The Sogitex 4X is mentioned in the Q&A below. You can find more info on the Sogitex here. The following is some info from that site: "The Sogitec 4X digital synthesizer at IRCAM during the 1980's, was the direct result of Luciano Berio's demand that live electronic sound ought to consist of at least a 1000 sinewave oscillators (1), in order to be interesting for the composer and the audience alike. It was the last huge hardware, that IRCAM instigated before it turned to the ISPW and subsequently to software solutions in the shape of jMax and Max/MSP. The 4X's eight internal custom-built processors were capable of altogether 200 MIPS (2), which being the equivalent of 1000 sinewaves, 500 filters or 450 second order filters (3). Each processor contained a data-memory, an address-memory, a microprogram-memory and a function-memory. For calculations it used 24 bit fixed point units consisting of a multiplier, a arithmetic and logic unit. It also had 256 internal (programmable) clocks and a large dual buffer for recording and playing. The algorithms were cross developed on DEC en Sun mainframes in Fortran and C. (4&5)."
Q&A on the Acxell II via an anonymous reader:
Q: If i understand well,its posible to acxelize a complex sound source from a pre recorded wav with the soundbuilder then use this resynthezized model in acxelink mode again to be able to control that sound it real time via an external audio input (1 or more?)( each one )
A: You are Right (can have up to 16 IN, 16 OUT adressable separately)
Q: modulary addresable to an 'element' or (preconfigurable) group of 'elements' (read group of cells,group of envelopes etc..) via a (or more) cell(s) in 'prog Rec/sample mode with the possibility to configure those/some as 'sample/hold(loop) and/or in real-time resynthesis mode (Because u mention also real time resynthesis.) ?
A: Yes
Q: If, are there preprogrammed preset 'modules (modes)' in wich matter the signal gonna be translated to modulate (envelope followers,pitch trackers,etc..)?
A: In the input elements, we implemented resources for level detection and pitch detections.
There are also math elements and filtering elements that can supply calculated results like: RMS, Band envelope followers, etc..
Q: And when i read future development at bottom of pdf that mean that real time audio signal at the input can pass/thru dry,morphing to real time resynthesized result of that sound also morphable to the loaded acxelized preset where it all started at beginning of my question above?
A: Morphing acts on different manner and levels, using Dual envelopes in combinations with cells. Or within cell by morphing the waveshape itself.
Q: Isn' t that not something the 4X [Sogitex 4X] could do at IRCAM? (have u talked to those guys,for help,support...?)
A: Not for support. We met them 1 or 2 times 20 years ago. At that time we had our Acxel 1 working yet. It is independent development.Our harware and software approaches, as far as I know is completely different.
Also as far as I know our morphing work really differently compared to the 4X, it acts at many levels internally and from outside world, at element level or global level.
Q: anyway if ur machine can do that,without blips glitches or other unpleased artifact,i want one for sure :) !
BTW,256 is the base pack of cells and and one card is 16 v able,but is it possible to adress a choosen amount of cells u want to use ,or is it limited to 32?
A: Yes, from single cell up to the maximum available, this is one of the major differences with the original Acxel
Q: If,If,if, makes sure to add the possibility to ad enough physical audio in/outputs and have the possibility (with buffer/jumpers inside) to be able to have very high level (voltage) outputs signal,so u can also directly use the machines outputs to drive elements on external analo modular synthesizer gear (for example)
A: In fact we actually look for many applications like Analog circuits emulations for example."
Update: I updated the post with the image of the Sogitec 4X. It previously was "a DEC (Digital Equipment Corporation) sales brochure for the DECSYSTEM-20 (AKA PDP-10) computer" according to Stephen Jones in the comments. BTW, do check out the comments of this post for more info.
Be sure to clic on the Acxel link below for previous posts.
The system shown is from a DEC (Digital Equipment Corporation) sales brochure for the DECSYSTEM-20 (AKA PDP-10) computer. Its a 36-BIT system with an 18 BIT buss. This picture shows two TU-44 9 track tape drives, two RP06 disk drives (max pack size is about 140MB) and a DECwriter II printing terminal (usually for the console) and a DECscope video terminal (does VT52/VT100).
ReplyDeleteThe system runs a few operating systems called 'monitors', typically TOPS-10 or TOPS-20 (very different despite the names).
Another important note is that a customized successor (the Foonly F1) was used to create some of the 3D graphics and animation in the movie TRON.
http://www.twenex.org is a public access TOPS-20 thats been running for many years.
I remember reading about these gigantic "academic research" digital synths in MIT's Computer Music Journal:
ReplyDeletehttp://204.151.38.11/cmj/
I believe that Stanford's CCRMA had a big digital synth called the Systems Concept Digital Synthesizer; I think it was nicknamed the Samson Box. David Jaffe did a great composition on it called Silicon Valley Breakdown which is on his CD called XXIst Century Mandolin.
The Foonly was used by a researcher for (non-real-time) score extraction from orchestral audio. Celemony, eat your heart out.
Thanks guys. I updated the post with the 4X.
ReplyDelete