Sega Genesis Audio Quality - The Problem

[Epicenter]

Your clips are awful; how could they discern anything from them? You should be analyzing the clock into the YM2612 before jumping to any conclusions. Because right now, there is no evidence the clock is faulty. It would cause the pitch to shift, and that isn't evidenced in any of your clips.

[evildragon]

these clips are fine enough to show the difference.. this is the one the Roland guy heard..

http://blackevilweredragon.spymac.com/before4.wav
http://blackevilweredragon.spymac.com/after4.wav

[Epicenter]

It sounds to me like you manually altered the files with a sound editing application on your PC, or you raised the clock to the YM2612 and altered the pitch (it sounds slightly higher). I find it highly unlikely you have an oscillator rated for exactly 7.67 MHz sitting around-- meaning you might've used an 8 MHz generator, which would increase pitch. As for any 'quality' improvements? I hear none.

[evildragon]

the files weren't altered, hence i left them as the raw WAVE files they were recorded in, so you could test the files any way you wanted...

you seriously don't hear the low notes being improved? perhaps you don't have good speakers/hearing, like i do then... (being my father used to be in a band, and i deal with audio recording equipment, i hear even the tiniest things) cause others here on this forum hear the difference...

[Epicenter]

If they've heard a difference, it was probably imagined because you told them it was improved, and to listen for it. This is not the point under discussion, however. The point under discussion is clock clarity to the YM2612 and the effect it has-- you've not proven there is any issue with your clock, as you haven't analyzed it.

And you NEVER stated the frequency of your crystal. What is it? We don't even know if your crystal is true to its specifications, as you've not checked it to any higher standard.

My main concern is, you state your findings as fact, when your method didn't even border on scientific in the least. This is severely misleading to anyone without a thorough technical understanding of how the hardware works.

[TheGZeus]

Originally Posted by Epicenter

The very idea behind this 'mod' is preposterous for a number of reasons.

#1, When a clock signal passes from a generator to an IC, then out and to another IC, it is not damaged in the process unless it encounters enormous resistance. A clock cannot be 'sucked up'-- at least use technical terminology. The signal going out of the first IC to the second should not be significantly changed. There is more current draw, but there is not more voltage draw, so the amplitude will be unaffected. Unless the signal line encounters significant high frequency noise it should have its frequency unchanged as well.

#2, The statements in the above paragraph are barely necessary, because in the MegaDrive/Genesis, VCLK is split down numerous paths before it is sent to the 68000 processor, YM2612 audio synth and VDP. The 68000 has no output clocks, only one input. There is NO opportunity for the 68000 to 'suck up the clock', whatever that even was supposed to mean.

Your recordings are both terrible, and each exhibits flaws not present in the other, all of which sound like faults of your recorder, cable, connections, recording medium, or other factors. The audio sounds heavily overamplified, even on these speakers, which are unable to produce any sort of high-volume audio, these samples STILL sound appallingly loud. Neither one sounds discernably 'better' than the other-- they're both bad.

#3, Another technical inaccuracy on your part-- the VDP does not produce a clock to feed to the 68000/VDP. The ~7.67 MHz clock (VCLK) is generated in the GLU chip by division of the input MCLK (which is around 53 or 58 Mhz depending on the region of the system, and generated from a crystal oscillator). No generation or division of this clock occurs in the VDP.

#4, The quality of the input clock to the YM2612 would have no impact on the achievable frequency range of its output. All that its amplitude matters is that it be high enough voltage to trigger the transistors and gates in the IC (probably between 3vdc-5vdc, ideally 5vdc). If it is too low-amplitude, the chip won't trigger and it didn't perceive reception of a clock pulse. If it is too high, the chip is more likely to just be damaged. If its frequency is incorrect, the pitch of the output will rise or drop.

I think you need to go back to the drawing board on this one.

[TheGZeus]

He's going to claim he's autistic again in a moment.
It's a pose.
These things are EASLY for people like us(autistic) to explain. It's ABSTRACTS that are difficult.

Voltsxamps=watts is a very concrete concept and easy to grasp.

The way he's described various other things here shows he does understand the concept and can express it.

I could link to some GameSX threads where he's demonstrated a total lack of understanding of the most basic concepts of electronics.


I recommend we all abandon this thread and let him make up some other random mod that's irreversable and pointless THEN get himself banned.

[Robivy64]

Whoa! Devon i'm amazed. I swear you have a radar for any Genesis hardware mod that is inaccurate!

[Melf]

That's two threads I've had to lock because you guys can't hold a mature discussion. I don't want to have to lock a third.