WinDAS White Point Balance guide for Sony Trinitron CRTs

I don't think the instrument could explain your problem. You calibrated with the same instrument you did the post-calibration measurement with, so something other than the instrument is responsible for the discrepancy.

Similarly, just because you didn't use ArgyllCMS for fine tuning wouldn't explain this.

When you're doing the post calibration grayscale test in HCFR, what is the target/reference white point? And what was the color temperature in OSD set to? If they're not the same, then the delta Es are of course going to be high.

As an aside, it's interesting that your monitor (Sony CPD-G400) doesn't have 6500K. Because of this, WinDAS doesn't target the D65 chromaticity for your tube. But I suppose you might be able to hack it by targeting the D65 chromaticity (chromaticity: x = 0.313, y = 0.329), for, say, the 5000K color temperature. That way, when you choose 5000K, you'll actually be getting 6500K (D65 is important as it is the standard used for mastering much content)

holy moly targeting that x and y worked! colors match a lot more to my modern screens, gamma around 1.5-2 looks great no black crush anymore!

i think the issue i had with 30% going wild with readings might've been a hcfr issue? if it was readomg below 8.2ish cd/m2 the rgb and dE levels would 0 out but the x, y and Y number values read fine so maybe a bug with hcfr? or maybe because youtuber guides had set a custom sRGB white point instead of your rec709?
anyway first image shows initial reading with 1.0 gamma, second image at around 2, massive relief thank you.

just curious is there any way to calibrate for 1.0 gamma? in case you're using these kinds of crts with very old hardware that don't have gamma options? I've got a couple other similar sony crts one i wanted to dedicate for dos/amiga or earlier systems with a scandoubler even if the screens a bit overkill for that era

OriginalDan said:

holy moly targeting that x and y worked! colors match a lot more to my modern screens, gamma around 1.5-2 looks great no black crush anymore!

i think the issue i had with 30% going wild with readings might've been a hcfr issue? if it was readomg below 8.2ish cd/m2 the rgb and dE levels would 0 out but the x, y and Y number values read fine so maybe a bug with hcfr? or maybe because youtuber guides had set a custom sRGB white point instead of your rec709?
anyway first image shows initial reading with 1.0 gamma, second image at around 2, massive relief thank you.

just curious is there any way to calibrate for 1.0 gamma? in case you're using these kinds of crts with very old hardware that don't have gamma options? I've got a couple other similar sony crts one i wanted to dedicate for dos/amiga or earlier systems with a scandoubler even if the screens a bit overkill for that era

Click to expand...

Good job.

Not sure what is causing the low light readings to act up in your case. Have you tried following the steps outlined in my guide regarding using HCFR?

Also, sRGB uses the D65 white point.

Not sure what you mean by 1.0 gamma. You rarely calibrate for 1.0 (linear) gamma - the only time I've done that is during grad school where I was doing experiments for visual psychophysics - linear gamma would look awful. Your gamma is probably somewhere between 2.1 and 2.6 depending on how you set your G2.

spacediver said:

Not sure what is causing the low light readings to act up in your case. Have you tried following the steps outlined in my guide regarding using HCFR?
Also, sRGB uses the D65 white point.
Not sure what you mean by 1.0 gamma. You rarely calibrate for 1.0 (linear) gamma - the only time I've done that is during grad school where I was doing experiments for visual psychophysics - linear gamma would look awful. Your gamma is probably somewhere between 2.1 and 2.6 depending on how you set your G2.

Click to expand...

just wanted to see if targeting srgb instead of rec709 would make a difference and it didn't really, attached a photo with the gamma measurement as well.

i just thought that almost all the screens I've got in the past (various samsungs/lgs acer etc) always looked great without touching gamma, albiet most had some colour or usually geometry issues which is partly why i jumped into this, but never needed the gamma up from 1. still a noob to this so just wondering why that is?

ill be moving onto the Dell P1130 soon, first I've got a viewsonic i need to replace the antiglare coating on but i may end up trying a more recent colorimeter just to rule out its not the colormunki, anyone know if the Display Plus/Pro HL/SL or 123 support refresh displays?

OriginalDan said:

just wanted to see if targeting srgb instead of rec709 would make a difference and it didn't really, attached a photo with the gamma measurement as well.

i just thought that almost all the screens I've got in the past (various samsungs/lgs acer etc) always looked great without touching gamma, albiet most had some colour or usually geometry issues which is partly why i jumped into this, but never needed the gamma up from 1. still a noob to this so just wondering why that is?

ill be moving onto the Dell P1130 soon, first I've got a viewsonic i need to replace the antiglare coating on but i may end up trying a more recent colorimeter just to rule out its not the colormunki, anyone know if the Display Plus/Pro HL/SL or 123 support refresh displays?

Click to expand...

Ah you're talking about the slider in the nvidia control panel. That explains the confusion. I leave that at 1 and use argyll to fine tune things (including gamma) post hardware calibration (see my guide for instructions).

Your gamma (i'm assuming the yellow line is gamma, it's been a while since I've calibrated) looks very low - I imagine the image looks very washed out?

What does your gamma look like when the slider is set to 1?

spacediver said:

Ah you're talking about the slider in the nvidia control panel. That explains the confusion. I leave that at 1 and use argyll to fine tune things (including gamma) post hardware calibration (see my guide for instructions).
Your gamma (i'm assuming the yellow line is gamma, it's been a while since I've calibrated) looks very low - I imagine the image looks very washed out?
What does your gamma look like when the slider is set to 1?

Click to expand...

ah right my bad, i didn't use the argyll process as i wanted to set it up for use with original hardware dos/95/98/xp etc which can't run 3d luts, since you mentioned "We need to tweak our luminance function. One very easy way to do this is just to adjust the gamma slider in your video card properties." i went for that, sadly i didn't get screenshots before and after changing gamma settings and since sold the compaq p900 but i will revisit this with my dell p1130s.

the image wasn't washed out, however when the nvidia control panel gamma set to 1 it had crushed blacks until it was up around 2 to 2.5ish then it looked good, side by side with my modern VA LCDs (Lenovo G34w-10 & LG 32UK550-B) looked great and actually found my LG screen being too pink in comparison lol, maybe the low gamma reading from HCFR was the reason behind why my meter struggled to read? strange since i followed each guide method, setting G2 to match the grayscale bars and dailed in the 0 and 30% and 100% so that shouldn't be low?

i did read conflicting information on whether or not to hook up the 5v line on the USB to TTL converter to the windas port, since your pcs usb port should provide 5v already and one of the youtube guides claimed connecting the 5v will elevate the black level, i wonder if that was skewing things as i didn't connect the 5v to the TTL converter during all this, i couldn't find any other mention of this though.
id presume that windas port would have the CRT providing a 5v line to power the original windas external ttl equipment? either way i had no issues with the windas software itself in terms of stability it was super stable throughout.

ill report back later with how the p1130s go and see if i get similar results, windas should be much the same as the fw900 since i think they share flybacks and somewhat similar chassis(?) but ital be a few weeks or months until i get to em, for now I've got a Viewsonic PF790 i need to replace it's antiglare film and i dont think viewsonics had an equivalent windas port?

I'm slowly getting a better idea on how this all works so the help is much appreciated, hopefully my noob fumblings can help others.

OriginalDan said:

ah right my bad, i didn't use the argyll process as i wanted to set it up for use with original hardware dos/95/98/xp etc which can't run 3d luts, since you mentioned "We need to tweak our luminance function. One very easy way to do this is just to adjust the gamma slider in your video card properties." i went for that, sadly i didn't get screenshots before and after changing gamma settings and since sold the compaq p900 but i will revisit this with my dell p1130s.

Click to expand...

The Argyll step in my guide doesn't do a 3d lut, it does three 1d luts.

OriginalDan said:

i did read conflicting information on whether or not to hook up the 5v line on the USB to TTL converter to the windas port, since your pcs usb port should provide 5v already and one of the youtube guides claimed connecting the 5v will elevate the black level, i wonder if that was skewing things as i didn't connect the 5v to the TTL converter during all this, i couldn't find any other mention of this though.
id presume that windas port would have the CRT providing a 5v line to power the original windas external ttl equipment? either way i had no issues with the windas software itself in terms of stability it was super stable throughout.

Click to expand...

My understanding is that the 5v pin is to power the circuitry inside the monitor, and that the USB provides the power to do this.

Hi, followed the steps and ended up with a useful calibration, thanks for you all your good work you put into this guide! There is just one thing nagging at me: it seems the screen is still capable of producing high luminance levels, at least during calibration it can do even 150 cd/m² or so. But at the end of calibration, I end up with 115 cd/m².

Do you know how I can increase luminance to a higher level ? (ps: contrast is at 100 %)

teufelskrapfen said:

Hi, followed the steps and ended up with a useful calibration, thanks for you all your good work you put into this guide! There is just one thing nagging at me: it seems the screen is still capable of producing high luminance levels, at least during calibration it can do even 150 cd/m² or so. But at the end of calibration, I end up with 115 cd/m².

Do you know how I can increase luminance to a higher level ? (ps: contrast is at 100 %)

Click to expand...

Good stuff!

Yes the screen is capable of higher values. There are a couple steps where you're asked to hit a certain luminance - one is 115 (CMAXBMAX) and the other is 95 (CMAXBCENT). I think c and b stand for contrast and brightness, and cent stands for centre value?

I never experimented, but I imagine hitting diff targets at these points will result in a higher luminance after calibration. Perhaps try bumping up the CMAXBMAX to 130 or something.

But be aware that this may age your tube faster.

spacediver said:

Good stuff!

Yes the screen is capable of higher values. There are a couple steps where you're asked to hit a certain luminance - one is 115 (CMAXBMAX) and the other is 95 (CMAXBCENT). I think c and b stand for contrast and brightness, and cent stands for centre value?

I never experimented, but I imagine hitting diff targets at these points will result in a higher luminance after calibration. Perhaps try bumping up the CMAXBMAX to 130 or something.

But be aware that this may age your tube faster.

Click to expand...

Pretty sure if your ABL circuit is doing what it's supposed to it won't let it get beyond a certain level, or it may throw a fault. I don't know - maybe the calibration adjusts the limit.

@spacedriver: thanks yes those are the parameters I needed to raise to boost luminance to a value higher than 115 but unsurprisingly the resulting image was pretty bad I didn't have much time to experiment with this yesterday though so I'll try to redo the whole process again and see if I can strike a balance between higher luminance and decent color calibration results. Otherwise I'll stick to the default targets.

Hello FW900 veterans. I’m about to buy my 6th FW900. I will want to run through WinDAS calibration as per this guide. I’ve done it many many times before, actually brings back fond memories of me trying to get everything right.

SpaceDiver or anyone else. May I please have the original PDF and the test patterns? The link in the OP no longer works.

congrats on the new FW900, hope it's a good one!

The guide, including test patterns, should be in the link in my signature.

spacediver said:

congrats on the new FW900, hope it's a good one!

The guide, including test patterns, should be in the link in my signature.

Click to expand...

Thank you. This FW900 is rather odd. But with some adjustment and maintenance I've made it a bit better.

I was wondering if you'd ever do another tutorial or two on WinDAS. Namely on focus and convergence in the procedure menu.

There is already a guide on dynamic convergence in WinDAS (not in the procedure menu like rest). I've been practicing it, but I feel perhaps the focus and convergence from the procedure menu could yield even greater results.

christpunchers said:

Thank you. This FW900 is rather odd. But with some adjustment and maintenance I've made it a bit better.

I was wondering if you'd ever do another tutorial or two on WinDAS. Namely on focus and convergence in the procedure menu.

There is already a guide on dynamic convergence in WinDAS (not in the procedure menu like rest). I've been practicing it, but I feel perhaps the focus and convergence from the procedure menu could yield even greater results.

Click to expand...

focus and convergence is not something I went deep down the rabbit hole. If you figure out a solid solution though, please share!

spacediver said:

focus and convergence is not something I went deep down the rabbit hole. If you figure out a solid solution though, please share!

Click to expand...

I have been playing around with the Dynamic Convergence procedure a few times. Ordered a special device to make it more objective. Will write perhaps a lite guide once I've ran through it with the device

christpunchers said:

I have been playing around with the Dynamic Convergence procedure a few times. Ordered a special device to make it more objective. Will write perhaps a lite guide once I've ran through it with the device

Click to expand...

That would be very cool! Curious - what kind of device have you ordered? Is it some sort of optical magnification device? (e.g. a loupe)

So on the last few steps of calibrating my Dell P991 using WinDAS in the sRGB section, it asks me to display a "Special WHITE" pattern with an IRE of 48 and 100% of something. Can someone explain to me exactly what it wants me to do during this phase? How do I get HCFR to display a 48 IRE pattern?

huh, that's a new one, never seen that before. I'm guessing it is asking you to show a 48 IRE (which would be RGB = [123, 123, 123], which is a medium grey), and the 100% may just mean full screen. What are the instructions after you load up the image? Is it asking you to meet a particular luminance target?

spacediver said:

huh, that's a new one, never seen that before. I'm guessing it is asking you to show a 48 IRE (which would be RGB = [123, 123, 123], which is a medium grey), and the 100% may just mean full screen. What are the instructions after you load up the image? Is it asking you to meet a particular luminance target?

Click to expand...

This is most likely the correct assumption. And great question.

spacediver said:

huh, that's a new one, never seen that before. I'm guessing it is asking you to show a 48 IRE (which would be RGB = [123, 123, 123], which is a medium grey), and the 100% may just mean full screen. What are the instructions after you load up the image? Is it asking you to meet a particular luminance target?

Click to expand...

I would have to go through the calibration again since I rushed through those last few screens. I know it asks for Special WHITE, Special BLACK and I think Special GREY. The rest of the calibration was perfectly normal aside from the sRGB section.

I probably wouldn't worry too much - I imagine that each mode you're calibrating for is independent. So if you don't plan on using sRGB mode, then it probably doesn't matter what you do during that phase.

Can someone do me favor and check something on their FW900 for me?

I know the default OSD brightness and contrast is 31/90.

But as you all know, there’s like 3 steps within each number range.

So for brightness: is it 31, just one button press shy of 32? Or is it 31, just one press above 30?

And how about contrast?

Please let me know. I have what appears to be a low hour SGI FW900 (FW9011 2002) on my desk. But the brightness and contrast isn’t resetting properly with the physical reset button.

christpunchers said:

Can someone do me favor and check something on their FW900 for me?

I know the default OSD brightness and contrast is 31/90.

But as you all know, there’s like 3 steps within each number range.

So for brightness: is it 31, just one button press shy of 32? Or is it 31, just one press above 30?

And how about contrast?

Please let me know. I have what appears to be a low hour SGI FW900 (FW9011 2002) on my desk. But the brightness and contrast isn’t resetting properly with the physical reset button.

Click to expand...

What do you mean it isn't resetting properly? Can you clarify? I believe they're all 8-bit values internally so make of that what you will.

I bought this P992 CRT monitor last October. Maybe a week later, I performed Windas calibration and convergence adjustment. However, since the adjustment, the monitor has developed a problem: whenever the power is cut off—even just using the power button to turn it off—the OSD becomes locked when I turn it back on, and I can only unlock it by reconnecting to Windas and flashing or exporting the configuration file. But as soon as the power is disconnected again, the problem reappears. Although I don't usually need to adjust the OSD on a CRT monitor very often and put up this issue for a while, I've recently found this quite inconvenient, and it's obviously abnormal. I don't know if this is a config file problem? I have the original .dat file ported with only some tweaking in OSD before, but I'm not sure if I might have corrupted the file or not.

That is strange, I've never encountered this before. Maybe try loading up WinDAS again and choosing procedure/final setting and then exit and see if that fixes it.

spacediver said:

That is strange, I've never encountered this before. Maybe try loading up WinDAS again and choosing procedure/final setting and then exit and see if that fixes it.

Click to expand...

I reloaded a config file then processed this step and the issue is fixed. Thank you so much. I really don’t know how I overlooked this step for so long.

good stuff