-

As a photographer, those "blue pixels" and the bright white clouds within are very similar to the typical overexposure "clipping." Could it be your output has been set to 16-235 instead of 0-255? While this setting is usually reserved for HDMI output, your symptom is way too similar to clipping at 235.

Take a look at the attached calibration chart with full-screen; where do blue pixels appear?

Somehow I have a feeling that red pixels might appear around black areas......

dormcat wrote on 2023-09-08, 11:45:

As a photographer, those "blue pixels" and the bright white clouds within are very similar to the typical overexposure "clipping […]
Show full quote

As a photographer, those "blue pixels" and the bright white clouds within are very similar to the typical overexposure "clipping." Could it be your output has been set to 16-235 instead of 0-255? While this setting is usually reserved for HDMI output, your symptom is way too similar to clipping at 235.

Take a look at the attached calibration chart with full-screen; where do blue pixels appear?
HDTVTestPattern.png

Somehow I have a feeling that red pixels might appear around black areas......

Nice, I'm going to look at it, but unfortunately I store this PC in my office as a non 100% hardware. Going to reply back on Monday. 😀
What if it has this calibration problem you mentioned? I mean what could be done to amend it?
Thanks

dormcat wrote on 2023-09-08, 11:45:

As a photographer, those "blue pixels" and the bright white clouds within are very similar to the typical overexposure "clipping […]
Show full quote

As a photographer, those "blue pixels" and the bright white clouds within are very similar to the typical overexposure "clipping." Could it be your output has been set to 16-235 instead of 0-255? While this setting is usually reserved for HDMI output, your symptom is way too similar to clipping at 235.

Take a look at the attached calibration chart with full-screen; where do blue pixels appear?
HDTVTestPattern.png

Somehow I have a feeling that red pixels might appear around black areas......

There it is.

That's interesting, so apparently it completely lacks red in that specific color gradient area, so everything that have red mixed in will display color without it. Probably an obscure driver bug.

kalanneni wrote on 2023-09-11, 06:13:

There it is.

So it's not clipping but missing at or around 219 of the red channel. Never seen this kind of malfunctioning before.

Do you have .inf/.icm files installed? https://www.driversforhp.com/download-hp/hp-7 … tware-free.html

giantclam wrote on 2023-09-11, 07:05:

Do you have .inf/.icm files installed? https://www.driversforhp.com/download-hp/hp-7 … tware-free.html

Done it now, but nothing changed. I'm starting to feel desperate.

have you tried dialing back to a lower color depth (e.g. 16 bit)?

drosse1meyer wrote on 2023-09-11, 12:11:

have you tried dialing back to a lower color depth (e.g. 16 bit)?

I remember trying it earlier and didn't help, but checked it now as well. It seems there are more of that "false" pixels in 16 bit.

kalanneni wrote on 2023-09-11, 12:35:

drosse1meyer wrote on 2023-09-11, 12:11:

have you tried dialing back to a lower color depth (e.g. 16 bit)?

I remember trying it earlier and didn't help, but checked it now as well. It seems there are more of that "false" pixels in 16 bit.

🙁

It may be worth trsting under real mode DOS with a VESA test suite, like the one thst comes with UniVBE or possibly Scitech Display Doctor .

If the issue is present too, it would exclude the Windows driver as a culprit.

I wonder it this might be due to a custom VBIOS meant for some industrial/scientific application.

Does yours have a NTSC/PAL jumper?

edit: just thinking about this ...if the screenshot from winxp is fine, that means the framebuffer is fine .... so this problem hints at a bad RAMDAC ; trying a different monitor would be a check

giantclam wrote on 2023-09-11, 14:07:

Does yours have a NTSC/PAL jumper?

edit: just thinking about this ...if the screenshot from winxp is fine, that means the framebuffer is fine .... so this problem hints at a bad RAMDAC ; trying a different monitor would be a check

I tried 2 different CRTs and a LED monitor as well. Same result. 🙁

If it does the same thing over DVI and VGA, it is not the RAMDAC .

I am guessing that it is either a weird hardware fault OR maybe something to do with a custom VBIOS.

What brand/model is the card and can you share VBIOS dump ?

darry wrote on 2023-09-11, 18:55:

If it does the same thing over DVI and VGA, it is not the RAMDAC .

I am guessing that it is either a weird hardware fault OR maybe something to do with a custom VBIOS.

What brand/model is the card and can you share VBIOS dump ?

Unfortunately I could only find this label on the card. Trying to gain VBIOS dump information on Thursday.

Sapphire 87-5C83

giantclam wrote on 2023-09-11, 14:07:

Does yours have a NTSC/PAL jumper?

edit: just thinking about this ...if the screenshot from winxp is fine, that means the framebuffer is fine .... so this problem hints at a bad RAMDAC ; trying a different monitor would be a check

By the way, there is an NTSC/PAL jumper, but does nothing regarding this problem. 🙁

kalanneni wrote on 2023-09-12, 12:27:

By the way, there is an NTSC/PAL jumper, but does nothing regarding this problem. 🙁

Yeah, likely only touches 'tv-out'.... I've got a jetway 9600XT.... the other thing I was thinking about, is does the 3D unit of the GPU display the same problem (in win98 we're only seeing the 2D unit output)...just curious