Re: [hatari-devel] 8 runes of aerillion doesn't have sound |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/hatari-devel Archives
]
- To: hatari-devel@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [hatari-devel] 8 runes of aerillion doesn't have sound
- From: Laurent Sallafranque <laurent.sallafranque@xxxxxxx>
- Date: Sat, 9 Mar 2024 23:36:18 +0100
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/simple; d=free.fr; s=smtp-20201208; t=1710023779; bh=1ZkIDTGm8nNW7Veqa9M9YhB0JPrk0scgENUymesTQ1Y=; h=Date:Subject:To:References:From:In-Reply-To:From; b=qSUfd0BkyyC0Cw9nLuX5NXIHFiU/Rfbp/OwGwiWgq/ZE0j0FJYcpo9fkH/F5pmvDm pE0jZjTeHj6PWDDtEMMkxVAuyYBZvtJspbhlPkho972o/KfBjwH2NC0E/bqHS0+Kpq 5TCwHvilPb4GPJYl8iPfP+/GMKgjXaNvUPpa/hxSA18smFgkDHp5wdlDwQkhy+UGrl mi7pPic4R3QMYXy432k/gT+fuE4+UgRLQHHyqFvRBLirc5a6WXtqUYCkM6q7SqzIt0 scJCeoG/BZ2hL1LBNVz3kehADOakMjLEcunkXBNbSnGl97ITyh4YwAnEzhFFNgUshv Tcu6to1PTeqow==
Hi
That's better, but not perfect.
There's a problem, it seems that there's a first sample that is
forgotten (everything is out of sync), the first sample should repeat
once, but the music goes directly to the 2nd sample.
After the blue intro magneto cassette screen, the sequence should run
like this :
dc.l 193548*0,193548*1 ; 01 -
00.raw 1st half Dune logo appears // 2nd half logo removed
dc.l 193548*0,193548*1 ; 02 -
00.raw Torus appears
dc.l 193548*1,193548*1+193548/2 ; 03 -
01.raw Logo Thadoss appears
And it runs like this :
; dc.l 193548*0,193548*1 ; 01 -
00.raw not played
dc.l 193548*0,193548*1 ; 02
-00.raw 1st half : Background // 2nd half garbage + Torus appears
dc.l 193548*1,193548*1+193548/2 ; 03 -
01.raw Logo Thadoss display
It seems that the 1st sample is not played and the interrupt start at
the 2nd one.
I've tested with the "HEAD-100" version of hatari and it runs
perfectly (if needed, I can bisect).
(git bisect start HEAD HEAD~100)
A youtube video of the demo : https://www.youtube.com/watch?v=aow302WXZak
Ask mr if you need more explanations
Regards
Laurent
Le 09/03/2024 à 22:26, Nicolas Pomarède a écrit :
Le 09/03/2024 à 11:29, Laurent Sallafranque a écrit :
I was pretty sure of it, but I've bisected electric night (the
problem is certainly the same for 8 runes as it uses the same sound
engine).
Hi
this is fixed now ; I forgot to update crossbar.SNDINT_Signal value
after it was changed, making GPIP7 value always 1, which prevented the
GPIP7 interrupt :(
Also, that's a good example of the "end of frame" interrupt that can
be set in $FF8901, it's useful to validate the recent changes in
crossbar.c
Nicolas