WHDLoad MantisBT - Pushover
View Issue Details
0004965Pushover[All Projects] Generalpublic2021-01-22 21:582023-01-21 17:40
Assigned ToStingRay 
PlatformOSOS Version
Product Version 
Target VersionFixed in Version 
ChipMem2 MB
FastMem10 MB
WorkbenchOS 3.1
KickROM40 - Kick 3.1
Summary0004965: Sound glitches with fast CPUs
DescriptionWhen playing Pushover on an accelerated Amiga, the music doesn't sound correct. Notes (and sound effects) are randomly skipped, and there's some occasional noise. The faster the CPU, the worse it sounds. Lowering the clock speed or disabling the accelerator card altogether seems to fix (or at least minimze) the problem.
Steps To ReproduceRunning the game on an accelerated Amiga (tested with 68030)
TagsNo tags attached.
Attached Files? PushOver.slave (6,192) 2022-12-29 18:46

2021-01-30 10:53   
I suspect you are using a (very) old version of the patch. What's the patch version you've been trying?
2021-01-30 12:01   
I'm using Version 1.23 (27.11.2017) from the WHDLoad website. Just to confirm, I did a fresh install today and ran a couple of tests.

First I tested in FS-UAE and wasn't able to reproduce the sound issues in the emulator at all. No matter how fast I ran it, the sound was fine, so this seems to be an issue that happens on real hardware only.

Then I've tested on an A1200 with an ACA1233n, which has the advantage that I can run it at different clock speeds:
- 40 MHz gave me very notable sound glitches, some missing sound effects (especially in in the intro), dropouts in the music etc.
- 26 MHz improved it a little, but it was still notable.
- At even lower clock speeds it was still happening, but considerably better. Sound effects seemed to be mostly fine and I still got occasional missing notes in the music.

So, it seems to happen only on real hardware, and it seems to get worse with higher clock speeds. The game is running fine otherwise, no timing issues or anything.
2021-01-30 23:48   
with winuae we can probably reproduce it with chipset_hacks=0x8 in .ini file

this dev option trashes the sound if timing isn't respected between writing to dmacon (to shut off the channel(s)) and new audio data fed to audx registers.
2021-12-11 22:46   
Recently I've tested a TerribleFire 1260 card when I remembered this bug report. So I tried to run Pushover, and got the same audio problems with that card. So it's not limited to the ACA card, and with the 060 it's even worse, there's not much left of the music. Which seems to confirm that the sound glitches get worse the faster the CPU is.
Then I also tested with WinUAE, and while the sound is fine when I run it normally, with "chipset_hacks=0x8" it causes very similar audio dropouts.
2022-12-29 17:52   
(Last edited: 2022-12-29 18:01)
Confirmed, I had already fixed a few CPU dependent delay loops but there are more, probably "hidden" in the external files the game loads. Currently trying to find the remaining delay loops that need to be fixed.

2022-12-29 18:47   
I have added DMA waits in the MED replayer, please try attached slave, does it fix the sound issues?
2023-01-11 22:25   
Thanks a lot! I'm not the original reporter but for me (A4000, 68060, OS 3.2.1) this fixes it.
2023-01-12 20:42   
Thanks for the feedback, RobDangerous. I'll release an updated package of the patch then.
2023-01-21 17:40   
The updated version of the patch is now available.

Issue History
2021-01-22 21:58probdenisNew Issue
2021-01-23 18:18JOTDAssigned To => StingRay
2021-01-23 18:18JOTDStatusnew => assigned
2021-01-30 10:53StingRayNote Added: 0009490
2021-01-30 12:01probdenisNote Added: 0009491
2021-01-30 23:48JOTDNote Added: 0009494
2021-12-11 22:46probdenisNote Added: 0010815
2022-12-29 17:52StingRayStatusassigned => confirmed
2022-12-29 17:52StingRayNote Added: 0012221
2022-12-29 18:01StingRayNote Edited: 0012221bug_revision_view_page.php?bugnote_id=12221#r1556
2022-12-29 18:46StingRayFile Added: PushOver.slave
2022-12-29 18:47StingRayNote Added: 0012225
2023-01-11 22:25RobDangerousNote Added: 0012296
2023-01-12 20:42StingRayNote Added: 0012305
2023-01-21 17:40StingRayStatusconfirmed => resolved
2023-01-21 17:40StingRayResolutionopen => fixed
2023-01-21 17:40StingRayNote Added: 0012338