Skip to content

Future VGM expansion: YM2128(0) ?

aka the DX7; wishful thinking

Technical discussion about the VGM format, and all the software you need to handle VGM files.

Moderator: Staff

  • User avatar
  • neologix Offline
  • Posts: 211
  • Joined: 2012-04-22, 4:03:45
  • Location: New York, NY, USA

Future VGM expansion: YM2128(0) ?

Post by neologix »

Ok, so file this under "wishful thinking," but being as big a fan of the sound of the Yamaha DX7 as I am I would personally love it if a future expansion to VGM incorporated the YM21280 / YM2128. AFAIK there aren't any available datasheets or official specs floating around (this page contains the most public official info available for the DX7), but we could start by using hexter / sixport as a basis if we do attempt this monumental feat.

Regarding the sequence data, I'm unsure of what format already exists for the DX7 exclusively short of MIDI and some specialized SysEx messages, but it could be very possible that we could take this opportunity to either encapsulate a slightly extended form of MIDI (ie, MIDI w/stuff to specifically handle DX7 messages; could theoretically be done w/straight MIDI and the right set of CCs and SysEx messages), use XMI, or to create our own sequence format that this and future not-yet-implemented not-quite-videogame chips could use. I personally vote MIDI or XMI.

Once again, I totally consider this "wishful thinking" and am fully aware the YM21280 isn't a videogame-specific chip. It is, however, a Yamaha FM chip that not only sounds awesome but has a huge following. I'm also fully aware this would likely be the largest undertaking the VGM format would have to date if followed thru, and that such a change would totally need to be vetted and such. I simply ask that this not be immediately shot down on the singular basis that "it's not a videogame chip," even tho for a format such as VGM (which is predicated entirely upon being the Video Game Music container format) it is indeed a very legit reason; also, I would totally understand if this particular chip was viewed as feature creep/bloat. The YM21280 could be the first of many not-quite-videogame chip formats the VGM format can encompass.

OR it could be the first of many chips in an offshoot container format related to VGM. I'll have to think abt this last bit...
  • User avatar
  • tails_ Offline
  • Posts: 170
  • Joined: 2012-03-18, 16:59:23
  • Location: Tomsk, Russia
  • Contact:

Post by tails_ »

I still don't see much sense in this, if you're about writing music then VGM definitely not good solution. Also DX7 is based around MIDI so importing patches will be another headache since they're in SysEx
  • User avatar
  • ValleyBell Offline
  • Posts: 4768
  • Joined: 2011-12-01, 20:20:07
  • Location: Germany

Post by ValleyBell »

If the DX7 uses MIDI commands for communication, VGM support doesn't make sense.

For sequence data, I would prefer MID. I don't like XMI for the lack of tempo changes.
If you want to make your own sequence format, you can ask me about the format of Lemmings 3D. It uses a MIDI-like format that's optimized for tracks without accords.
  • User avatar
  • neologix Offline
  • Posts: 211
  • Joined: 2012-04-22, 4:03:45
  • Location: New York, NY, USA

Post by neologix »

ValleyBell wrote:...Lemmings 3D. It uses a MIDI-like format that's optimized for tracks without accords.
That amt of "optimization" sounds too close to "GM Lite" for my personal comfort; I've always found GM Lite to be too restrictive, and Android-based smartphones are the only phones I've ever encountered that even come close to fully supporting it.

So to summarize arguments so far:

Me: Let's emulate the DX7 in VGM format!
tails_: SysEx-to-VGM patch = headache.
VB: DX7 already uses MIDI, and XMI is restrictive.
Others: TBA

At least the convo isn't immediately shut down due to potential stubbornness and obstinacy :D .
  • User avatar
  • ValleyBell Offline
  • Posts: 4768
  • Joined: 2011-12-01, 20:20:07
  • Location: Germany

Post by ValleyBell »

You got me wrong.
The format itself is fully capable of playing doing GS/XG. (Aside from the lack of SysEx commands, but there are lots of unused commands.)
The optimization is, that a byte between 00 and 7F defines a note, followed by a volume. 80-FF are control bytes. (like 'Turn off last note', 'Select channel', 'Loop Start/End', etc)

i.e. you get very small files, if there are no accords in the track. This is the case, if you want to control sound chips like the OPL2 or the YM2612. And I assume this is also the case for the DX7.


That said, I have nothing against non-VGM chips in the VGM format. It just doesn't make sense for me to encapsulate MIDI commands in VGM files. RIFF MIDIs didn't make much sense either.
  • User avatar
  • neologix Offline
  • Posts: 211
  • Joined: 2012-04-22, 4:03:45
  • Location: New York, NY, USA

Post by neologix »

ValleyBell wrote:You got me wrong.
The format itself is fully capable of playing doing GS/XG. (Aside from the lack of SysEx commands, but there are lots of unused commands.)
The optimization is, that a byte between 00 and 7F defines a note, followed by a volume. 80-FF are control bytes. (like 'Turn off last note', 'Select channel', 'Loop Start/End', etc)

i.e. you get very small files, if there are no accords in the track. This is the case, if you want to control sound chips like the OPL2 or the YM2612. And I assume this is also the case for the DX7.
Ok, I think I get what you mean on the optimization bit. Straight-up General MIDI has something very similar w/its "running status" ability, and that for the most part is also affected by the existence of chords in a track. I'd still personally sacrifice the optimization provided by running status in favor of the clarity of the actual instructions (also one of the reasons I write "note off" messages instead of "note on velocity=0").
ValleyBell wrote:RIFF MIDIs didn't make much sense either.
The only thing I really like abt RIFF MIDI over plain SMF is the ability to embed a sound set w/its sequence (most commonly done w/DLS apparently). This didn't look to be used all that often in practice, however.
  • User avatar
  • Sik Offline
  • Not a musician
    Not a musician
  • Posts: 75
  • Joined: 2011-12-12, 12:43:15

Post by Sik »

neologix wrote:
ValleyBell wrote:RIFF MIDIs didn't make much sense either.
The only thing I really like abt RIFF MIDI over plain SMF is the ability to embed a sound set w/its sequence (most commonly done w/DLS apparently). This didn't look to be used all that often in practice, however.
The problem was that practically all players added "support" for RIFF MIDI by... completely ignoring all the new features. So for all we know they could have just been standard MIDIs, and thereby everybody stuck with those. I think I only had one program that bothered with properly implementing RIFF MIDI, and it involved using a separate codec.
Sik is pronounced like "seek", not like "sick".
http://www.mdscene.net/
Post Reply