zyzzle wrote on Yesterday, 02:10:

Thanks for information about "P" capturing screenshots. That explains why I was only ever able to get a single screenshot no mat […]
Show full quote

Thanks for information about "P" capturing screenshots. That explains why I was only ever able to get a single screenshot no matter how many times I tried! The file is named something like 20260321.bmp which is the max DOS 8.3 character limit. Perhaps modify code to make it something like: 26032100.bmp to 26032199.bmp (or in mmddyy format which would be 03212600) which would allow for 100 screenshots per day!

Yes, I have Intel HDA + SBEMU sound. Even the mono music is good; I can't wait to try the stereo. Thanks for your hints and help.

I'll modify sdlpal.cfg as you specified and re-post my new logs when I get a chance (this will be in 2-3 days, as I'm out of town now and that system with SDLPal DOS is at home).

I'm surprised there hasn't been more commentary in this topic. The game is outstanding!

Perhaps we could post more screenshots.

Testing this on a X99M Killer/3.1 board modded with dISAppointment v0.2 to connect a real Sound Blaster 16 (CT2950). So far it doesn't work.

I'm using the freeware (Win9x) data files you linked. I could get it play the first few cinematics fine, just without any sound. After that I get a black screen with the FM playing a note and everything apparently hung. No response from keyboard (LEDs don't change when I press CAPS LOCK) and I can only do a hard reset.

Using MS-DOS 7.10 (Win98SE) and FreeDOS, with and without JEMM386, CWSDPMI and HDPMI32 (from HX). No difference. Note that JEMM386 must not be run with NOVCPI parameter as that blocks DPMI servers from running.

I'm using v2026-03-15-g839ffaf workflow build. The workflow artifact only has the executable itself, but no SDLPAL.CFG. Where can I obtain this file? I don't see a default configuration file created on-the-fly.

I'm able to see the SDL log. So far it's able to list all video modes I have, and the last few lines (before the game hung the system) appeared like this.

1SDL not built with thread support
2Created renderer: software
3Audio subsystem is not initialized
4Invalid audio device ID
5SDL not built with thread support
6Invalid audio device ID
7Invalid audio device ID
8Invalid audio device ID

The beginning of the SDL log per run has only VClock-related stuffs before listing video modes. Not sure if there's anything out of ordinary on my system.

1[VCLOCK] [vclock] lazy_init(): starting
2[VCLOCK] [vclock] pit_divisor = 11931
3[VCLOCK] [vclock] IRQ0 vector got (sel=2b, addr=3c40)
4[VCLOCK] VCLOCK ISR code size is 232 bytes
5DOS: Keyboard ISR code size is 304 bytes

EDIT: I just looked at the mode listing output... Some were ignored due to "bad pixel format". Which mode does SDLPAL DOS port explicitly require?

LSS10999 wrote on Yesterday, 16:34:

Testing this on a X99M Killer/3.1 board modded with dISAppointment v0.2 to connect a real Sound Blaster 16 (CT2950). So far it d […]
Show full quote

Testing this on a X99M Killer/3.1 board modded with dISAppointment v0.2 to connect a real Sound Blaster 16 (CT2950). So far it doesn't work.

I'm using the freeware (Win9x) data files you linked. I could get it play the first few cinematics fine, just without any sound. After that I get a black screen with the FM playing a note and everything apparently hung. No response from keyboard (LEDs don't change when I press CAPS LOCK) and I can only do a hard reset.

Using MS-DOS 7.10 (Win98SE) and FreeDOS, with and without JEMM386, CWSDPMI and HDPMI32 (from HX). No difference. Note that JEMM386 must not be run with NOVCPI parameter as that blocks DPMI servers from running.

I'm using v2026-03-15-g839ffaf workflow build. The workflow artifact only has the executable itself, but no SDLPAL.CFG. Where can I obtain this file? I don't see a default configuration file created on-the-fly.

I'm able to see the SDL log. So far it's able to list all video modes I have, and the last few lines (before the game hung the system) appeared like this.

Copy code to clipboard

1SDL not built with thread support
2Created renderer: software
3Audio subsystem is not initialized
4Invalid audio device ID
5SDL not built with thread support
6Invalid audio device ID
7Invalid audio device ID
8Invalid audio device ID

The beginning of the SDL log per run has only VClock-related stuffs before listing video modes. Not sure if there's anything out of ordinary on my system.

Copy code to clipboard

1[VCLOCK] [vclock] lazy_init(): starting
2[VCLOCK] [vclock] pit_divisor = 11931
3[VCLOCK] [vclock] IRQ0 vector got (sel=2b, addr=3c40)
4[VCLOCK] VCLOCK ISR code size is 232 bytes
5DOS: Keyboard ISR code size is 304 bytes

EDIT: I just looked at the mode listing output... Some were ignored due to "bad pixel format". Which mode does SDLPAL DOS port explicitly require?

Thank you so much! We’ve been eagerly awaiting the test results for the original Sound Blaster.

We are currently addressing the issue of FM music not functioning properly on old sound cards. It now works with the software-emulated FM implementations in SBEMU, VSBHDA and VSBCMI, as well as with the limited hardware support they offer. Your test results are invaluable. They help us confirm that this is likely an issue with our implementation rather than a compatibility issue with the PCI sound card.

SDLPAL.CFG is written to disk when the game exits normally, and the game itself also has a set of default settings. If you downloaded the English Patch, it includes an SDLPAL.CFG file for Windows, but I think this actually allows you to bypass the current sound card bug in the DOS version by running it in silent mode.

The relevant settings should be something like this:
UseSurroundOPL=0
Music=RIX
OPLCore=MAME

Set OPLCore to an emulated one to disable hardware OPLchip access.

It's great that the AVI cinematics plays. The lack of AVI sound is to be expected, since Sound Blaster DMA audio hasn't been implemented yet (thus the emulated FM cores do not work, either).

Regarding pixel mode, the game renders at a resolution of 320x200 with RGB 888 color; the actual output resolution depends on the settings.