Reply 40 of 110, by Cacodemon345
https://github.com/Cacodemon345/VSBHDASF/releases/tag/v1.5.1
New release featuring GM/GS Reset and Master Volume control support.
https://github.com/Cacodemon345/VSBHDASF/releases/tag/v1.5.1
New release featuring GM/GS Reset and Master Volume control support.
I couldn't find any info in project documentation about the soundfont file - where to get it. Maybe could be good to add it, if it came from windows or original SB driver CD, etc...
Gigabyte GA-P67-DS3-B3, Core i7-2600K @4,5GHz, 8GB DDR3, 128GB SSD, GTX970(GF7900GT), SB Audigy + YMF724F + DreamBlaster combo + LPC2ISA
@RayeR any SF2 file is valid, for example I'm using 2GMGSMT.SF2 from the AWE64 cards
Cacodemon345 wrote on 2025-01-28, 08:35:It's probably an issue of your configuration then, unfortunately. Make sure the executable is on v1.9. I don't know what else would cause the output to get garbled up.
One thing of note is that the Sound Blaster emulation code is somewhat of a mess; it's entirely possible to encounter weird machine-specific bugs.
It's ok, I'm using the v1.9 shareware, thank you for the help. When you say that the sound blaster emulation is a mess, are you talking about doom or VSBHDA?
wierd_w wrote on 2025-01-27, 22:59:I noted with git released sbemu that I *need* to restrict the amount of XMS memory for it to not crash when games use it. It is simply not optional.
I would suggest using the max xms flags with jemmex and limiting to something 'sane' for a dos pc, like 32mb.
Thanks, this sounds promising. Looking at the JEMM documentation, this would be something like
Config.sys:
DEVICE=C:\JEMMEX.EXE X2MAX=32767
Or simply DEVICE=C:\JEMMEX.EXE MAX=32767 ?
So I found the answer for this point back in the VSBHDA.TXT file - under the notes for extended memory - the switch to limit the reported memory for HDPMI32i -x. So I edited the startp4.bat file to
lh hdpmi32i -b -a -x
Cacodemon345 wrote on 2025-01-28, 07:30:Run JEMMEX with NOVCPI after loading VSBHDASF, because FastDOOM won't detect anything otherwise.
I only tested with Vanilla DOOM 2 v1.9 executable, not anything else. It may be running into strange sample rate issues that I forgot to take care of, since I only extensively tested it with the Sound Blaster 16 option.
Is there an explanation anywhere for what these options are and what they do? The JEMM documentation seems to assume a lot of knowledge, but searching doesn't seem to help with much practical advice. Wiki explains what VCPI is and its history, but is there any guide to why it should or shouldn't be used for different things? (and the same goes for most JEMM options!)
Now I've got it working in multiple games, can confirm it worked perfectly on a Core i3 Sandy Bridge Thinkpad. It also worked well on an "Athlon Neo" powered Thinkpad X100 (basically a netbook) so that bodes well for many netbook type machines! At the other extreme though, General MIDI was horribly garbled on a Via C3 600Mhz with 8237R south bridge.
One thing with the Thinkpad though, the first line of VSBHDA's output reported 'using headphone output', and indeed it only worked via headphones. How to switch the output?
New release is out, which attempts to address reports of garbled sound output from Doom v1.9.
justin1985 wrote on 2025-01-28, 13:07:Thanks, this sounds promising. Looking at the JEMM documentation, this would be something like Config.sys: DEVICE=C:\JEMMEX.EX […]
wierd_w wrote on 2025-01-27, 22:59:I noted with git released sbemu that I *need* to restrict the amount of XMS memory for it to not crash when games use it. It is simply not optional.
I would suggest using the max xms flags with jemmex and limiting to something 'sane' for a dos pc, like 32mb.
Thanks, this sounds promising. Looking at the JEMM documentation, this would be something like
Config.sys:
DEVICE=C:\JEMMEX.EXE X2MAX=32767
Or simply DEVICE=C:\JEMMEX.EXE MAX=32767 ?So I found the answer for this point back in the VSBHDA.TXT file - under the notes for extended memory - the switch to limit the reported memory for HDPMI32i -x. So I edited the startp4.bat file to
lh hdpmi32i -b -a -x
Cacodemon345 wrote on 2025-01-28, 07:30:Run JEMMEX with NOVCPI after loading VSBHDASF, because FastDOOM won't detect anything otherwise.
I only tested with Vanilla DOOM 2 v1.9 executable, not anything else. It may be running into strange sample rate issues that I forgot to take care of, since I only extensively tested it with the Sound Blaster 16 option.
Is there an explanation anywhere for what these options are and what they do? The JEMM documentation seems to assume a lot of knowledge, but searching doesn't seem to help with much practical advice. Wiki explains what VCPI is and its history, but is there any guide to why it should or shouldn't be used for different things? (and the same goes for most JEMM options!)
Now I've got it working in multiple games, can confirm it worked perfectly on a Core i3 Sandy Bridge Thinkpad. It also worked well on an "Athlon Neo" powered Thinkpad X100 (basically a netbook) so that bodes well for many netbook type machines! At the other extreme though, General MIDI was horribly garbled on a Via C3 600Mhz with 8237R south bridge.
One thing with the Thinkpad though, the first line of VSBHDA's output reported 'using headphone output', and indeed it only worked via headphones. How to switch the output?
NOVCPI option for JEMM completely hides the VCPI interface from programs. It should never be used by anything except HDPMI32i and VSBHDASF, period.
Also, VIA C3/Cyrix III CPUs before Nehemiah are kind of infamous for featuring slow FPU units; you shouldn't count on General MIDI options working well on those.
As for output, there's a /O parameter option, that ranges from 0 to 2. Try one of each and see what sticks.
Downloaded the latest version, no changes unlucky: https://youtu.be/AHZFEOF-o2M
It's not garbled, but seems to be kind of echoing. Since music output appears to be indeed fine, it looks like it's the resampler being problematic.
You should try fiddling with the /F parameter (sets the frequency, e.g. /F44100 sets the sound hardware output frequency to 44100) and see if it helps.
Cacodemon345 wrote on 2025-01-28, 19:58:It's not garbled, but seems to be kind of echoing. Since music output appears to be indeed fine, it looks like it's the resampler being problematic.
You should try fiddling with the /F parameter (sets the frequency, e.g. /F44100 sets the sound hardware output frequency to 44100) and see if it helps.
Yes music works fine.
I tried F44100 and F22050, no changes but on the other hand VSBHDASF output still shows the output is 48000 hz despite the /fxxx switch.
Given the fact that I don't have any real AC97 hardware, this will probably not be easily fixable. But I will try.
ViTi95 wrote on 2025-01-28, 11:44:@RayeR any SF2 file is valid, for example I'm using 2GMGSMT.SF2 from the AWE64 cards
OK, I used SOUNDFONT=4GMGSMT.SF2 from SB live driver CD (there are 2/4/8MB version).
Tested in Doom and Blood and wavetable music plays fine and clean on my i7. Good job!
Gigabyte GA-P67-DS3-B3, Core i7-2600K @4,5GHz, 8GB DDR3, 128GB SSD, GTX970(GF7900GT), SB Audigy + YMF724F + DreamBlaster combo + LPC2ISA
I attempted reproducing this bug with output frequency forced to 48000 in the code. Safe to say, unfortunately, I couldn't reproduce the echo issues.
Last thing I'd say is to try with other executables (especially the normal one for all CPUs); I suspect misoptimizations are at fault here.
Cacodemon345 wrote on 2025-01-29, 06:07:I attempted reproducing this bug with output frequency forced to 48000 in the code. Safe to say, unfortunately, I couldn't reproduce the echo issues.
Last thing I'd say is to try with other executables (especially the normal one for all CPUs); I suspect misoptimizations are at fault here.
I will test the other executables and report back. It could very well be a chipset issue, the ecs + sis combo is very flaky.
EDIT:
Done some testing, lead to no change:
Other exes
Rem most of the complementary stuff on config.sys and autoexec.bat
update to latest jemmex and jload version
removed the switches I use with HDPMI32I
I tried sbemu (b5 and b4) they do not exibit the issue (also sbemu is able to set the frequency).
VSBHDASF: https://youtu.be/PWR3e7JGW4I
SBEMU: https://youtu.be/r_JVz1H8lBo
Testing on VirtualBox's AC97 option, it has the exact same garbled sound problems as well, and it doesn't happen on upstream. Seems like the backport didn't work out well. I'm seeing what I can do to deal with this.
Cacodemon345 wrote on 2025-01-29, 09:08:Testing on VirtualBox's AC97 option, it has the exact same garbled sound problems as well, and it doesn't happen on upstream. Seems like the backport didn't work out well. I'm seeing what I can do to deal with this.
Thank you.
It's just a trivial thing but the colored output of sbemu about what is enabled (green) or disabled (red) seems nicer and more readable (if something is wrong) would it be possible to have it?
New release that fixes AC97 playback problems on Vanilla Doom 1/2. Anyone with a SiS 7012 audio chipset, please test.
Nemo1985 wrote on 2025-01-29, 09:11:Cacodemon345 wrote on 2025-01-29, 09:08:Testing on VirtualBox's AC97 option, it has the exact same garbled sound problems as well, and it doesn't happen on upstream. Seems like the backport didn't work out well. I'm seeing what I can do to deal with this.
Thank you.
It's just a trivial thing but the colored output of sbemu about what is enabled (green) or disabled (red) seems nicer and more readable (if something is wrong) would it be possible to have it?
Possible but it'll be considered later.
Cacodemon345 wrote on 2025-01-29, 11:16:New release that fixes AC97 playback problems on Vanilla Doom 1/2. Anyone with a SiS 7012 audio chipset, please test.
It doesn't work at all.
Doom doesn't produce any sound output (not even music works).
Duke3d doesn't boot complains about invalid or conflicting dma channel (and then says something about 16bit dma) same does blood.
https://github.com/Cacodemon345/VSBHDASF/rele … es/tag/v1.5.1.3
New release for SiS7012 audio chipsets. Tested to be working correctly.
I tested VSBHDASF on a Dell D600 laptop, which has an AC97 chip. Wavetable music and Sound Blaster sound fx worked on Duke3d, Doom, Descent, Warcraft2, Dark Forces, Tie Fighter and SimCity Classic. Dark Forces will crash on fast cpu's when using general midi, so I used CpuSpd to slowdown the Pentium M chip. Tie Fighter may have the same issue. Let me know if there are any games you'd like me to test.
Cacodemon345 wrote on 2025-01-28, 09:37:https://github.com/Cacodemon345/VSBHDASF/releases/tag/v1.5.1
New release featuring GM/GS Reset and Master Volume control support.
Thanks! It works great.
TUF B450PRO/RyZen 3800X/7900GS/ALC S1200A/SB0060.
To ALC S1200A the output:
I packed an external driver, and I started this way with SBEMU, just to plug into any machine. Input for 7 types of smooth sbemu, smooth vsbhda, 16bit vsbhda (to Tyrian), and now vsbhdasf, the latter just for the B450, but for the time being, I use it for dos anyway.
GRUB+I deciphered 1.44FDD-> HDD emulation and then wrote in external M2 NVME SSD.
2 FAT32 partition, 8GB (boot)+90GB (the latter is a lot of tools+game).
The iron:
The SB0060 configuration follows.
The 3 body problems cannot be solved, neither for future quantum computers, even for the remainder of the universe. The Proton 2D is circling a planet and stepping back to the quantum size in 11 dimensions.