From tb303+@cmu.edu Tue Sep 27 17:21:11 1994 To: Electronica <+dist+/afs/andrew.cmu.edu/usr1/tb37/list/list.dl@andrew.cmu.edu> Subject: Re: FM/additive/Vector synthesis Content-Length: 2453 X-Lines: 55 Status: RO Damien Miller writes: > 1) Can anyone reccomend a good FM synth? Depending on how much money you have to spend there are about three choices. If you're completely skint then I think your best bet would be a TX802? I'm not sure about the number but it's basically a one unit rack mount with a six operator, eight waveform FM synth in it. The other two choices would be an SY-99 or a TG-77. The SY-99 was the Yamaha flagship synth. It does 6 operator, 16 waveform FM with RCM and filters, and you can download samples into it. The TG-77 is a three unit rack mount with basically the same synth engine the differences being less flexible effect routing and no downloading of samples but an additional 8 individual outputs. I've got a TG-77. > 2) Can anyone reccomend a good Additive synth? (don't kawai make one?) The Kawai additive was the K5 (no longer in production) they also did a module version called the K5m. The other alternative is an old Kurzweil machine called the K150. I know even less about these but they sound more powerful than the K5 (256 partials with a 256 stage envelope for each). Apparently the unit is only editable via sysex or (for the K150fs) via an RS232 interface hooked up to an AppleII. I'm currently seeking one of these but they seem quite rare. > 3) What is vector synthesis? It depends who you ask. > How does it work? I think the VS version was basically about crossfading between different waveforms. Each voice on the VS had four sample sources which you could crossfade between in realtime using a joystick. > What keyboards (apart from the SC prophet VS) implement it? The Korg Wavestations have it as an integral part of their archetecture. There are also some cheapy Yamaha synths that do it. (SY-33?) I can't remember the exact model but it's easily recognisable from the joystick on the from. You could use the joystick to record complex two dimensional envelopes which control the crossfading between the different waveforms. Besides these most half decent workstation type synths are capable of it with a little programming effort. For instance on the K2k you'd use a voice with 4 layers, set the layers up in pairs with the crossfade parameters set up appropriately and use a different controller for each pair. Not as good as having it built into the archetecture but ok if you put the effort in. Ab. -- From tb303+@cmu.edu Wed Sep 28 02:41:01 1994 To: +dist+/afs/andrew.cmu.edu/usr1/tb37/list/list.dl@andrew.cmu.edu Subject: Spectral Interpolation Content-Length: 1505 X-Lines: 31 Status: RO Thomas C Butcher writes: > The sound is sampled, then broken into tiny time-slices (sort of like > breaking an hour into minutes... or seconds for that matter). Then, a > Fast-Fourier Transform is performed on each timeslice, and the spectral > characterisics of the sound are filed away. Once this process is > completed, the data can be used to re-create the sound using additive > synthesis techniques. This is why I'm looking for a K150fs. On a very related note there is some software which will take a sample and generate patches for the K5. My friend has this but apparently it isn't that good due to the limitations of the K5. > What's more, apparently some teams are using neural networks to enhance > the quality of the sound approximations. (you there, sami?) I did some work on using neural net speach recognition techniques for interpretting musical performances when I was at college. What I was basically doing was using a using a feature map first to generate feature for various musical performance techniques and then using this feature to recognise the techniques. You could use basically the same technique to do vector quantisation on sound spectra. > I think this technique of re-synthesis was used in designing the Technos > Axcel synthesizer and the Technos 16pi.... since I really know close to > nothing about the machines, I can't comment on them. Well I know absolutely nothing. Could you tell me close to nothing, I'd be interested. -- From tb303+@cmu.edu Wed Sep 28 07:08:01 1994 To: +dist+/afs/andrew.cmu.edu/usr1/tb37/list/list.dl@andrew.cmu.edu Subject: Technos synths and a little rant Cc: Content-Length: 2807 X-Lines: 56 Status: RO > This is why I'm looking for a K150fs. About 3 years after they came out (what, '90?), I saw one at a music store in Houston.. I don't remember how much they wanted for it, but it wasn't a measly fee. It was rather large. Anyway, the thing sounds quite nice, and if I knew anything more about it, I would have paid more attention to it at the time. > On a very related note there is > some software which will take a sample and generate patches for the > K5. My friend has this but apparently it isn't that good due to the > limitations of the K5. Ah, that's the thing. Spectral interpolation!! It's nice stuff. > > I think this technique of re-synthesis was used in designing the Technos > > Axcel synthesizer and the Technos 16pi.... since I really know close to > > nothing about the machines, I can't comment on them. > > Well I know absolutely nothing. Could you tell me close to nothing, > I'd be interested. OK, well as far as I know, the company Technos is out of business. The trade rumors are that once synthesizer manufacturers found out how powerful the machine was, they did some big industry move and dissolved the manufacturer. I'm talking about the Acxel thing. The 16pi was a keyboard that really never made it to the market- vaporware I think that was advertised around 1985. The Axcel was this huge box with a screen of like 500 led/buttons. They called it a 're-synthesizer,' and quite frankly, I have no idea about its specifications, operation, or anything else. I do remember reading something about the machine being able to do the spectral interpolation, so that's why I mentioned it. As far as I know, some of them were actually sold to educational institutions, but none were bought by individuals before the company went under. Everything I heard (just little notes, BTW) made it sound like a revolutionary product. Which it may well have been ... anyone remember the PPG Realizer? Industry sucks! Instead of having powerful synthesis machines now, we have soundcanvasses, JV-30s, and M1 clones. Oh well. I could ramble on and on about how I'm extremely dissatisfied with the synthesizer market, but would it change anything? Don't think so. Programmers these days are a minority. In fact, I've wanted to make this list for a long time for programmers- people who are interested in sound, synthesis techniques, and who are willing to spend time experimenting. There are so few programmers now, most synthesizer companies overlook us. Oh god I'd better stop before I step up to the soapbox! But anyway, I do need to say this: stick together! Keep programming! There are a lot of machines out there that have a TON of untapped potential. Maybe someday the manufacturers will wake up and make synthesizers for synthesists again! Tom From tb303+@cmu.edu Wed Sep 28 15:39:48 1994 To: +dist+/afs/andrew.cmu.edu/usr1/tb37/list/list.dl@andrew.cmu.edu Subject: Technos synths and a little rant Content-Length: 2025 X-Lines: 52 Status: RO Thomas C Butcher writes: > > This is why I'm looking for a K150fs. > > About 3 years after they came out (what, '90?), I saw one at a music > store in Houston.. I don't remember how much they wanted for it, but it > wasn't a measly fee. Ball park figure? I haven't a clue what they're worth or what I'd be expected to pay. I've got an entry in the gear wanted list for $400 as this would be a good starting price to haggle for a K5 but I get the feeling that this is far too low for a K150. > It was rather large. I heard it looks like an old military radio. > Ah, that's the thing. Spectral interpolation!! It's nice stuff. I've got a couple a synths that would be capable of doing some noddy additive stuff. The TG-77 could do 24 partial additive with complete control of each partial but with only 4 voice poly. If I thought the results would be worth while I'd write some software to do it but I don't think 24 partials would be enough. The K2k is in theory cabable of doing 96 partial additive with even more control but with only one voice poly. This would be even more hassel to set up. However the idea really does apeal to me. Does anyone know anything about square wave additive? I know that just as any waveform can be represented as the sum of sine waves you can also do the same thing with square waves. I think the maths might be harder though. Does anyone know much about this? Does it take more or less square waves to reconstruct a signal than sine waves? > OK, well as far as I know, the company Technos is out of business. etc... > The Axcel was this huge box with a screen of like 500 led/buttons. No wonder - it must have been a nightmare to manufacture and would have cost an arm and a leg. > Programmers these days are a minority. In fact, I've wanted to make > this list for a long time for programmers- people who are interested in > sound, synthesis techniques, and who are willing to spend time > experimenting. Hear hear! Nice one. Ab. -- From tb303+@cmu.edu Tue Oct 18 03:58:24 1994 To: electronica@andrew.cmu.edu Subject: Re: Minneapolis Things For Sale Content-Length: 716 X-Lines: 24 Status: RO PPGWave@aol.com writes: > KURZWEIL K150 ADDITIVE synth - $350 I've been looking for one of these for months and now I've got one. It should be winging it's way to me this very night. Thats another one ticked off my wish list. So now I have examples of all of the following synthesis paradigms: Analogue modular - Roland System 100M plus Obi SEMs Digital FM - Yamaha TG77 Digital time domain/sampler - Kurzweil K2kRS Digital frequency domain - Kurzweil K150 All I need now is a wavetable synth (ideally a Wavestation A/D) and my list will be complete. Are there any K150 owners out there who could give me some tips on the beast. (I'm planing on writing some analysis/resynthesis software for it.) Ab. -- From tb303+@cmu.edu Thu Oct 20 22:25:52 1994 To: ELECTRONICA@ANDREW.CMU.EDU Subject: Re: Kurzweil K150 and Korg poly800 Content-Length: 3108 X-Lines: 66 Status: RO Barry Klein writes: > On AH I saw reference to the K150 indicating $350 for it was a great deal. > Is it? The $350 price was no haggling too but I didn't want to haggle I just wanted the thing before somebody else picked it up. It's a great deal to me because I've been looking for one of these things for ages and this was the first I'd seen. I've never seen one for sale in the UK at any price so I didn't know what to expect to pay (I had an offer of 600UKP in the gear wanted list). > What kinds of sounds does it have? Is it a sample playback type of > synth? It's one of only two widely available additive synths produced. Ok there were other additives made - Synclaviers could do it, so could Fairlight CMIs - neither of these I would say were widely available. Synergies and other exotic beasts could do additive but these are about as rare as chicken teeth. There are some other synths that would let you build up static waveforms based on additive synthesis - notably the OSCar and the Kawai K4 (I think), but only the K150 and the Kawai K5 could do dymanic additive synthesis. Choosing between the K150 and the K5 five is actually quite tricky - the K5 has a nicer display, individual outputs and is fully programmable from the front pannel. The K150 OTOH has a one line LED display, only a single mono output and has only limited programming capabilities from the front pannel. Where the 150 wins over the K5 is in the actual power of the synth engine. A single voice in the K5 could consist of two layers each capable of producing 64 partials which where controlled in groups by (I think) four eight stage envelopes per layer. A program on the K150 OTOH could have up to seven layers of 256 partials, each partial having it's own 256 stage envelope. This means that the 150 is capable of excellent quality resynthesis. ie given the right software you could take a sample, analyse it using an FFT and then recreate almost exactly the sample on the K150. This recreation would not be a sample but would be a synthesised version of the original sound. An important point to note about the K150 is that as it stands you cannot actually program the additive engine. To do this you need to upgrade it to an fs version. Given this modification you can then program it via sys-ex. Hoever the software to do this was originally written for an Apple][. So why did I buy the thing? Well I'm going to write some three dimensional spectral interpolation software for it and make some wild sounds you can only get out of an additive box. > And on the Korg poly800. I saw one for $130. It has a midi interface. > What midi information does the synth send? Probably not much more than note-on/off, pitch bend, modulation and program change. > With 8 voices I was thinking it may be a cheap source for spares if > nothing else. Bear in mind that it it only has one filter for all it's eight voices and that the oscilators themselves are rumoured to be generated in software. Ab -- From analogue-request@postbox.acs.ohio-state.edu Wed Nov 2 14:04:18 1994 Sender: analogue-request@magnus.acs.ohio-state.edu To: analogue@magnus.acs.ohio-state.edu Subject: PPG modular, DIY modular and K150 info Content-Length: 3893 X-Lines: 95 Status: RO This weekend I met a guy, Ruediger Lorenz (Malte, you know him), a real collector, not to say a freak. He owns virtually tons of gear, six or so modulars, and 20-50 synths of other proveniences, from Vermona to Morpheus. OK, he has a PPG 300 Modularsynthesizer and he has built one by himself, which is very interesting. I took pictures of both, which can be viewed via WWW at http://www.physik.th-darmstadt.de/~georg/synths.html Read till end, there is some K150 info waiting. --------------------------------------------------------------------------- Description of the PPG (from top to bottom, from left to right): First Row: 313 Sequential Switch, 314 Clocked Sequential Control, 310 Random Voltage Generator, 309 Signal Mixer, 314 Clocked Sequential Control, 313 Sequential Switch Second Row: 301 Voltage Controlled Oscillator (4 times), 309 Signal Mixer, 301 Voltage Controlled Oscillator (4 times), 309 Signal Mixer Third Row: 305 Low Frequency Oscillator, 317 Voltage Controlled Filter, 307 Dual Envelope Generator, 315 Dual Analog Multiplier, (VCF-ENV/FBY02) not PPG, (LFO'S AND PITCH VOLTAGES) not PPG, 308 Dual Envelope Generator (2 times), 303 Voltage Controlled Modifier, 311 Power Supply 322 Duophonic Digital Keyboard --------------------------------------------------------------------------- Description of the DIY synth, (I think he called it Modulog) First Row: Noise/Tri Proz, Osz, Graphic, Harmonic, Filter, Amp 1/Amp 2 Second Row: LFO, Osz, Graphic, Harmonic, Filter, Amp 2/Amp 4 Third Row: Waveform Proz, Osz, Graphic, Harmonic, Filter, Filterbank Fourth Row: Ringmod, Waveseq, Karplus Strong, Sheppard, Hall/Chorus, (Power/MIDI) ?? He build the modules from a mixed bag of kits from various manufacturers (Paia, Formant, some German magazine stuff). Here some additional info: Osz: The oscillator is mainly used to drive the Graphic/Harmonic Generators, it's a VCO, I think from Paia, probably the CEM one. The topmost Osz contains a digital osc too, from Digisound, waveforms stored in a EPROM are read out, clocked by the VCO. The two rows of LEDs and the two red buttons belong to the digital osc. The other Osz are prepared to hold DOs, if he can get them from Digisound, anyone knows the actual address? Graphic: With the 16 slider you can "draw" the waveshape. All sliders can be controlled by control voltages. Harmonic: With 16 sliders you "draw" the spectrum, again controllable by CVs. He uses two analog sequencers (one Korg SQ-10 , one PPG 314) which are triggered when pressing a key to trigger lag processors (seen on top of the unit) from the gate outs of the sequencers. The lag processors out is sent to the CV ins of the Harmonic Generator. Sounds pretty cool. The three HGs are driven in different orders, the sound is a bit PPGesque. Filter: Traditional VCF from Formant. Same for VCAs, don't know the source. Filterbank: Sort of Graphic EQ, but you can set the volume of the bands to zero. The output of each band is available via the row of jacks. I forgot to ask about the Waveseq and the Karplus Strong module, but you may know these buzzwords. The Sheppard module is from Paia, they had in their 1992 catalog, the one and only I have. And finaly, the guy is 53, so he had some money back in the heyday of analog synthesis, and he is recording music, up to now 5 LPs and 5 CDs. So, if you wanna hear the PPG doing the rhythm-loop, ask me, I forward you his address. BTW, he paid DM 500,- for the PPG in 1988, someone had it sitting unused in his loft, gathering dust. A last information for you Kurzweil K150 junkies. He sent the mainboard of his K150 to Hal Chamberlin, who upgraded it to the FS version for $25. And he has the Apple ][ software to program the unit. So, if anybody is interested, ask me. Thanks for reading, Georg. From tb303+@cmu.edu Tue Nov 15 08:43:58 1994 To: electronica@ANDREW.CMU.EDU Subject: Re: K5m additive, any tips ? Content-Length: 2889 X-Lines: 67 Status: RO Herbert Janssen writes: > Anybody know when this came out/was discontinued ? I'm make a total guess at 86/88 on the grounds that I remember seeing one in a music shop in Manchester at around that time. > Also anybody have an idea how this machine works internally, > there can't be 128*16 sine osc's in this box, right ? I don't know about the K5 but I just got the developers documentation for the Kurzwiel K150 and I'm dieing to tell someone about it. The K150 is additive too. It's hardware is split roughtly into two parts: a CPU board which implements all the on board editing, midi interpretation, and controlling of notes (envelopes etc). The other part is the sound engine which actually implements 240 virtual oscilators half in software, half in hardware. Yes, I was shocked too. The sound board consists of two lookup tables, one containing a sinewave, the other containing two noise samples; some RAM to act as resgisters, and a universal timer IC which actually does most of the work. For each oscilator there are four registers: phase - a pointer into the lookup table; phase inc - the increment added to the phase register on each clock tick, used to control frequency; level and level inc which together are used to create envelopes. I'll have to read up on this but I think the timer chip has enough power to process the contents of each of these registers once every sample period. The CPU meanwhile writes to these at a much slower rate in order to control the envelopes, pitches and modulation etc. An interesting feature of the K150 is that it actually implements partial stealing instead of voice stealling - if it runs out of notes it steals partials from other notes rather than killing them entirely. > Is there any editing software for the Atari ST ? I think the FFT software ran on the ST but as usual I don't remember who wrote it. > Any starter tips for getting into editing those ... 200, 200+ ? ... > parameters ? Well aparently all the original voice models in the K150 were generated by editing hugh lists of numbers generated by analysis programs. Consider yourself fortunate that you have a graphical editor at all. > How do these wierd filter parameters work: > "cutoff" - OK ! > "slope" - really ? > "flat" - what ? Good question. I think the filters operate mathematically on the levels of the partials in the voice so the parameters probably bear little resemblence to anything from the analogue world. OTOH this could be an advantage. They might let you do some wacky things that you couldn't do with less `virtual' filters. > Does MIDI expression CC#10 always just affect volume ? Don't know. Happy hacking. When I get some editing software for the K150 we should be able to exchange some general additive tips. Ab. -- From tb303+@cmu.edu Tue Nov 15 16:15:43 1994 To: electronica@andrew.cmu.edu Subject: Re: K5m additive, any tips ? Content-Length: 1247 X-Lines: 32 Status: RO Hi, Ab wrote: > I don't know about the K5 but I just got the developers documentation > for the Kurzwiel K150 and I'm dieing to tell someone about it. Yeah, very interesting description, especially about the partial stealing and the filters being implemented as level modulation of the partials. I'd love to have some "developer description" for my K5. > Well aparently all the original voice models in the K150 were > generated by editing hugh lists of numbers generated by analysis > programs. Consider yourself fortunate that you have a graphical editor > at all. Well I don't have *any* software for it at the moment - let alone a graphical editor, altough I have still hope someone will have mercy ;-) I began to write some software though yesterday evening, accompanied by loud cursing because I really hate hacking together some dirty code just to get things running (I want to use the patches in glib format from the net). > Happy hacking. When I get some editing software for the K150 we should > be able to exchange some general additive tips. Yes, I'd love to, you will certainly hear from me ;-) ciao Herbert P.S. what does your 150 look like now ? really like a huge old tube radio ? From tb303+@cmu.edu Tue Nov 15 22:59:12 1994 To: electronica@andrew.cmu.edu Subject: Re: K5m additive, any tips ? Content-Length: 1657 X-Lines: 38 Status: RO Herbert Janssen writes: > Well I don't have *any* software for it at the moment Whats it like editing it from the front panel? There is now way to edit the voice models in the K150 apart from via sys-ex. The original editor looks pretty powerful and you can even get the source code for it - 800K of 6502 assembler - my idea of fun. So I'm either going to have to track down an Apple][ or write my own from scratch. What platform are you writing for? I'm not scared of getting my hands dirty but for the past 5 years I've only written C++ on Suns so hacking some horrible implementation of C on a 1Meg ST isn't going to be fun. Do you know anyuthing about developing on the ST? > P.S. what does your 150 look like now ? > really like a huge old tube radio ? If it had more knobs on the front it would. It's actually a big (4 unit) black metal rackmount box with rack handles on the front (not unlike a military radio). The user interface consists of a one line alpha/numeric LED (good and visable at least) and two twelve button keypads. The only knob on the thing is volume. The thing looks like it would withstand a nuclear blast at close range. Compared to the K5 it looses on user interface and outputs (the K150 has only a single mono output), but gains in terms of power of the sound engine (one 256 step loopable envelope per partial per model, and you can put a different model on every key if you want to.) However all this poer does need some good software to really get to grips with it. Ab. PS This is getting a bit off thread. Maybe better to take follow ups to private mail. -- From tb303+@cmu.edu Wed Nov 16 15:04:46 1994 To: nwilson@morgan.com (Ab Wilson) Cc: electronica@andrew.cmu.edu Subject: Re: K5m additive, any tips ? Content-Length: 5581 X-Lines: 109 Status: RO Ab Wilson writes: > PS This is getting a bit off thread. Maybe better to take follow ups > to private mail. I was really talking about the detour into Atari ST programming here. However I've had a couple of private mails asking me to keep this thread on the list so I'll keep the stuff which is vaguely synthesis related here. (Herbert and Mikael do you want me to CC to the list the mails you sent me?) Timothy J. Prezzano writes: > So what sort of sound have you been getting from it? I don't have the ability to program the voice models yet (thats what I need the software for) so all I've been able to do is play around with the models it came with. The K150's archetecture is like a sample playback synth. You have the voice models which you can't touch without an external editor (almost like samples), then you have the actual programs which are built out of key ranges and layers and voices. Mine seems to have come with a full set of voice models according to the documentation I got. These aren't really very interesting in my book. There is a really nice piano model which has to be the best synthesised (as opposed to sampled) piano I've ever heard (it was created by the analysis of a sample so I suppose it doesn't count). There are slap basses and saxes and chimes and things which all basically sound like good quality multi samples except there are no detectable looping noises. And there are resonant sweepy type sounds which Hal Chamberlin must have put there just to demonstrate what you can do with his software. They have a very sharp resonant spike but none of the nasty self oscilation noise you get with real filters or their digital simulations (don't flame me - self oscilation is fine when you want it but a pain when you don't). This is one of the possible advantages of additive synthesis - you only get partials where you want them. However since the models are preset you can't really do a lot with them. The next level up is the layer which is where most of the front panel programming goes on. You can do more with a layer than you would think. Each layer has it's own very strange LFO which you can only really use for vibrato. However the LFO is noteworthy in that you can modulate it's shape from a midi controller (I think just about all the program level parameters are controllable in real time over midi). It has two basic shapes - square and ramp - but you can warp the square wave into pulse and the falling ramping into a triangle, into a rising ramp - quite a bizarre ammount of control for a simple LFO. There's also parameters for psuedo delays which let you set a number of repeats, time between repeats, level change amount (you can make echos that get louder as well as softer), and pitch change ammount (you can make echoes bounce around in pitch both up and down at the same time). As is usually the case with Kurzwiel all the parameters are in sensible units like dB, cents, ms etc. About the most interesting thing (which really sets the unit apart from a sampler) is that you can timbre shift any of the models and you can do this in real time from a midi CC. Timbre shifting is basically about playing the timbre of one pitch at another pitch. This is what happens when you transpose a sample except with a sample the play back speed changes too. On the K150 the playback speed stays the same so you can transpose a note from the top octave right down to the bottom without the envelope getting all messed up. Anyway this basically opens the door to morphing - just create a model that alters radically across it's keyboard range then modulate the base key number thus changing the timbre. I played around with this for a bit - you'd be surprised at just how much you can fuck up a piano sound like this. > I'm very curious to one day hear what that synth engine can sqwank > out... So am I. Herbert Janssen writes: > Probably such macro editing facilities are a good thing to extend > when making software for it. Hal Chamberlin's software let you draw response curves in all dimensions - the envelope for a single partial, the excitation over all partials for a single timeslice, and the response to velocity for all partails for a given timeslice. I don't know what tools there were for integrating these different views but these are the sort of things I'd want out of a graphical editor. I like the sound of the macros in the K5. I've always wondered what it would be like to play with the levels of particular series of partials (ie all odd) rather than the series as a whole. I've been thinking about this quite a bit. There are really two ingredients that make up a patch on an additive synth - the frequency series (ie all odd, all even, some arbitrary collection), and the formant curve superimposed over this. I'd like to be able to take the frequency series from sampled sounds - since these would be a pain to set up manually - and superimpose my own formants over it - since this could be done simply by drawing a curve. (Stuff related to ST programming to be sent by private email.) > I played around with the EQ section yesterday: quite easy to get > formant-like effects from a sawtooth basic wave. The K150 has an EQ section too. I haven't tried playing with this yet (since like the K5 it isn't mod'able). I've got models with the basic waveforms so I'll have a go at building some formant type voices myself. Ab. -- From owner-emusic-l%AMERICAN.EDU@vm.gmd.de Mon Dec 12 15:28:50 1994 Reply-To: Electronic Music Discussion List Sender: Electronic Music Discussion List Subject: additive synth hardware Comments: To: Electronic Music Discussion List To: Multiple recipients of list EMUSIC-L Content-Length: 1730 X-Lines: 36 Status: RO Andy Farnell writes: > I am looking for an additive sythesis card/hardware. Either > stand alone or platform specific for PC/Mac/Sun. Well Flange me old mate I may have just the thing for you. I've got a Kurzeil K150fs complete with full documentation. I brought it back form the State last time I was over there. > What I need is real time (I mean REAL real time, like > instantaneous) control of a good few hundred partials. You'd have to write some new system software to do this (68000) but it shouldn't be that difficult, I've wanted to do that myself. The basic hardware has two hundred digital oscilators. These are really just pointers into a look up table. Each oscilator has four registers: phase, phase increment, level, level increment. Phase stores the current point in the lookup table. Phase increment controls the change in phase at each clock tick and hence the frequency. Level controls the output level of the oscilator (gasp), and level increment allows you to create envelopes without using the main processor except when there is a break point in the envelope. A K150 voice consists of a list of instruction (yes a program) which modify these registers. That's why you can't do realtime control of the partials. To do so would be a fairly simple hack. You write an interrupt routine which picks up the midi CC. This then inserts new instructions into the program to make the necessary changes. > I have heard that the IRCAM 4B is good bet.. does anybody > have any info on this ? Don't know but I'll bet it's pricy. If you come up to London to collect it I'll let you borrow the K150 for the price of writing the new system software. Ab. From owner-emusic-l%AMERICAN.EDU@vm.gmd.de Wed Dec 14 18:39:43 1994 Reply-To: Electronic Music Discussion List Sender: Electronic Music Discussion List Subject: Re: additive synth hardware Comments: To: Electronic Music Discussion List To: Multiple recipients of list EMUSIC-L Content-Length: 1972 X-Lines: 40 Status: RO > The only big problem I can think of is this: It might be quite > possible to get the m/c to 'respond' to real time control but if > the only input stream is via MIDI it is quite another matter > to then stuff the (potential) 200+ 16bit real time controllers down > the MIDI link, I know cos i've tried it with a modified K5 :-) What is it you're trying to do? Specifically why do you need realtime control over all the partials? If you want to do resynthesis then the 150 is pretty capable without any modification (much more capable than the K5 for instance). The OS I've got lets you download instrument models over midi with an envelope per partial and no fixed limit to the number of stages per envelope (only 64 partials - though in practise this seems to be plenty. The grand piano model for instance is very realistic). This gives you pretty good control. If you want realtime control above and beyond this then it's probably sensible to be using some sort of abstraction anyway. If I recall your PhD was to do with abstractions for sound desciption anyway wasn't it? > Thats why I was really looking for a dedicated card, ie with a > direct parallel/bus conex. It sounds to me like the K150 is already doing a lot of the processing you want to do on the host computer. I think with a little bit of thought you could work out a way to share the processing load between the host computer and the K150 and keep the communication overhead to within the limits of midi. > Anyroads, I will come up to town and take a look, haven't seen you > for quite a while, plenty to catch up on. Hmm private mail I think... > Anyone know much about Kyma/Platypus system and what it can do ? I contacted the Kyma people when I heard about this thing and the sent me all the bumf. I'll see if I've still got it. What I remember about it wasn't that impressive for the price. The Mars systems sounds a lot better to me. Ab.