VOGONS


First post, by vasyl

User metadata
Rank Oldbie
Rank
Oldbie

I've submitted a little patch to emulate HiColor DAC on Tseng ET4000, a.k.a. Sierra DAC. The emulated DAC is not exactly Sierra version, WHATVGA identifies it as an UMC clone but everything I could find works fine. So far, in addition to WHATVGA itself, there is an old animation tool called PCAnimate and freeware Targa viewer TTGA. Both work perfectly with this patch. If anybody knows of any games or other tools that have HiColor DAC support, please let me know. Also, if anybody has any good documentation, that would help too -- the only detailed documentation is a file in VGADOC and, as it usually the case, some important things are not there.

Reply 1 of 7, by DosFreak

User metadata
Rank l33t++
Rank
l33t++

He lives! Thanks for the patch. May get the time to try it out next week.

DOSBox Compilation Guides
DosBox Feature Request Thread
PC Game Compatibility List
How To Ask Questions The Smart Way
Running DRM games offline

Reply 2 of 7, by DosFreak

User metadata
Rank l33t++
Rank
l33t++

Only game I can think of off-hand:

Don't know if you can see this thread or not:

Topic 15028

DOSBox Compilation Guides
DosBox Feature Request Thread
PC Game Compatibility List
How To Ask Questions The Smart Way
Running DRM games offline

Reply 3 of 7, by h-a-l-9000

User metadata
Rank DOSBox Author
Rank
DOSBox Author

There's also the Transgression 2 Demo (start with tgr2 -hc), which uses an unknown DAC. My Megabuild 5 has a half-assed implementation for it in machine=vgaonly.

Have you seen this datasheet? It might have additional info.
http://www.datasheetarchive.com/SC11487-datasheet.html

1+1=10

Reply 5 of 7, by vasyl

User metadata
Rank Oldbie
Rank
Oldbie

There are some issues with Doom 0.5 besides the DAC -- those illegal reads/writes are not related, at least, directly. The game does switch to 15-bit color mode but locks up immediately after that.
Transgression 2 demo behaves oddly. It expects DAC at least very similar to the implemented but the image on the screen is still not right. I'll take a look, it might be something trivial.
That datasheet is useful to understand what that DAC can do but it does not explain how the DAC is integrated into the VGA circuitry. I'll have to bite the bullet and decipher WHATVGA sources -- it's hard to believe that Pascal code can be that unreadable.

Reply 6 of 7, by h-a-l-9000

User metadata
Rank DOSBox Author
Rank
DOSBox Author

On real hardware, TGR2 was closest to perfect on a S3 864 card with 86C716 DAC. The final issue was 32k colors vs 65k. The demo maker seems to have had some weird DAC.

1+1=10

Reply 7 of 7, by vasyl

User metadata
Rank Oldbie
Rank
Oldbie

That demo has only two issues with my patch. One is kinda expected -- the emulation does not handle this type of geometry change properly atm so the screen width is incorrect. The other one is exactly the same issue of 32k vs 65k. The demo is definitely generating 16-bit image but the values sent to DAC command register correspond to 15-bit color modes for most DACs. I could not find any DAC that uses 0x80 for 16-color mode. By hacking the sent value to 0xc0 I was able to get the right colors on the screen. This sounds very strange -- the expected DAC is a close relative to Sierra SC series but uses different bit assignments... More likely, just a typo or a bug in the demo itself.