Beiträge von 0xdeadbeef im Thema „Level of SID Implementation?“

    Playing samples on the 6581 was done by exploiting a bug in the hardware which caused a distinct popping sound when playing with the volume bit of the noise channel. So anything that fixes that bug including Commodore's own 8580 will produce a more muted output on those test cases.

    I am very well aware of this. When I first got a short board with an 8580, the first thing I tried was checking the sample in the intro of "Ghostbusters" and expected it to be inaudible. Well, it wasn't. Actually, I couldn't really tell a difference from a 6581. As far as I recall, this also sounds OK on an UltiSID in both settings. Probably a somewhat different approach to play the samples compared to other games and BreadAmp.

    Anyway, try BreadAmp with an UltiSID, both in the 6581 setting and judge for yourself. IMHO this is not just a volume thing. It sounds terrible and nothing like a real 6581 at all. Surprisingly, it sounds much better using the samples meant for the 8550 which makes you wonder if there is really just a 8580 emulation anyway with slightly different filter settings for the 6581 "emulation".

    Just for the record: the UltiSID implementation in the Ultimate 64 is not bad for normal games, but far from perfect. What seems to be a worst case scenarios is playing samples in BreadAmp. When using the samples for a 6581 and configuring the UltiSID in 6581 mode, the resulting sound quality is somewhat terrible and not even close to that of a real 6581. An ArmSID does a much better job here. Admittedly, this is a very specific scenario but it shows that some details of the 6581 are emulated much worse by the UltiSid than e.g. by an ArmSid. So using the current state of the UltiSID as a general benchmark for SID emulation seems like a bad idea.