MSX1 vs C64

look, a port from SID to AY
https://www.youtube.com/watch?v=OwdAyVqcUbU&t=1m10s

Assuming I've understood the question (and the VIC-II documentation I've just hastily parsed): the C64's sprites are bigger, and it can display more of them per line. You need to accept halving the resolution if you want them to be more colourful though. However it achieves these differences from the TMS in two ways:

• it elevates sprite details other than their pixels to registers, which effects a limit in the quantity of them that can be handled automatically: eight in total, rather than thirty-two; and
• to deal with the amount of memory bandwidth necessary for eight wider sprites by repeating a trick it uses for getting 40 columns of colourful text per line while sharing its memory footprint with the CPU: stealing cycles. The more you have the VIC doing, the less processing time you have available. And it's very slightly worse than optimal because it uses the 6502's RDY signal for that which means losing an additional 1/32nd of the line for 6502 reasons*.

The only amelioration is that the VIC provides per-line interrupts, so you can jump in and tweak those sprite registers to get more than the normal eight on a display, even on the original machine. You need an MSX2 for that level of interrupt timing. An MSX1 would have to busy wait, in which case it's questionable that you're gaining anything a hardware sprite, but 32 is probably plenty so it probably doesn't actually need to think about it at all.

Also potentially of interest: playing raster tricks can put hardware sprites into the border area.

EDIT: so, that's why I spoke well of the SID, but declined to mention the VIC-II or the 6502. The timing interactions between the two somewhat starve the machine, I think. For my money, the Atari 400/800 have the most interesting 8-bit video hardware, being a blitterless proto-Amiga.

* the 6502's RDY signal tells it to stop working the next time it tries to read from the bus. The 6502 is permitted to write regardless of the RDY signal. Two cycles is the maximum number of contiguous writes on a 6502, so you have to signal RDY two cycles before you want to grab the bus to be completely sure. If it would only have read during those two cycles then, tough, you've lost time. And in a C64, a line is only 64 cycles because 6502s tend to be clocked lower in general, and the C64 is at the low end of the range.

Admittedly if you discount all "sound effects achieved by manipulation of SID registers" then, no, there aren't any attempts at instruments. Or any audio at all, in any Commodore game.

As I said before, I think the SCC and OPL both beat the SID, but that the SID is probably better than a plain AY. I don't think that introducing artificial tests is helpful in evaluation.

The SID is on par with scc or opll imho, fewer channels, but compensated by its other features. I'd say depending on situation it can be better than scc or opll. And it runs circles around the AY, no doubt about it.

Yeah that was what I meant, I've seen that demo before and I thought "WTF???" . Its the arpeggio's that give it away, otherwise I would have thought that first song was some indie/underground dance track made with VST's/synths etc.

Also amazing Hyperbased cover at the end.