Son of SID?

[middleman]

I wonder how much it would cost to get some new sids manufactured? Would you go with the old SID or a more up to date version? Or maybe an up to date version with a SID bit for compatibility? What new features would you include?

Edited October 19, 2011 by middleman

[m00dawg]

This comes up from time to time so I would recommend searching the forums to figure out why this hasn't been done (cost, licensing). If you want the closet thing to a modern reproduction, check out the SwinSID stuff. It's MidiBox compatible as in, you can use and control the extra waveforms from it not found on the original SID. I haven't built one yet but could be something to look into, particularly if you couple it with an analog filter (such as the SSM2044), thereby getting around some of its inherent limitations.

SwinSID is glorified emulation on a hardware chip with some extra features. If you wanted to go a step further, there have been projects to design a SID on an FPGA.

Probably all you can hope for right now, I'm afraid.

[middleman]

Yeah, I've been watching the swinsid threads for a while, but development and interest seems to have fallen off lately. I would love to replace the empty sockets in my 64's with something like that, but I get the feeling that without a decent PCB burnt pic service like we have here, it would scarcely be feasible. The FPGA/FPAA route looks very promising, but who has the specialised skills to accomplish this kind of task? What I was thinking is that if some kind of super SID type chip were to be specc'ed out, maybe one of the large scale synth manufacturers would get involved, and this community could benefit with the spin-off. Dunno. Just dreaming really. Now that it is getting so hard to acquire chips, it would be nice if there were a substitute that the community could focus on.

[m00dawg]

The SwinSID might not be updated all that frequently, but it still seems alive and well to be and still a viable alternative. SwinSID is ARM and AVR, not PIC; but making a burner should not be terribly difficult. If you ask nicely on the SwinSID forums, perhaps someone would be willing to burn it for you and ship it your way.

The FPGA route is more an expansive and large solution than it is complicated. You can likely find a SID written in a hardware description language already somewhere in the wild. I have seen multiple projects (in passing) that implement the SID in an FPGA.

Otherwise, your dream is probably just that - but it is a dream shared by many.

[Hawkeye]

For me, knowing to have some 25+ year old sound chips, that have certain (very hard to emulate) imperfections, is half of the fun. Thinking, that these very same chips have played the tunes of Hubbard/Galwyn/Huelsbeck to the total awe of us 80ies computer kids (me was one, too) is the other half :). They have souls. And you put them in your gear. How much better can it get? :-)

The SIDs have some unique imperfections,e.g. some internal distortion when all oscs are active - very hard to simulate... also, when emulating the filters digitally, there might be some harsh "aliasing" tones... which you could offset with a nice analog filter to achieve "perfection" again... but then... was it ever "perfect" ? :-). I like the filters of the SIDs a lot... no, more than a lot... they are imperfect, with audible "dents", but that defines their character :-).

Personally, I think the SIDs will still last for a while (the prices will rise, though - good investment to stock some). There have been many, many C64s.

Greets,

Peter

Edited October 20, 2011 by Hawkeye

[m00dawg]

Agreed, the SID has a charm about it that would be hard to keep with modern fidelity were a "SID 2.0" to be made. The idea, in it of itself, is intriguing, but for different reasons than why I like SIDs to begin with. The NES and GameBoy are the same thing. Even the squarewave generators between the two have a different inherent sound that my modern Virus can't reproduce.

That said, I still find the SwinSID compelling for what it is. I would call is different though instead of better/worse. Probably a suitable replacement for the SID in a real C64, though. In a MidiBox, the SwinSID would be neat for making sounds different than the real SID and am excited to have it in my arsenal. Not to replace, but to augment, but original SIDs.

[middleman]

For anyone who has not got a cache of sid's its very hard to get going. Locally 64's are selling for around €80 on average, adn they are not too plentiful. On ebay the postage kills the deal. I got stung a couple of times before I learnt to beware. (for 'untested' read 'dud'.)

I have built a Sammitch (brilliant piece of design), but the idea of anything more ambitious is out of the question for me, and most others who are a bit late to the party, due to the net cost of the SIDs

A swinsid based MB-6582 with an analog filter bank would be a lovely project, but would definitly need a dedicated pcb in order not to drive a constructor insane.

I'll head over to the SwinSid people and see what's happening. If anyone is interested in following this line, keep posting. I don't want to let the dream go just yet!

Edited October 20, 2011 by middleman

[m00dawg]

You can use the port on the map to wire up CV inputs into an analog effects box if you wanted to go that route. I've been going back and forth on putting my analog filters inside my rackmount SID or making it a separate box (so as to be used for other synths). Part of that reasoning was for cost and sourcing the SSM2044 chips which are also somewhat expensive to find.

[middleman]

I got a 2204 on ebay, but I havn't tested it yet. Thats another area which needs looking at. Is it possible to program a pic to work as a DSP? or is is necessary to go down the fully blown DSP route?

There is a DSP course currently being run in Elector magazine, but with the hardware at £115 its not something to just jump into.