This might be the wrong time to start doing that; I highly suspect Pxx for N163 will change in the release version, or some subsequent version. Currently P81 does not correspond to a +/-1 in the pitch register.
It is possible to tune odd wave sizes with Pxx, but I think that would be very impractical. I planned to remove the odd sizes initially because I didn't see the use, but left them in by request. (I'm still not sure why, though.)
rainwarrior: Pitch offsets are shifted to increase the sensitivity, as very small changes are mostly inaudible. I don't plan to change this actually, would that be desirable?
Hmm, when I was trying it out it seemed too sensitive, but I guess it depends on what range you're applying it in, and which sample length you're using makes a difference too. Honestly, I don't know what I'd suggest.
Thought to report that the beta version crashes quite often when I messing with buttons around the right Win button - can't get what exactly happens, but usually I accidentally press something in that keyboard area that causes the Windows 'keyboard blocked' blip, then either crash happens right away or when I close the editor. It never happened with previous versions and does not happen with the latest stable version (so I had to roll back to this one).
Another bug report. If you have a Fxx effect at a second effect column, it won't work if the column is hidden. The problem is also present in previous release version.
By the way, there is a situation that happens quite often if you have patterns of different length (using D00) - sometimes editor skips a half of a normal pattern if you start play on a shortened pattern or something. Can't figure out how this happens exactly, but it was like this for many versions.
By the way, there is a situation that happens quite often if you have patterns of different length (using D00) - sometimes editor skips a half of a normal pattern if you start play on a shortened pattern or something. Can't figure out how this happens exactly, but it was like this for many versions.
It pretty much happens as you described, I think. Basically if you play a pattern that is shortened by Dxx, then manually skip to the next frame during playback, the next pattern is going to be shortened by the same amount, regardless of whether that one has Dxx in it (unless it's shorter than the first frame).
Dxx is pretty broken anyway... It doesn't let you skip to a row further down than the one it's used on. Say for instance you want to use the command on row 0x1F, then you can only jump to row 0x1F or earlier (so writing D30 on row 0x1F will jump to row 0x1F. I told jsr about this bug, but it doesn't seem to be high on his priority list.
The same goes for all effects - anything hidden is ignored. WYSIWYG.
Hm, never noticed that. So it is not a bug, but rather a counterintuitive behaviour for me. I would expect it to work like a code editor - if you fold a section, it still works, just not displayed.
Well, it could be a bug if it's unintended... But again, I believe that's how it's always been, so yeah I guess it's just something you're not used to. Perhaps it'd be possible to alter the source to make the effect columns always work regardless of visibility?
Shiru: I fixed a crash related bug recently so I hope it will work now, and hidden effect columns works as intended. It's supposed to enable/disable columns rather than just hiding.
--
I've uploaded yet another beta. This time with only minor changes, including:
1. I've fixed a crash bug that I think was the source to some bug reports. (I'd like to have this verified.)
2. Some more changes to N163: The suggested resampling method which utilizes the three unused memory bytes/channel has been added, which will cause a bit more aliasing than earlier but this time it's caused by the chip itself. This has been verified on a real chip.
I've also disabled channel multiplexing when using less then seven channels, to avoid the problems caused by the resampler.
