voice cards also makes up a significant part of "The Sound", but not all of it. A factory fresh Series II has 8 voice cards, each with just 16K of RAM, bringing the total RAM available for holding samples to 128K (8 x 16K). However, no single sample could be greater than 16K. Infact EVERY sample is EXACTLY 16K in size. If you want a longer sample time you must lower the samplerate, better quality = shorter sampletime.
If memory serves, at the 24Khz Samplerate max sampletime was just 2/3 of a second.
RAMsize/Sampelrate*bitdepth = time
16000 / (24000 *
= 0.66 Sec. There's a lot of good technical information to be found here http://members.tripod.com/kmi9000/kmi_cmi.htm#kmi_cmi (http://members.tripod.com/kmi9000/kmi_cmi.htm#kmi_cmi) so I'm not going to spend too much time going over information which most people already are aware of or have access to.
The reason for the low pitched samples on the CD. It was to attempt to capture the strange way that the CMI handled pitch. Most modern samplers interpolate when pitching down, and skip samples when pitching up. The CMI doesn't do this, it simply adjusts the frequency that it plays back the samples. For example, if you have a sampled drum hit that was recorded at 24KHz, play that sound back 2 octaves lower and you hear ALL the sample points, however they're being played back at 6KHz! Nyquist states that for a perfect system that the upper frequency limit would be 3KHz in that case, and with all kinds of alaising for anything above that. That grungy steppy sound while widely cursed at the time is now looked back at fondly.
It also produced phase coherent transposition, which is again something that most "Modern" samplers don't do. Just bear in mind though, with 8 bit recording and low sample rates the fidelity of the Series II is awful despite this nice feature.
For management each sample was divided into 128 equally sized "Segments" over its length. The loop point could ONLY be set on one of these boundaries. How long each of the slices is depends on the sampling frequency used at the time of creation. The net result is that most loops are really, REALLY, /R/E/A/L/L/Y bad. Crossfade looping was still years down the road, so delay and reverb are your friends.
On the Library CD if there was a looped sample the note was left to play long enough for the loop to cycle at least once. If there is no loop in a sample on the CD, there was no loop in the original. Hope that clears this up for people!
Moving on, voices couldn't be assigned dynamically, you had to declare them. Essentially what you're doing is loading the sound you want to play into each of the voice cards. So if you want a 4 voice Poly Syn Choir, 1 voice Kick, 1 voice Snare, 2 Voice Guitar the following would happen:-
Voice Cards 1 - 4 would each contain its own copy of the choir sample
Voice Card 5 would contain the Kick
Voice Card 6 would contain the Snare
Voice Cards 7 & 8 would each contain its own copy of the Guitar
On the back of the CMI there is a balanced, Mono output from each of the voice cards, and Mono "Mix Out" which is the result of analog summation of ALL the individual outs.
The practical upshot of which is that unless you want all your sounds to come out as mono with no individual level control of each part you need bring the output of each Voice Card up onto one of 8 faders on your desk.
Now things get interesting quickly.
Just like an Analog Polysynth every time a note is pressed the CMI cycles through the assigned voice cards (Oscillators on the Analog Synth). So the 4 note poly Syn choir I described above would be have like this:-
A stacato, 1/4 note pattern would cycle around the Voice Cards in order: 1, 2, 3, 4, 1, 2, 3, 4 and so on.
Hit a 2 note chord and the first time it would play on Cards 1 & 2, the second time on Cards 3 & 4.
Hit one note and then a 2 note chord and you'll get Voice 1 then Voices 2 & 3 playing the chord.
Now, hit another 2 note chord and it will be played by Voices 4 and 1.
If you have the faders on your desk at unity, panned center then you have a nice, simple, 4 voice choir.
BORING!
Here's where you get creative
pan each of the 4 outs, effect them sepertately, EQ them seperately and with a bit of luck, patience or skill you now have a REALLY INTERESTING synth choir sound. This is where a little MIDI plugin could mimick this behaviour. Now that WOULD be both worthwhile and cool, not to mention much simpler and far more practical than a whole CMI emulation.
The MIDI filter would take a stream of notes in and then output them cyclically through a predefined number of MIDI outputs. On the receiving end would be n x instances of a Mono sampler, each with its own output enabling the processing of each "Voice Card" seperately...
Alternately "Insert VSTi Creator Name here" could add this "Voice to individual output" feature to any existing sampler.
So as I said before, the limiations of the Series II make for some really amazing creative posibilities, many of which are not tied to its sonic signature, but do constitute a major component of its overall character.
To wrap up, for a long time I dreamed of a Series II VSTi, but if it was to be totally accurate it must have all the limitations of the original. Those limitations severely hamper the practicality of such a thing, but if they were to be "Improved upon" (More sample memory, more parts, better control of filters and envelopes, loop points, etc) then the soul of the emulation would be lost.
My vote is to remember and cherish the CMI I, II, IIX for what they actually are, and not what people fantasize about. Thinking about the voice allocation is something I haven't done for a long time, and how the initial workaround for a limitation opened many creative avenues. That aspect of the CMI would certainly be really cool to emulate.
Sonically my library disk and the JJ disk give you pretty much everything from the instrument, however it's in the working methods and creative processing that a star is born. Setting up similar processing chains with modern samples and control methods would rock.
There is a whole otherside to the CMI which I haven't touched on at all, and that's Mode 1 additive synthesis and the ability to resynthesize partials from samples. I've not really seen anything other than "Experimental" attempts to do this, and Fairlight's early approach to the ways of handling such a complex thing as far as I can tell has not been bettered to date.
I have a sneaky feeling that Celemony are using similar techniques for their DNA tool, however that's only scratching the surface of what can be done with this technology.
Anyone wanting to try their hand at such a thing is welcome to PM me for more background / detail, but I think this post pretty much covers everything that's needed to know about some little understood, but significant parts of what makes the Fairlight different to modern samplers and how they're USED as opposed to dry technical specifications.
Finally, here's my real wish for the future instead of dwelling in the past...
Fairlight turned the entire world upside down with the CMI, but over the years lost ground to the onslaught of much cheaper and freely available hardware samplers from the likes of Akai, EMU and others. They moved to non-linear recording, and then to integrated digital recorders & consoles for high end Post-Production. Again though they lost out to cheaper and more agile competitors (Digidesign I shake my fist at you!).
Now though they're back with the FPGA powered Crystal Core CC-1. With the Xynergi controller, and an audio interface they're competitively priced to a bigger Digidesign setup but with much much more power and stability. I really hope Fairlight can come back up off the ropes and kick some serious ass again in the music arena.
A CC-1 powered CMI Series IV would rip apart everyone's preconceptions of what can be done in exactly the same way the original CMI did 30 odd years ago. The power of that card is insane, and all the guys producing closed, DSP based boatanchors had better start playing catchup if they want to keep their jobs.
A couple of strategic alliances between Fairlight and some plug-in manufacturers would make life really tough for the competion.
http://www.fairlight.com.au/ (http://www.fairlight.com.au/)
No, I don't have any affilliation with Fairlight other than having used their gear for a long time and had a lot of fun in the process.
Roll on the next 30 years.
Cheers,
Rob
[/url]
