It's true; I don't support the modulation table yet, so it will only accurately do low frequency modulation. I was planning to wait for 0.3.7b source and and at it then because the Hxx effect makes it a lot easier to cover all the parameters.
(Also, tell me what NSF that you noticed it was off on.)
Inb4 lazy covers and wannabes posting modified stuff calling it their work.
Have you tried this thing? Imported NSFs are pretty much impossible to work with unless you're a pretty advanced FT users, and I doubt advanced FT users are the kind of people who would blatantly plagiarise other NSFs anyway. Very unlikely IMO.
In fact jrlepage, probably the main thing anyone using this tool will use it for is acquiring instrument data and making songs in similar style. It wouldn't make sense for some to plagiarize songs in a small community where everyone they show their work off to will recognize it and ostracize them.
Inb4 lazy covers and wannabes posting modified stuff calling it their work.
And "I TAKE REQUESTS!!11"? I suppose it's somewhere in between posting GXSCC as '8-bit remix' and tossing soundfonts over other people's MIDIs and call them your work on the 'please don't do this' scale.
Nevertheless, that's great stuff rainwarrior. I'm guessing emulating the sound of games like Megaman and Castlevania is now more accessible to everyone. It's also nice to check out an .ftm representation of some Tim Follin.
now if only someone would take a .gbs input plugin and do this for that. Map squares 1:1, noise would be approximate, and gb wave channel could either go tri or fds >:D
now if only someone would take a .gbs input plugin and do this for that. Map squares 1:1, noise would be approximate, and gb wave channel could either go tri or fds >:D
rainwarrior; you're a true genius. I can finally analyze the music from the Hero Quest prototype's NSF!
I'm surprised at how accurate the import was, considering that this track was only pure 2A03. I'll try something with MMC5 & VRC6.
Xyz_39808 wrote:
You said a month.
LIES!
(why can't all lies be this sweet)
It could obviously take months before any bugs can be found & any further improvements can be made (but judging by the speed & skill of rainwarrior's coding, I may be mistaken; in any case, the ETA is up to him).
_______________________
Technology: the one thing that's hated & cursed at by all engineers, technologists, scientists & technicians!
Bumparoo for a good cause; bug found when importing MMC5 tracks.
See the attached files; proper_mmc5_track.ftm & proper_mmc5_track.nsf are the source files. The MMC5 squares are set to use V00 for a duty cycle of 12.5%.
Passing the NSF through the importer does something really interesting; it increased the MMC5 squares' duty to 50%, replacing the V00 effect with a V02 effect. Other than this, the entire song is 100% accurate. You'll see this in importer_error_mmc5.ftm.
I've also noticed that an extra (unneeded) instrument is created for the MMC5 squares, which has the exact same settings as the other instrument used for the default 2A03 channels. In both cases where DPCM is concerned, the DPCM samples play properly, yet there's no explicit note assignment within any of the instruments.
_______________________
Technology: the one thing that's hated & cursed at by all engineers, technologists, scientists & technicians!
I too have found a few bugs. The most notable one is with the VRC7. Pitch detection seems to be a bit off. I'll attach an NSF and the corresponding FTM so you can take a look at it.
) 