I always love hearing the YM2608 / OPNA / PC88 original versions of Koshirosan's [adapted for Genesis] pieces! My understanding is that Koshirosan never actually touched a YM2612 / SN76489, and did all the work on the OPNA, which was converted for him by [data missing] to the Genesis sound hardware. I love the "different but the same" nature of the PC88 renditions. Cleaner FM, -MUCH- cleaner and less smashed PCM, and slightly different PSG, but still the same basic sound in the final Genesis output, just more lo-fi. One thing that I think is extra fun about this comparison is that even though it's probably the part of the sound where you actually hear the smallest amount of real-world difference, I imagine the part of the conversion that was the most difficult and time-intensive is the PSG. I don't think the Texas Instruments SN76489 PSG can simply accept the instructions being fed to the YM2608's built-in YM2149 PSG (which is basically just a glorified rebranded AY-3-8910), and so I'd suspect that the PSG portions of these pieces would have to be reprogrammed from the ground up note by note. Not to say that the YM2149 isn't clearly superior to the SN76489, just that for these tracks, it was only used in ways that the lesser chip could match. So any differences between the simple square waves between the two are exceedingly subtle, and doubly so in the mix with the FM and PCM! I would imagine that that PSG portion would be a lot more work than the PCM conversion, which I always presumed was "let's make a hundred tiny recordings of the PC88 PCM, smash them down to where they'd fit on the YM2612's DAC (and within ROM considerations) and just use that. And the FM part, I'd imagine, would be the easiest part of the conversion of all since, so far as I know, the two chips are 100% identical "on the back end" of the FM process (in terms of operator / algorithm/ envelope capabilities etc), and that the coding / those instructions would probably just port straight over with potentially zero adjustments at all. To be sure, the YM2612's FM is lower-fidelity, but so far as I know, the difference is purely "on the front end" - in terms of "output fidelity", kinda like playing the same keyboard through a really expensive amp and a cheaper amp - even though I know one uses an adder and the other doesn't (kinda like the difference between a mixing board, and the distributor cap on an old car). The YM2608's FM is much cleaner, and without artifacts such as "the ladder effect", but in terms of their raw "sine-wrangling" capabilities per se, I am under the impression that there is literally ZERO difference between them whatsoever, which I'd think would make that part of the conversion a simple matter of copy/paste. Well, either way - I LOVE to hear the PC88 originals of Genesis classics. Funnily enough, I actually like the female vocalization "Uhh!" better in lo-fi on the Genesis. It's cleaner here, but sounds somehow cheesier, and most remarkably of all, somehow "smaller" here.
![YoungBoy Never Broke Again - Missing Everything [Official Video]](http://i.ytimg.com/vi/dTYeMZGzOzw/mqdefault.jpg)
Nice oscilloscope view of the original PC88 version of the song!😄🤩😎👌
I always love hearing the YM2608 / OPNA / PC88 original versions of Koshirosan's [adapted for Genesis] pieces!
My understanding is that Koshirosan never actually touched a YM2612 / SN76489, and did all the work on the OPNA, which was converted for him by [data missing] to the Genesis sound hardware. I love the "different but the same" nature of the PC88 renditions. Cleaner FM, -MUCH- cleaner and less smashed PCM, and slightly different PSG, but still the same basic sound in the final Genesis output, just more lo-fi.
One thing that I think is extra fun about this comparison is that even though it's probably the part of the sound where you actually hear the smallest amount of real-world difference, I imagine the part of the conversion that was the most difficult and time-intensive is the PSG. I don't think the Texas Instruments SN76489 PSG can simply accept the instructions being fed to the YM2608's built-in YM2149 PSG (which is basically just a glorified rebranded AY-3-8910), and so I'd suspect that the PSG portions of these pieces would have to be reprogrammed from the ground up note by note. Not to say that the YM2149 isn't clearly superior to the SN76489, just that for these tracks, it was only used in ways that the lesser chip could match. So any differences between the simple square waves between the two are exceedingly subtle, and doubly so in the mix with the FM and PCM!
I would imagine that that PSG portion would be a lot more work than the PCM conversion, which I always presumed was "let's make a hundred tiny recordings of the PC88 PCM, smash them down to where they'd fit on the YM2612's DAC (and within ROM considerations) and just use that.
And the FM part, I'd imagine, would be the easiest part of the conversion of all since, so far as I know, the two chips are 100% identical "on the back end" of the FM process (in terms of operator / algorithm/ envelope capabilities etc), and that the coding / those instructions would probably just port straight over with potentially zero adjustments at all. To be sure, the YM2612's FM is lower-fidelity, but so far as I know, the difference is purely "on the front end" - in terms of "output fidelity", kinda like playing the same keyboard through a really expensive amp and a cheaper amp - even though I know one uses an adder and the other doesn't (kinda like the difference between a mixing board, and the distributor cap on an old car). The YM2608's FM is much cleaner, and without artifacts such as "the ladder effect", but in terms of their raw "sine-wrangling" capabilities per se, I am under the impression that there is literally ZERO difference between them whatsoever, which I'd think would make that part of the conversion a simple matter of copy/paste.
Well, either way - I LOVE to hear the PC88 originals of Genesis classics. Funnily enough, I actually like the female vocalization "Uhh!" better in lo-fi on the Genesis. It's cleaner here, but sounds somehow cheesier, and most remarkably of all, somehow "smaller" here.