Right for the Ultrasound, i've missed it. Yes, it might be trivial, but as with every hardware piece there is a sound to it, which I like and want to know what's possible with them (specially since I can't afford a nice analog synth at the moment). I used to think of these devices as hobbyst's things, but (specially regarding the MOS SID) their demoscene users tend to take some art out of them. And they actually were designed with a musician's perspective (according to Bob Yannes). And if you get into hearing the sound of the waveforms and the filters, they have an undeniable character, which you can probably match by drawing waveforms on an editor, or with a capable synth like zyn, but then every other sound in the world is synthesizable or samplable. It's not only nostalgia, I want to believe.

On Thu, 10 Jul 2014, Rob wrote: Actually FM can be done in pure analog synthesis. It's nothing more or less than one oscillator's output being used as a modulator to be sent into the input of another oscillator on the same synth, and lots of analog synths can do that. You can do FM patches on a Minimoog. Yamaha likes to act like they invented it of course, and they've been acting that way so long now that we tend to think now that it has something to do with digital. :) Actually where Yamaha came into is was this: The part of a synthesizer where analog is most needed for the sound is the filter. Making a digital oscillator is not that tough and doesn't sound too bad, and neither is a good digital envelope generator hard to make either, but in the 80's, it was proving really hard to make a decent, musical-sounding digital resonant filter. By the mid-80's, most keyboard manufacturers were making hybrids (Roland JX series, Ensoniq ESQ-1, Korg DW series, Sequential Prophet-600, etc.), where everything was digital *except* the filters, which were still analog. Mostly all people wanted from digital if it came right down to it was stable tuning, and since the filters don't affect your tuning, these hybrids would have been enough to meet what people were expecting, and they could have just kept on making them. Of course, the keyboard companies kept going, and we got pure digital (with digital resonant filters) later, but for what it's worth, the hybrids pretty much fulfilled everything people were actually asking for, and we could have settled on that and everybody would have been happy. Anyway, Yamaha, unlike all the others, decided to try to get ahead of the game by making a pure digital synth before anyone else, and they tried to solve the filter problem...by NOT HAVING ANY FILTERS. Bahaha! Their idea was that if you had enough oscillators cross-modulating eachother, the resulting wave would have so many interesting harmonics in it that you could forgo the usual resonant filter completely, and it would still sound good enough for most people. The DX7 had six digital oscillators per voice (though Yamaha calls them operators instead of oscillators). Most people now agree, now that a popular taste for analog has returned, that the DX series was pretty cold and sterile sounding. However, they were dependable. You could count on a DX7 to not need any warm-up after you turned it on, and a patch on a DX7 would sound the same every time you used it. As for the sound, in pop music it was rarely heard without some kind of effects processing, so usually you never heard raw dry DX7 that much anyway. People actually bought zillions of them. The irony is, Yamaha was theoretically right. With enough oscillators cross-modulating eachother, you could make filters optional. Where they got ahead of themselves was thinking they could do that with only six per voice. Dozens or hundreds, yeah maybe. Six, no. That's true -- even digital synths have character. An OPL3 chip is probably not the hardest thing in the world to make an emulation of, but there's always some quirk of the hardware that gets lost in the translation when you try to make a software equivalent of almost anything. I used to have an Elektron SidStation in my studio (sold it), but it was a device like you're talking about -- a little Midi-controlled silver box with a real SID chip in it. -- + Brent A. Busby + "We've all heard that a million monkeys + Sr. UNIX Systems Admin + banging on a million typewriters will + University of Chicago + eventually reproduce the entire works of + James Franck Institute + Shakespeare. Now, thanks to the Internet, + Materials Research Ctr + we know this is not true." -Robert Wilensky

On Thu, 2014-07-10 at 10:36 -0500, Brent Busby wrote: The DX7 does some sounds very good and it does/doesn't sound cold and sterile (it's relative to the kind of sound that it should produce), but it's very time consuming to program a DX7 sound. Some sounds can't be done with a DX7, at least not without external filters/modulation and for sure, some sounds even can't be done with external filters/modulations. Even a Prophet 5 doesn't sound that full and warm, as usually an analog Roland, even a Juno-106 or Oberheim etc. do sound. It's wanted that not all good analog synth produce the same kind of warmth. OTOH compared to a Linux sound sampler, sequential samplers 2000/Studio 440 don't provide that realistic sounds, but they provide an unique kind of warmth, that a Linux sound sampler doesn't provide. IOW it's good to have a few devices that sound very full and warm, but some devices should provide a less full and warm sound. In professional music even some of the cheapest synth were (are) used in addition to good synth. The Yamaha TG33 generate cheap sounds that can be mixed by a joystick and than the sound becomes a good and full sound. Mixing several thin sounds to produce a full and warn sound was often done by using several synth. Btw. the TG33's Vector Control would be a nice feature for e.g. Yoshimi. The TG33 does record a sequence of mixing and/or fine tuning 4 sound sources. Regards, Ralf

On Thu, 2014-07-10 at 11:03 -0400, Rob wrote: I used the C64 with Speech Basic (or what ever the FLOSS sound sampler BASIC was named), build the inexpensive sound sample circuit that was using a joystick port (the circuit diagram and PCB layout were available for free too) and in 65xx assembler I programmed a MIDI extension for Speech Basic and than wrote a MIDI sound sampler. I bought a MIDI interface, but it's also easy to DIY. The sound was produced by SID. The gear is boxed, theoretical available for me, especially useful books such as the C64 Intern, but I dislike to search the cartons for the C64 gear. The sound of the SID is fuller than digital emulations of analog synth are, but a very, very serious drawback of the SID integrated to the C64 is secondary noise, so the less full sound by emulations of analog synth IMO is more usable than the SID integrated to the C64, that nearly is unusable regarding to it's secondary noise. You can't completely filter out the secondary noise and as soon as it's halfway filtered out, so that you might be able to use the sound for production, the sound quality has lost the fuller sound and will be less full than an analog synth emulation. JFTR in the 80s the C64 was more expensive than a PC nowadays is and some good analog synth based on CEM/Curtis chips were not more expensive than a C64, OTOH sound samplers were very expensive, so the lowest quality C64 sound sampler I programmed was useful, but IMO it never was useful to use the C64 as an analog synth. IMO no Linux analog synth emulation can compare e.g. to my Oberheim Matrix-1000 and even FM synthesis, e.g. Hexter can't compare to my first generation DX7, but Linux analog synth emulations are better than using a SID integrated to the C64. Using a C64 nowadays isn't worse the hassle and using it as an analog synth never was worth the hassle. 2 Cents, Ralf

Today, in my country, the cheapest decent synth is around the $5000 mark (pesos argentinos). This is for novation digitals, analogs start at $8000 if you are lucky. On the other hand, a yamaha based card is $140 and a c64 is around $700. In my mind that makes it worthy. Plus the simple architecture (at least simpler than today's computers) makes it lots more accessible to code it in assembler. Well, that depends on what you are after. Actually there are at least a few rock bands using commodores, so at least for them it's worth it. How did you achieve sample playing on a 3 oscillator chip?

On Thu, 2014-07-10 at 19:34 +0200, Ralf Mardorf wrote: ... for recording ;)

On Sat, 2014-07-26 at 07:09 +0000, Kaza Kore wrote: Much effort/money to get too much secondary noise. Instead of using a C64 and DIY, I would take a look at Ebay and e.g. get http://www.ebay.com/sch/i.html?_nkw=Roland%2BMT-32 The MT-32 provides the sounds that are provided by many free sound fonts, I guess most fonts sampled the MT-32 sounds. Less people are aware that the MT-32 pseudo analog section can be programmed and that it does produce some very good virtual analog sounds. I own one, but don't use it anymore. But I used it a lot with the C64 and Atari ST. IIRC there also is a Linux software available to edit it, but I might be mistaken.

On Sat, 26 Jul 2014, Ralf Mardorf wrote: I don't know 20 years earlier, but my first MIDI HW build was late 80s (88 or 89) A MIDI filter based on a Z8. Open if there was anyone to give it to :) 16 switches on the front... whichever switch(s) were up that channel got through the rest got bit bucket. Maybe 100 lines of code... probably less. (it has been a while) It also had lots of through ports... because they were easy to add. I also made a MIDI switchbox so I didn't have to change my plugs to change masters (keyboard/computer/tape sync). Their use has gone away pretty much now that most things can be done internal. I have a DX7 and cheap drum pads for input. A Roland mks50 and the DX7 could be used for output (I also have a TG100, but I think the internal sounds are probably better). -- Len Ovens www.ovenwerks.net

On Sat, 2014-07-26 at 20:06 +0200, Ralf Mardorf wrote: My apologies for replying to myself, but I right now noticed that the C64 book provides a PCB layout for a stereo SID. It's a book in German.

On Sun, 2014-07-27 at 05:37 +0000, Kaza Kore wrote: http://www.ucapps.de/mp3/midibox/bsp3.mp3 http://www.ucapps.de/mp3/midibox/bsp2.mp3 There isn't the background noise I know from the C64, sounds usable, OTOH the sound quality isn't better than that of a virtual synth. Without the background noise I experienced with my C64, this project is interesting for some stand alone devices, emulating those old Roland drum and bass machines, there maybe are such projects: http://midibox.org/forums/gallery/image/2171-mb808-sequencer-round-caps/ Resume: In combination with a Linux audio computer I would consider to use a virtual synth to get this sound quality, but since there seems not to be the background noise the C64 produces, the midibox is interesting as a standalone device.

I really wish things were as simple today as they were in those days. That's were I think the RPi and similar projects fail; they are fully fledged 32bit computers with modern architectures. Programming them barebones is possible, but I can guess it's more difficult than it's worth. Retro style makes it for a cool learning experience. Actually the c64 deserves it's own special place in my list. When I get to hold one of them, it will be to program it within it's machine code monitor, or maybe find some way to actually type on my PC. Learning the assembler and the registers seems hard enough to add making an interfacing kit (but then there is the possibility that learning the register leads me to better understand how to interface it...).