VOGONS


First post, by Kahenraz

User metadata
Rank l33t
Rank
l33t

Unfortunately, this isn't something that comes across well with a photo but I can assure you that text that should be white is a very pale shade of yellow. What's odd is that it isn't consistent. For example, white on gray looks OK but white on black is a shade of yellow on black instead.

The attachment IMG_20210815_022333.jpg is no longer available
The attachment Capture.PNG is no longer available

If I boot into Windows and open a command prompt the color is fine until I make it full screen text mode where it becomes yellow again. It seems to be a text mode problem specifically.

Reply 1 of 5, by shamino

User metadata
Rank l33t
Rank
l33t

I don't have a solution, but pale yellow seems to imply that the Blue component of the RGB signal is not as high as it should be. I'm not a color person, I'm just saying that based on playing with RGB values in GIMP.

Assuming this is an analog electronics problem, then my best guess is that Blue is having some problem changing fast enough - which would explain why it can saturate in time if the background is gray, but not black.
In the closeup photo, it looks like there's a yellow fringe on the leading edge of the white, but only after a stretch where the image was previously black. After it has time to "warm up", it eventually reaches white. So it looks like it's a problem with the Blue line being slow to change it's signal voltage.

I wonder if there's some color test pattern that would add any evidence of a problem on the Blue signal. If you displayed vertical bars in some pattern like
BLACK-RED-BLACK-GREEN-BLACK-BLUE
where each of the colors is a fully saturated (bright) value
then I wonder if there would be a noticeable shading effect on the left edge of the Blue (and whether anything similar would show up with Red and Green)

Anyway that's not an answer on how to fix it, just a theory of what the problem is.
The problem with this idea is that it doesn't explain why it works better in a window.
Windowed mode is probably higher res, so maybe there's a difference in clock frequencies somewhere that factors into this somehow. Or I'm just wrong about what's happening.

I don't know how bad it looks in person, but it's possible this is an artifact of intentional filtering on the card, not necessarily a component failure. Some NVidia cards were known for blurriness caused by filtering they put on the RGB signal lines, probably for FCC compliance reasons.
It could also be a problem with the monitor. Does it happen with other monitors, or with other cards connected to the same monitor? Have you tried a different cable?

Reply 2 of 5, by Tiido

User metadata
Rank l33t
Rank
l33t

This might be fixed with another video BIOS version for the particular card, if available. Windows drivers talk to hardware directly and program things they see fit, without normally respecting what the video BIOS does, but in DOS only thing that does matter is what video BIOS does when video modes etc. are being set.

T-04YBSC, a new YMF71x based sound card & Official VOGONS thread about it
Newly made 4MB 60ns 30pin SIMMs ~
mida sa loed ? nagunii aru ei saa 😜

Reply 3 of 5, by Kahenraz

User metadata
Rank l33t
Rank
l33t
shamino wrote on 2021-08-15, 22:43:

Assuming this is an analog electronics problem, then my best guess is that Blue is having some problem changing fast enough - which would explain why it can saturate in time if the background is gray, but not black.
In the closeup photo, it looks like there's a yellow fringe on the leading edge of the white, but only after a stretch where the image was previously black. After it has time to "warm up", it eventually reaches white. So it looks like it's a problem with the Blue line being slow to change it's signal voltage.

I can confirm that the issue is as described with fringing where the color changes from black to white. This can be demonstrated by comparing vertical bars to horizontal bars of white on black.

The attachment IMG_20210815_191421.jpg is no longer available

I do most of my testing hooked up to a 4:3 1280x1024 LCD which I think makes the problem more pronounced due to the hardware filter and scaling but the fringing is still visible connected to CRT.

I've been swapping around all kinds of different graphics cards using the same cable and monitor and this is the first to demonstrate this issue. Something is probably slightly out of spec on this card. Unfortunately, I don't have any other card of this model to compare or to swap BIOS chips.

Reply 4 of 5, by BitWrangler

User metadata
Rank l33t++
Rank
l33t++

I'd clean all the pins in the socket and touch up the solder points just in case there's a tiny oxide film making it that tad weaker.

Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.

Reply 5 of 5, by Kahenraz

User metadata
Rank l33t
Rank
l33t

If it was a signal problem then you would think that it would be visible even in Windows. The affect only seems to occur in text mode.