I released my first commercial game album, Music from 1001 Spikes a couple weeks ago.
I'm not going to release it in NSF format as way too many .ftm files were used for that to be even remotely possible, plus several songs use different expansions (FDS, VRC6, MMC5, 2a03-only).
It is possible to use multiple expansions in a single .NSF. The only thing to be careful of is not to write to a sound register which is also FDS RAM if you are not trying to change the program data at that address too.
There appears to be less than 255 songs so you might put it in a NSF as long as it also fits in less than 255 ROM banks.
I am not interested in downloading rendered audio of 69 tracks, whether or not it is for pay, although perhaps a physical CD might do.
RushJet1 wrote:
several songs use different expansions (FDS, VRC6, MMC5, 2a03-only).
Allow me to quote the Team Fortress 2 Spy. Nothing is more nauseating than a game whose music is inconsistent as all hell. I thought people would never tire of dickriding the VRC6, let alone mashing every single expansion audio together.
Let's map a MIDI output and an MP3 player into $47xx while you're at it
Moero Pro Yakyuu came dangerously close to the latter with its ADPCM voice playback chip.
I'll agree that consistency is the mark of professional game design, but I don't think it's worthwhile to trash talk the use of different expansion audio in a soundtrack considering 1001 Spikes isn't specifically trying to be an NES game. If memory serves right, it was just a pixel art platformer that didn't adhere to any specific graphical limitations other than pixelation.
Besides, TurboGrafx-16 sound as I understand it is pretty much a superset of MMC5 sound, VRC6 sound, and FDS sound, so in a way it is consistent. And if "pixelated" is the only limit, you could even bring in VRC7 and call it Sega game music or bring in a freaking 8-voice sampler and call it Super NES music.
You could also even call it an indie game.
You might be able to do a reasonable recreation of all of those expansions using N163, which would make a single NSF / cart implementation more feasible.
(...not that it'd be worth spending the time on, esp. when it's so easy to listen to already.)
It is possible to do a single NSF with multiple expansion audio like I have mentioned above; you only have to be aware of conflicts with FDS RAM. In case the music will not fit directly, you can choose to compress it, and to decompress the music at runtime, during the INIT routine, into RAM (which, due to FDS, you should have enough of, although you should reserve some banks to be used only for RAM in order that conflicts are not caused due to the implementation of the NSF player).
tepples wrote:
Let's map a MIDI output and an MP3 player into $47xx while you're at it
Moero Pro Yakyuu came dangerously close to the latter with its ADPCM voice playback chip.
MIDI output is real though, it's done over joypad #1.
To interface with a piano, yes. I imagine Tepples was referring to doing something like including an on-cart MT-32 as expansion audio, however.