When trying Colonization I noticed that the first intro music (Bonny Morn) at some points of this there are always some cracking noise. However none of the other music of the game seem to do this. I therefore tried some other speakers but same thing. Wanting to confirm it was not an issue with the A1200 I tried the game on an A500 and same thing happened. I also tried the game on FS-UAE and again in the intro music I can hear such cracking noise but the other music of the game looks to be ok.
I don't know the exact theory behind. It seems to happen at the transition between two sounds (or from silence to a sample, or from a sample to silence, sometimes even when looping) when the output level moves abruptly.
Amiga doesn't have "volume ramping". Volume ramping is a kind of blending of the frequencies. So instead of cutting a tone off completely (resulting it a click), it fades down really quick. Amiga doesn't have this. So if you stop a note, you'll probably hear a click.
@recedent, post #2
@Umpal, post #1
@Umpal, post #1
@Jazzcat, post #7
@Umpal, post #11
@Umpal, post #11
@Jazzcat, post #18
@Jazzcat, post #20
These clicks are clearly happening because - even though volume is at 0 on the previous sample, volume it turned back up on this sample just before playing the next one, causing the click. That's a bug. The previous sample should have been stopped before playing the next one.<-- that could be a plausible explanation
@Umpal, post #21
@Adinfo, post #24
Czy rozwiązaniem mogło by być dodanie dodatkowych (zbliżonych wartością) bajtów w kolejnym samplu? (taki trik maskujący)
@Adinfo, post #24
- does pt/amiga do something weird like loop the first byte of a non-looping sample when it ends? So you'd want to ensure the first byte is 0
- every sample loops. Non-looping samples loop the first two bytes after finishing. They are zero and this is not part of his problem so you'd want to ensure the first 2 bytes are 0 then :) And they are
- ok
- this is why replen is 2 on non-looping samples btw. Bet you wondered why replen aka. loop length
when you set data and length on Paula, it will just play that chunk over and over, and there's no way to know exactly when the data ended, so what they do is to first set data and length like normal, and then when the DMA had latched, they instantly set data to sampleStart and length to 2, and that will take effect after Paula has done one cycle of the sample (played through)
so non-looping samples on Paula is very stupid
the same goes for loops. After one cycle, they set data to loopStart and len to loopLength
or rather, instantly, but Paula re-reads the registers after a cycle
(and the reason replen is 2 and not 1 is because Paula works on words, not bytes. Length = words)
which also means that 2 is the lowest amount it can do.
(which internally in the PT replayer is half of that value, not the value you see on screen)
@mschulz, post #12
@Don_Adan, post #27
No nie wiem, wszystkie sample Protrackera maja 2 lub 4 puste bajty na poczatku, tzw. null.
@Umpal, post #26
@mschulz, post #28