Deprecated: mysql_connect(): The mysql extension is deprecated and will be removed in the future: use mysqli or PDO instead in /storage/content/49/145849/famitracker.com/public_html/forum/classes/dbHandler.php on line 29
FamiTracker
After a [url=http://nesdev.parodius.com/bbs/viewtopic.php?t=5473]discussion over at NESDev about the peculiar tuning of the DPCM samplerates, I decided to make a short test tune that tries to be as in-tune as possible with them.
Strangely, it looks like the DPCM samplerates were chosen to make a major scale (+ some extra random notes) with a looped 16-byte DPCM sample; but because the DPCM actually works with 17-byte samples (or any multiple of 16n+1), they are all off by 16/17 (an out of tune semitone).
So... I wrote this piece in FamiTracker, then after exporting to NSF I replaced its pitch table with a new table tuned to B-flat = 440hz * 16 / 17, instead of A = 440hz. You can [url=http://rainwarrior.ca/projects/nes/lately_tuned.py]download the python script that will generate this pitch tables, if you want to see how I did it.
The FTM version does not have the replaced pitch tables, obviously, since that's not a FamiTracker feature, but you'll see that the difference is fairly subtle anyway. The NES is not a precision-tuning instrument even at the best of times.
TLDR? Just listen to the NSF, not the FTM. This is an example of what the DPCM sounds like when it's in-tune as possible (i.e. not great).
The pitch table isn't really that different; for most of it it's only around a P7F. The tuning difference is almost too subtle for the NES.
Anyhow, the tuning is to the "ideal" intended frequencies that the DPCM pitch set is tuned to, the square channel that is doubling it gets much closer. (The DPCM's internal divider only has 8 bits of accuracy; it's really not very good. Better than TIA I guess.)
