Skip to content

(LLM-generated content removed) - #280

Closed
GLITCHKISS wants to merge 1 commit into
devkitPro:masterfrom
GLITCHKISS:asnd-voice-update-on-send
Closed

GLITCHKISS wants to merge 1 commit into
devkitPro:masterfrom
GLITCHKISS:asnd-voice-update-on-send

Conversation

@GLITCHKISS

@GLITCHKISS GLITCHKISS commented Oct 29, 2026 •

Copy link
Copy Markdown

LLM-generated content removed

VOICE_UPDATE was cleared only for voice 0 (in the DMA callback) and by the
look-ahead, which clears the next active voice's flag while the voice before it
is still being mixed. A voice set with ASND_SetVoice after the look-ahead had
passed it was sent to the DSP with the flag still on. When its mix came back,
__dsp_requestcallback took the flag to mean the voice had been set again during
the mix, discarded the DSP's progress and sent the voice from its start once
more, so the first 1024 samples of the sound played twice.

Clear the flag at the point the voice is sent, as the voice 0 path already
does. A voice set while it is on the DSP still sets the flag again and is
restarted as before.

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
@fincs fincs closed this Oct 29, 2026
@fincs fincs changed the title libasnd: clear VOICE_UPDATE when the voice is sent to the DSP (LLM-generated content removed) Oct 29, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants