Viewing Issue Advanced Details
|ID||Category [?]||Severity [?]||Reproducibility||Date Submitted||Last Update|
|00187||Sound||Minor||Sometimes||Jan 26, 2008, 04:29||May 7, 2022, 05:58|
|Version||0.101u5||Fixed in Version||Build|
|Fixed in Git Commit||Github Pull Request #|
|Summary||00187: looping: Sound glitches.|
|Description||When you enter the second part of the game, after destroying the terminal, most of the time music begins to play and no sound effects are played (normal I guess), but other times no music is played and sound effects continue to play.|
|Steps To Reproduce||
Destroy the terminal.
A voice will tell 'game is over'? Sometimes this voice sound isn't played entirely. If that's the case, continue to the 2nd part.
No sound effects will be played anymore in 1st part, no music will be played in the second part of the game.
|Affected Sets / Systems||looping|
|There are no relationship linked to this issue.|
May 7, 2022, 04:56
edited on: May 7, 2022, 04:57
|Was this also fixed by https://git.redump.net/mame/commit/?id=9afc6012e0a3d6d131ed63196fec3b1bd0bd605d ?|
May 7, 2022, 05:58
edited on: May 7, 2022, 06:07
Destroying the terminal does not trigger "The game is over" anymore (I even guess that was unemulated protection in MAME 0.37 or such where it resetted itself after that), but the speech delays and sometimes skips certain words (noticeable with highscore text and spoken numbers of remaining lives(?)). Apparently the sound/speech hardware gets out of sync with the game CPU. Occasionally the speech gets stuck and also mutes the main sound. Later (after starting another game?) speech makes glitch noises and then normal sound continues.
It sounds like when data pulses (by IRQ?) are too long and so pause the sound chips. (I collect music keyboards and do circuit bending. When I experimented with sound chip data lines I often got similar behaviour.) May it be that the original hardware limits IRQ pulse duration e.g. by edge sensitivity or analogue means (serial capacitor?).