animalstyle: Hey thanks for your advice and for the info on re-capping. Also, a hardware VGM player sounds sweet!

it's pretty easy to "prosound" just the master system chip.  I did it.  maybe have a look at emu-docs.org.

I lied i got 10 HD graphics 5 normal, I'd like to contribute to figuring this out.. if there is any value to the stuff I did

i've got like 15 high definition graphics models.. i did a bunch of research into this using board schematics i found online. theres an amp for the headphone port on the front and theres another dual op amp that amplifies the merged audio channels and something else from some other part of the board, if you pull that 2nd amp completely the genesis wont boot

I assume this dual op amp working on completely different signals is introducing some noise, its an LM317 or something like that, dirty amplifier really not meant for audio from my research.

I havn't done exhaustive testing, but i picked up the L and R audio channel before they were merged and put them through my own LM317 on a separate board and it sounded pretty good, my next tests were gonna involve using a dual op amp meant for audio.

i really dunno what i'm talking about though, learned through experience, sorry if what i'm saying is dumb

I've got a couple of mega drives at home now and maybe one more I could borrow. I would be up for making a sound comparison, but I'd need some help from anyone who's got a different machine than I to make it more complete. I have a borrowed everdrive as well, so I could play some VGMs. Although I would probably construct some sort of "test sound" / pattern. I will most probably do this.

It would be a part of my secret wiki that currently is a wip cataloguing all the different nanoloop versions in a reader friendly way.

Helpful thread thanks, been trying to figure out the best option for this. Think I'll give re-capping a go.

Slight aside, but does anyone know how easy it would be to split off the SN76489 to a separate output? / What exactly I'd have to do?

I think the model 3 has a stereo mod that can be done. It actually does sound fantastic in general too. Not sure why it doesn't get the love it deserves.

The revisions really are a bitch. I would be willing to run the test myself if people were down to send me units to barrow. I have 2 model 2s that were given to me and access to one other hi deff model ones. 

Weeeeee can doooo it if we pool our resources.

yea - the problem is all the damn board revisions.  not all models and boards are equal - makes getting every detail a much more daunting task than the gameboy.  someone with a bunch of money should do it

If someone does do it should they compare square waves like Weixelbaum, though?  would that even make sense for an FM synth?

yea - model 3 isn't on Little Scale's page either - and I think it has a better output by itself than any model 2 (even modded).  Only downside is the mono output.

nah that guy, while the most complete available, still is missing model 3 and different sound mods available.

Idk, this is pretty good.  It's missing board revisions but that's about it.

http://milkcrate.com.au/_other/sega_comparisons/

The genesis really needs a good sound comparison page. the current ones are kinda meehh and very incomplete. would love to see this in full swing.

yes it's a digital chip that is a descendant of the high quality yamaha OPN family.  There is a diference between hearing the actual YM2612 and the ASIC version's approximation in later models.

My friend luis made a hardware VGM player on a breadboard with the isolated chips + arduino at @partty.  That thing sounded amazingly clean.  There was no VDP noise and just about no AC hum.  Electrolytic caps have a life span of about 10 years...  less if they're shitty.

The original filter actually chills out the tin on the high parts of the FM because of the built in design flaws of the DAC in the actual chip (in your wiki article) - i like that better.

again - it's probably not all the chip itself because you find OPN chips in high quality keyboards.  I realize that the wiki is talking about one specific type of noise when channels are blended together.  What I'm talking about is the general hiss, hum and VDP noise funneling through the sound chip.  You basically hear AC line leakage and VDP noise when the processor crunches (i love that sound actually - its nice when it's subtle but annoying when your caps are old).

the first version of TFM did not support panning and when the option was added later... you couldn't hear it inside the tracker.

Seriously - recapping your High Definition Graphics model one sounds really really nice.  The bass sounds better to me too.

This is very helpful information.  Cap values are very helpful.  I have a couple of GenMDM's and was looking to modify a couple of consoles here very soon.

First of all, I think your complaint about the sound being more or less "computery" is a bit odd. What are you expecting exactly? This is a digital synth chip in a video game console, not a state-of-the-art professional analogue synthesizer.

Secondly, I'm not positive about this, but I was under the impression that some of the noise comes from the poor quality DACs inside the YM2612, which you can't change. Actually, it seems this is the case. Have a look on wiki.

I think CCAM actually bypasses the amp and filter section as a better sounding equivalent.

For me, the inherent noise, though it's really a design flaw, is part of the sound character of the synth.

Finally, TFM does supported panning. I've used it for every track I made in it.

I forgot to mention I actually replaced the huge cap in the wall wart too...  which did help some noise go by by.  I also tried to isolate the YM2612's power supply as well but i failed and then ran into some monetary issues - it was hella difficult but i probably went about it the wrong way.  I think the noise is not really being picked up by the YM2612 because those chips are pretty solid and are found in many many expensive keyboards of the 80s - i think its actually the CXA-1034 or the CXA-1145 that's the culprit.  I think that's why tiido basically ripped his out.  I don't mind the sound of that op-amp...  it's dirty in the right frequencies and i also like the capacitor values - it think it filters out some of the high frequency harshness of the tinny sounding FM.  Although - it doesn't help that CXA-1145 is both an audio amp and a video encoder...  doing both of those so close to eachother in an IC could probably lead to some noisey results.