i think i found a bug with teh Cxx command... but it just seen to affect the notsofatso plug-in.
If you USE Cxx command and you delete them, somehow it still stoped the song even if there is no Cxx command anywhere. It just happend when you use Cxx and use notsofatso. There most me something wrong with boths or with famitracker or notsofatso. But i still report it! :D.
Also, the 7xx is broken? or is working in a different way, because some of my old nsf are totally broken now with the 7xx command! ((
I'd like to bring to attention a strange bug that I've been having, that also finds its way into the NSF renders.
I'm writing a module that has a long intro, with a large number of Fxx tempo changes. Because Fxx has been having issues with snail-pace speeds lately, I've re-written some parts of the song into their own windows using double the pattern space and twice the speed I need, cutting to the next patterns with Dxx command.
Ever since I've done this, when it gets to the start of the actual song, the volume envelopes are messed up. They fade out much more quickly, and there is a noticable difference in sound when the pattern starts at that point as opposed to running the whole thing.
I've scoured the effects I call in those patterns multiple times, and haven't found anything that should cause an issue. Perhaps Fxx is causing some error with the envelope settings? I cannot say for sure.
I cannot provide the FTM module right now, but I will get something up here as soon as possible so that others may see what I mean.
That sounds like a pretty weird bug. Never heard of anything like it before... I guess an FTM (and possibly the resulting NSF too) would be necessary to help solve the issue.
Did you check if it happens in 0.3.8 beta 4 as well? Maybe the bug has been fixed already... The chances are slim, but who knows.
My apologies for the double post, but I've tried things out in the beta to no avail. While I was able to de-butcher the pattern jumping that I had to use because Fxx works properly at low tempos again, there's still this issue that I'm having.
Included is a sample with the introduction and the very first part afterwards. For comparison, listen to the pulses and triangle going from the introduction into the first part, and then start it from pattern 03 and listen to it. There is much more brevity, and doesn't sound as full when it plays from the intro.
It sounds like, for some reason, the pulse cycle envelopes on Pulse 1 in Frame 03 are being overridden by something else. But this only happens when Frame 02 is played before then, in its entirety, or at least more than the last two bars. (the last 32 lines) But I can't tell what's causing it.
Upon closer inspection, it almost sounds like the last tick of the duty macro (where it goes to duty 0) is getting cut off and not played when the full intro is played. At the BPM the main section runs at, it actually should get cut off a few times, but it seems that it gets cut off each time when the intro is played.
Proposed solution: Overclock your NSF. That way the macro is sure to run. It may make some other instruments sound a bit off, so you'd need to tweak things a bit, but it's looking like the format is getting in the way.
What happens is your tune has a tempo that causes irregularities in the NES's audio updates. Basically, at the tempo it's going at in pattern 03, each row alternates between during 3 frames, then 2 frames. However, your macros last 3 frames, so every other row the envelopes get cut off.
Like icesoldier said, overclocking your NSF would be a potential fix, but it's not recommended if you want to export to NSF eventually (not all players support overclocking). Another solution would be to just slow your tune down, but that might not go well with your artistic view of what the tune should sound like, and I can perfectly understand... Or, you could leave it like that since I think it actually sounds quite well the way it is.
Oh and personal recommendation: leave the tempo at 150 and only mess with Fxx to change tempo. For instance, alternating between F03 and F02 lets you achieve the same speed (@ tempo 150) as using FF0, with the benefit of letting you know where and when the "frame skip" in the NES audio engine updates takes place.
It's a personal thing, but I think it helps workflow immensely, so I thought I'd share the tip!
Lastly, I really hope you'll be sharing the tune with us when it's complete. It sounds awesome already, and I'd really love to see where you can take it!
I'm listening carefully, and it really doesn't sound like a problem with the duty cycle. I swear, it's got to be something with the volume . . . The channel volume even seems to behave differently.
I think that it's also a problem with the macro, but it's strange. It seems to play the full extent of the macro when I start the tracker from there, but if it plays by itself it doesn't play the full extent.
This whole problem started when I changed the way that the tempo rises in the introduction. I had it at one point where there wasn't this kind of glitch present, and I was running it at the same tempo.
With some further experimentation, I have actually determined that the same thing happens with the triangle channel, actually. The note duration with the skips isn't as long. It seems like something in the introduction is causing a minor frame skip, I don't know how or where.
Also, I'll be sure to share at least the NSF here when I'm done. Ideally I'd like to put this in an album of sorts, maybe try to sell it for cheap? I don't know.
((
