#1092 closed defect (fixed)
IMUSE: Pitchbend isn't reset when changing tracks
| Reported by: | eriktorbjorn | Owned by: | SF/jamieson630 |
|---|---|---|---|
| Priority: | normal | Component: | Engine: SCUMM |
| Version: | Keywords: | ||
| Cc: | Game: |
Description
This bug report will describe a specific case in Fate of Atlantis, but it probably applies to iMUSE in general.
While walking around the outer circle of Atlantis, I suddenly noticed that one of the parts was way out of tune. It turned out that it had a pretty large _pitchbend value. If I set it to 0 it sounded fine again.
I believe I've figured out what must have happend. I even managed to reproduce it once, but it's tricky:
There seem to be three different tracks, numbered 0, 1 and 2, which are picked randomly when entering a room. Changing tracks is done with the Player::scan() function.
Track 1 starts using pitchbend after about 35-40 seconds. The part that is being bent is very audible in track 0. I entered a room, getting track 0, left the room, and re-entered it again after not quite 37 seconds, this time getting track 0. By timing it just right - this took several attempts - I managed to get the change to happen in mid-pitchbend.
If the track had been allowed to continue on its own, it would have reset the pitchbend to 0 of course. But since the track changed, this never happened.
So, to summarize, there is a problem, I know how to reproduce it (though I may not be able to do so at will), and I have a good guess on how to fix it. But I'd prefer if someone who actually understands the iMUSE code would do it. Particularly since there may be related bugs there as well.
(I can provide a savegame, I just forgot to bring one.)
Ticket imported from: #780924. Ticket imported from: bugs/1092.
Change History (4)
comment:1 by , 21 years ago
| Owner: | set to |
|---|
comment:2 by , 21 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
comment:3 by , 21 years ago
comment:4 by , 5 years ago
| Component: | --Unset-- → Engine: SCUMM |
|---|
Well, despite knowing exactly where the pitchbend occurs, I had to run the music at 50% speed to reproduce the problem. Fix is in CVS. Thanks, eriktorbjorn -- I'm impressed that you caught this!