Plantronics 256 Color Mode (Technically Likely Less, but I didn't calculate the pal, I captured it)

[chjmartin2]

Hi,

I'm not sure why I didn't post this before but watching 8-Bit Guy's Super CGA Video it occurred to me that there was a mention of an odd 136 color mode and he commented "that must be some kind of composite thing." Well, this made me remember that I played with composite output on my Plantronics Compatible PVC4 card to display really nice 160x200x256 color images. I don't know if the color combos actually equals 256 colors. I used the below image and captured it, then scaled it down to the right count and took those color values. I am sure @reenigne would be able to tell us the right math but this is how I did it. Anyway the converter takes in a 160x200 RGB BMP and outputs a BMP file, then it has to be converted to GIF. The file can then be viewed using Compushow (but make sure to turn off dither!) Here is the demo file I used:



Anyway, it outputs this image on a composite screen:

I took this capture and then calculated RGB values for it. Again there is certainly better math but you'll see it works. Here is the Paradise PVC4 card (I don't know if that is actually the card, but certainly it has the chipset. Take note of the composite output which is part of the PVC4 with some additional circuitry. This is why you can modify a computer that uses the PVC4 to put out composite.)



Here is a quick demo view:



There is a zip file attachment of the GIF files in the slideshow ready to use on real hardware. I don't believe there is an emulator out there that would support this. I'll release the converter if anybody expresses interest.

Here is a capture from tonight of all of the images in the zip file on this post.

Thanks,

Chris

 

[VileR]

I'm not sure why I didn't post this before but watching 8-Bit Guy's Super CGA Video it occurred to me that there was a mention of an odd 136 color mode and he commented "that must be some kind of composite thing." Well, this made me remember that I played with composite output on my Plantronics Compatible PVC4 card to display really nice 160x200x256 color images. I don't know if the color combos actually equals 256 colors. I used the below image and captured it, then scaled it down to the right count and took those color values. I am sure @reenigne would be able to tell us the right math but this is how I did it.

Ha, nice work! Although I gotta wonder whether that palette is the same across the original ColorPlus board and its various compatibles. CGA clones had that problem: the color and pixel clocks don't necessarily have the same phase relationship as on the original IBM boards, so different cards produce different sets of artifact colors.

Nice video from the 8-Bit Guy... although if I see this "composite = 160x200" misconception repeated ever again, I might just end up writing a windy, rambling blog post about why that's not the case

BTW, that 136-color mode on the QuadColor II wasn't a composite thing - it was done with temporal blending of RGB colors on alternating frames, much like a certain demo did on the Tandy 1000 in software (I think that was yours as well?)

 

[Timo W.]

BTW, that 136-color mode on the QuadColor II wasn't a composite thing

Speaking of that, I own the original Dr. Halo disks he could not find. But I don't think he's making a follow-up video. At least, he did not react on my offer to dump them. I may dump them anyway once I have one of my 5.25" disk drive-equipped systems set up again.

 

[chjmartin2]

Ha, nice work! Although I gotta wonder whether that palette is the same across the original ColorPlus board and its various compatibles. CGA clones had that problem: the color and pixel clocks don't necessarily have the same phase relationship as on the original IBM boards, so different cards produce different sets of artifact colors.

I am not even sure my palette is correct as it is. I only have the one card although I think I may have a Tandy that has composite out so that would be a test as well because the 320x200x16 mode would do the same thing.

Nice video from the 8-Bit Guy... although if I see this "composite = 160x200" misconception repeated ever again, I might just end up writing a windy, rambling blog post about why that's not the case

I would appreciate that blog because I am under the same delusion apparently. CGA composite in 640x200 uses 4 pixels of 1 bit patterns hence the 16 colors, so 640/4 is 160. Help me out here??

BTW, that 136-color mode on the QuadColor II wasn't a composite thing - it was done with temporal blending of RGB colors on alternating frames, much like a certain demo did on the Tandy 1000 in software (I think that was yours as well?)

Ohhh! So yes it’s the same thing then. Just using the 30 hz flicker - concept ‘stolen’ from the C64 tweakers. I have a few with CGA that aren’t released.

 

[chjmartin2]

Speaking of that, I own the original Dr. Halo disks he could not find. But I don't think he's making a follow-up video. At least, he did not react on my offer to dump them. I may dump them anyway once I have one of my 5.25" disk drive-equipped systems set up again.

Ugh - we don’t need a follow up video but for sure we have to see this 136 color mode in action?!?

 

[chjmartin2]

Speaking of that, I own the original Dr. Halo disks he could not find. But I don't think he's making a follow-up video. At least, he did not react on my offer to dump them. I may dump them anyway once I have one of my 5.25" disk drive-equipped systems set up again.

By the way, for sure you need to dump them for posterity as it seems pretty clear that they are not available.

 

[jafir]

Those disks may or may not have what is needed. I’ve got a copy of “Dr. Genius” which is a rebranded version of Dr. Halo, and it doesn’t seem to have the quadcolor support.

 

[Plasma]

There are about a dozen versions of "Dr. Halo" supporting random combinations of graphics cards. I also own original Dr. Halo III disks and there is no Quadcolor driver. There are copies of Dr. Halo II/III on WinWorld but they don't have the driver either. I don't think there's anything wrong with the dumps.

 

[Timo W.]

Wonder where all you people were when he asked for dumps of the original disks? But I guess I don't have to bother dumping the disks then.

I have not checked the disks for ages, so no idea if the driver file is on the disks. But QuadColor II support is certainly there, as it is stated in the manual. I don't see why they would omit the driver. I'd assume it was missing from the dump he used because someone freed up some space on the disk for saving pictures.

 

[jafir]

I’d recommend dumping it. Like I said, it might have it or might not. All I was saying is that you cannot be sure it’s there without looking.

I never saw any request from him, if he posts here, I have no idea what his username is.

 

[Plasma]

Wonder where all you people were when he asked for dumps of the original disks? But I guess I don't have to bother dumping the disks then.

I have not checked the disks for ages, so no idea if the driver file is on the disks. But QuadColor II support is certainly there, as it is stated in the manual. I don't see why they would omit the driver. I'd assume it was missing from the dump he used because someone freed up some space on the disk for saving pictures.

Well I don't subscribe to 8-bit guy and he's not a member here...

I don't think any of the dumps are modified. I know my disks were not modified because they came out of a sealed box, and they don't have the driver. I have seen versions taking anywhere from 1-4 disks, so I assume that the publisher goofed when they were removing drivers to fit it onto 2 floppies.

 

[jafir]

I just checked another set of images I have, from an Emerson mouse, and these do list quadcolor II and quadscreen in setup, and neither driver is on the disk set. There are also no user data files on either disk. I dumped these before I knew what I was doing, so I don't have the original stream files. I can dump them again. Normally the kryoflux tools can detect if an MFM disk has been modified after it's original write. I'm not exactly sure how they do it, but it seems to be accurate. For example disks that are original but have had registration information written show those sectors were modified.

 

[VileR]

Nice video from the 8-Bit Guy... although if I see this "composite = 160x200" misconception repeated ever again, I might just end up writing a windy, rambling blog post about why that's not the case

I would appreciate that blog because I am under the same delusion apparently. CGA composite in 640x200 uses 4 pixels of 1 bit patterns hence the 16 colors, so 640/4 is 160. Help me out here??

The thing is that these two situations don't exactly correspond in terms of a simple division: in digital RGBI, a pixel's color and its position are completely independent factors. In NTSC the color of a pixel depends on its horizontal position (modulo the color carrier cycle), so you're not really dealing wth two separate variables. (Some illustrations here.)

The color carrier frequency is exactly 160 cycles per visible scanline, so that's your effective color resolution (chroma). But composite video has different resolutions for chroma and luma, and higher frequency detail still shows through in the latter. How much really depends on the image content and decoder bandwidths - you can choose not to decode color at all, and you'll always get 640 distinguishable pixels (so 80-col CGA text is perfectly legible on monochrome composite displays).

As a good example, here's one game using CGA on composite: https://www.mobygames.com/game/pc-b...t-simulator-v20/screenshots/gameShotId,39387/
Compare to the same game at a real 160x200 resolution (PCjr): https://www.mobygames.com/game/pc-b...-simulator-v20/screenshots/gameShotId,160197/
The first shot clearly shows much more detail than the second one: see the circles and digits on the instruments especially.

I've seen some graphics artists treat it as 'only' 160 pixels per line... that just misses out on the potential for much nicer images.

 

[Trixter]

I just checked another set of images I have, from an Emerson mouse, and these do list quadcolor II and quadscreen in setup, and neither driver is on the disk set. There are also no user data files on either disk. I dumped these before I knew what I was doing, so I don't have the original stream files. I can dump them again. Normally the kryoflux tools can detect if an MFM disk has been modified after it's original write. I'm not exactly sure how they do it, but it seems to be accurate. For example disks that are original but have had registration information written show those sectors were modified.

I helped him with trying to find quadcolor-supporting programs, and the reason for the Dr. Halo requests is because that's the only program I could find that actually mentioned support for it. I didn't realize my Dr. Halo II disks didn't actually have the driver, and my disks seem the same as the ones floating around.

Maybe a better question we should have asked: What software actually supported the Quadcolor II?

 

[chjmartin2]

The thing is that these two situations don't exactly correspond in terms of a simple division: in digital RGBI, a pixel's color and its position are completely independent factors. In NTSC the color of a pixel depends on its horizontal position (modulo the color carrier cycle), so you're not really dealing wth two separate variables. (Some illustrations here.)

The color carrier frequency is exactly 160 cycles per visible scanline, so that's your effective color resolution (chroma). But composite video has different resolutions for chroma and luma, and higher frequency detail still shows through in the latter. How much really depends on the image content and decoder bandwidths - you can choose not to decode color at all, and you'll always get 640 distinguishable pixels (so 80-col CGA text is perfectly legible on monochrome composite displays).

As a good example, here's one game using CGA on composite: https://www.mobygames.com/game/pc-b...t-simulator-v20/screenshots/gameShotId,39387/
Compare to the same game at a real 160x200 resolution (PCjr): https://www.mobygames.com/game/pc-b...-simulator-v20/screenshots/gameShotId,160197/
The first shot clearly shows much more detail than the second one: see the circles and digits on the instruments especially.

I've seen some graphics artists treat it as 'only' 160 pixels per line... that just misses out on the potential for much nicer images.

Yeah this needs to be shouted from the rooftop. The pictures prove out the comparison. Makes me wonder how to exploit it. I have noticed that the ‘solid’ colors actually have some bars in them depending on the pattern.

 

[Plasma]

BTW ZSoft Paintbrush IV Plus claims to support the Quadcolor II at 640x200 16 colors.

 

[Plasma]

I have located the elusive Dr. Halo Quadcolor II driver on the DFI-branded "Halo Paint and Graph" disks. I believe this to be the most complete version of Dr. Halo III, supporting 34 graphics cards (plus VGA if you select IBM PS/2). Attached for posterity.

1. Ahead EGA 2001
2. Ahead Wizard
3. Amdek MAI
4. AT&T DEB
5. AT&T Indigenous
6. ATI Wonder
7. Atronics MegaGraph
8. Conographics 40
9. Everex Micro Enhancer
10. Genoa Super EGA
11. Genius VHR
12. Hercules InColor
13. Hercules Mono
14. IBM CGA
15. IBM EGA
16. Mylex
17. Number Nine
18. Paradise Autoswitch
19. Photon Mega
20. Quadcolor II
21. Quadscreen
22. Sigma Color 400
23. Sigma EGA 480
24. SigmaVGA
25. STB EGA PLUS
26. STB Graphix Plus II
27. STB VGA EXTRA
28. Tecmar EGA 480/800
29. Tecmar Graphics Master
30. Toshiba T3100
31. Tseng Labs EVA
32. Tseng Labs EVA/480
33. Verticom CAD480
34. VEGA/Deluxe

 

[Trixter]

Oh wow! Even the AT&T DEB is here. Thanks so much for finding this, now I have something to test my Quadcolor II with!