Commodore 64 and Raspberry Pi 4 = Synth6581 | Hacker News Reader

Commodore 64 and Raspberry Pi 4 = Synth6581

raspberrypi.org

79 pointsfangorn24 comments

With this, and the dual-SID add-on boards for real Commodore 64s there are not enough original SID chips to go around. Already hawkers on eBay rip them out of working C64s and sell them separately at jacked-up prices, to people who have managed to buy C64 motherboards at a lower price because they are missing their SID chips, probably in fact the very same SID chip they are now paying over the odds for separately.

There are a few pseudo-SID hardware projects[1,2,3] that can produce 'good enough' SID-like sounds, maybe the more honourable thing is to rework this project into something that uses those ?

---

[1] https://www.c64-wiki.com/wiki/SwinSID

[2] https://github.com/Sgw32/UltiSID/tree/hal_usid

[3] https://www.retrocomp.cz/produkt?id=2

I used to know a guy who collected C64s in 2009, and he was complaining back then that buying them on eBay was an exercise in trial-and-error, for the exact same reason that the SID chip was missing so often. :(

You could probably blame the SIDstation (and its successors) for that. People have been putting SIDs into various synths (including DIY like the MIDIbox SID and others) since the late 90s.

There's also the fact that these chips are prone to failure, partly because they presumably are sensitive to static electricity and partly due to the design of theC64 which offers little protection.

The worst part is I actually have a few SIDs around I intended to use for projects, but because of my chronic disorganization... I couldn't find a couple of them last time I looked. Lying in a pile of parts. I'm sure there are people who would hang me for this :-)

Thanks for those links; I was about to ask if we could just make new SID chips, but the RetroComp link in particular appears to be the answer “yes”.

The ARMSID is a software emulator, running on an ARM microcontroller. While it's plug-and-play with the real thing, I would not equate it with a "new SID chip".

For some it's probably fine, for others it's just not the same.

There are FPGA recreations too which are not software emulation. Not quite an ASIC but realistically as close it at gets at this volume.

[1]https://www.fpgasid.de/

[2]https://www.facebook.com/plugins/post.php?href=https%3A%2F%2...

I'm working on the latter one, Sidmax, so if there are any questions happy to answer.

Since a lot of the distinctive sound of a SID comes from the analog components which still need to be emulated in a FPGA solution does an FPGA solution really get closer to the sound of an actual SID than a quality software emulation?

FPGA does not really offer advantages of accuracy, though it does offer advantages in time precision. This is important when putting it inside a C64. Of course a much faster CPU can also be used to handle real cycle accuracy in sync with the original clock.

Part of the sid sound comes from the non-linearity due to approximation in its design. For example the state variable filters are very non-linear on the 6581. It even differs significantly from each individual chip to chip[1]. The best software models I'm aware of are in residfp[2] and jsidplay[3]. After some work in octave I found the filter model can be recreated without the large lookup tables used in this software.

Of course getting the digital part correct is also important. There is a lot of documentation on this resulting from the decap [4].

Some features like bitfade, multiple driven bits in the wave corruption etc do not translate well to FPGA. They can be approximated with timers and lookup tables.

[1] Filter response chart: https://csdb.dk/forums/?roomid=14&topicid=105719

[2] residfp: https://bel.fi/alankila/c64-sw/index-cpp.html

[3] jsidplay2: https://haendel.ddns.net/~ken/

[4] Sid internals: https://sourceforge.net/p/sidplay-residfp/wiki/SID%20interna...

Although I haven’t gotten any obviously resurfaced SID chips (something that, for one reason or another, does happen often) the majority of SID chips I’ve gotten off of eBay (sample size <10 for what it’s worth) have been junk; either not working at all, or only partly working.

Still managed to get myself a working dual SID configuration with one 6581 and one 8580. But even That is obviously not ideal, since you can’t do cool stereo SID with that, at least if you want to it sound good. I thought maybe it would be OK, but a lot of tracks really need one or the other, the filters are usually pretty mucked up on one or the other. But you know, beggars can’t be choosers. Maybe I could do a stereo setup with one SwinSID.

Even though I didn’t grow up with a C64 though, man, nothing really beats the real thing when it comes to SID. It’s a shame it’s so damn challenging to emulate.

Some people in the Parallax Propeller community have been hard at work on a high quality emulation. First gen is SIDCOG, and a second one, running at some 400khz is in development.

You're forgetting about FPGASID[1], which provides dual SID capability and high quality simulation of the original SID's analogue parts, not to mention can be toggled between the two different SID chip versions. It also supports paddles, like the real SID chips, but unlike SwinSID.

[1] https://www.fpgasid.de/

I've sometimes wondered if neural networks could be used for replicating the differences between different 6581 chips, by having a computer automatically interrogate the chip and the neural network.

Optimally the end result would be a 6581 chip model (or possibly just some parts of the chip) with 2d input indicating what kind of chip to use today, possibly another input for temperature, and then driven by those existing 6581 emulators.

It could potentially use insane amounts of compute power just to run it, if it uses the same approach as those researchers that did a speech synthesizer that produced one sample at a time from the network.. But it could also have very compelling results.

Still waiting on a neural network sampler. I'm worried that by the time someone makes one all of the sweet artifacts will be ironed out though.

See also the AY-3-8910 if you cannot find a SID https://en.m.wikipedia.org/wiki/General_Instrument_AY-3-8910

Well the SID has a powerful resonating filter, this one has no filter at all. It's a bit like using a banjo instead of a Les Paul...

I would gladly shell out for a kit of this.

pi has enough grunt to perfectly emulate 20 sid chips /slaps forehead

Perfectly? The SID is semi-analogue isn't it?

What you probably meant is no two SIDs sound the same. Emulated SID will fit in perfectly :) cant claim it sounds wrong when they all sound wrong in some way ;p

An emulated SID won't fit precisely because two emulated SIDs would sound the same :)

You can customize filter curves for all of your emulated SIDs.

Alternatively, it could emulate the entire system. Do you see this as a flaw?

Reply on news.ycombinator.com