WinDAS White Point Balance guide for Sony Trinitron CRTs

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
I'm not sure if that monitor has a service manual or service menu. If you can hunt one down that may help.If you're somehow able to adjust the G2, you might be able to achieve what you want. Maybe post in a separate thread with the CRT mode and service manual in the title
 

xykreilon

Weaksauce
Joined
Jul 22, 2020
Messages
81
I'm not sure if that monitor has a service manual or service menu. If you can hunt one down that may help.If you're somehow able to adjust the G2, you might be able to achieve what you want. Maybe post in a separate thread with the CRT mode and service manual in the title
Okay, I'll see if I can find one.
Also, very interesting, the G2 stuff. So, if I've wrapped my head around it right, adjusting the G2 adjusts what the minimum brightness of black on a CRT is via undervolting, without compromising the gamma somehow? I would appreciate an explanation on this a lot.
 

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
Okay, I'll see if I can find one.
Also, very interesting, the G2 stuff. So, if I've wrapped my head around it right, adjusting the G2 adjusts what the minimum brightness of black on a CRT is via undervolting, without compromising the gamma somehow? I would appreciate an explanation on this a lot.

It does affect gamma. If you set it too low, the image will be crushed (gamma too high). This can be compensated by a software calibration after the hardware calibration, but this works best if you have a 10 bit DAC (nvidia cards with analog out have this).

It's been a while since I studied this stuff, and even back then there were gaps in my understanding. I think in WinDAS, the other steps also affect gamma, but not sure exactly how.
 

xykreilon

Weaksauce
Joined
Jul 22, 2020
Messages
81
It does affect gamma. If you set it too low, the image will be crushed (gamma too high). This can be compensated by a software calibration after the hardware calibration, but this works best if you have a 10 bit DAC (nvidia cards with analog out have this).

It's been a while since I studied this stuff, and even back then there were gaps in my understanding. I think in WinDAS, the other steps also affect gamma, but not sure exactly how.
Okay, interesting.
 

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
also, you might be able to adjust the voltages (G2, drive), directly with a screwdriver if your tube supports it. I think some tubes have a potentiometer for G2 on the flyback that you can adjust with a screwdiver.

Do your due diligence before opening up the tube though (safety, etc.)
 

xykreilon

Weaksauce
Joined
Jul 22, 2020
Messages
81
Update-
I calibrated my 751 set to half brightness and full contrast in the OSD and am thoroughly impressed. Crazy intensely good black levels, while being bright enough to still be usable in a well lit room.
I also got way better color performance this time- 92% sRGB coverage vs. my previous 84%!
Screenshot_20200729_092521.png

Though, that weird dip in the lime green area wasn't there in my last calibration. Oh well, that's probably a compromise made in order to make the rest of the colors perform well.
With all this in mind, I'm actually decently comfortable doing some photo and video editing on this over 2 decade old monitor. I say that coming from a laptop monitor that performs this well after calibration:
Screenshot_20200729_092735.png
 

Boohyaka

n00b
Joined
Aug 4, 2020
Messages
17
Many thanks for the process and it's a success!

A few things I figured out that may help others:

1- After the first WPB, reaching almost perfect values, it did save the settings in spite of my initial feeling. The only step I wasn't able to properly do was the very last one (sRGB). It asks you to reach a Y=80 value with a "CONTRAST" slider, and with the slider to the max (255) I was only able to reach 72. I did the Final Setting step, and.....everything looked like shit, in spite of perfect values at every single step. Is it because of that last sRGB step? I was annoyed and couldn't understand what went wrong. So I reloaded my EEPROM backup I did before the WPB, and tried from scratch again...and I reached exactly the same shit result (including the sRGB step).
So this time I decided to keep those new settings and start a new WPB straightaway on top of it. Again, hitting all close to perfect values, reached the final sRGB step...and this time it was totally different! I could properly reach the Y=80 with the slider around 200. I did the Final Setting again, saved it, and... bingo! Everything looks damn perfect. I'm not sure if that makes sense to be honest. For me, the iterative nature of the process was to always try to get slightly closer to perfection, I didn't expect a first run to end up all off, then the next one perfect, it's pretty counter-intuitive to me. In any case, if it happens to you, just do another round or two on top of each other.

2- The "verification" step from the procedure on the main PC with HCFR, where it says "select View Images from the generator dropdown" is wrong (maybe HCFR changed since it was written?). This is how I did it:
- Open HCFR, go File->New
- "Select a generator from the list": pick Automatic and next, select your sensor in the list (i1DisplayPro in my case), "create a new meter correction file"
- In the main video, in the "View" section, "Gray Scale" should be selected already, if not, pick it up in the dropdown. Then press the "Go" button on the right-hand side, a popup will ask "Run gray scales on 11 values?" and press yes. Your screen will go fullscreen and 11 different shades of gray will automatically be displayed and measured by the sensor. When it's done, you'll be back at the main view, with the "deltaE" table colored and the RGB/gamma graph, pretty similar to spacediver's procedure. Hopefully, your deltaE will be green in the whole range :)

3- If you're doing the procedure on an older computer (I used XP on a Pentium3 myself), PowerStrip is extremely useful to view real-time screen timings and validate settings (gamma ramp, LUT, etc..)

4- Finally, I am attaching the templates I adapted for my 1600x1200 screen, including a i1DisplayPro centering template, if that's useful for someone.

Many thanks!
 

Attachments

  • Templates_1600.zip
    64.6 KB · Views: 0

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
Thanks for this valuable information and the templates. I'll update the guide in my signature with the new HCFR flow (it may have been due to older version of HCFR like you suggested, I'll have to confirm), and will include your templates.
 

NBAPwns13

n00b
Joined
Aug 22, 2020
Messages
8
Made an account just so that someone here can verify this for me.
Does changing the contrast and brightness prior to calibration screw up the whole WPB process? Is it mandatory to have contrast and brightness set to default before beginning the calibration process? I have a CPD-G520, which is a CR1 Chassis. I set the contrast and brightness to 50 and 50. (If someone can verify me the default contrast and brightness I would appreciate that. I'm guessing its Contrast: 85 and Brightness: 50? Since it changes to that value after calibration).

Problems during calibration:
For all 3: 9300, 6500, and 5300, I cannot hit a (Cont Min) minimum contrast for any of them. The slider is either very close to value of 255. Maxing out to 255 isn't able to reach the target given on Windas. This is also relevant for the last step SRGB. Using a white pattern, i'm unable to reach the proper luminance value even after having the slider maxed out.


SRGB Problem.PNG


Disgusting results after calibration:
 

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
Made an account just so that someone here can verify this for me.
Does changing the contrast and brightness prior to calibration screw up the whole WPB process? Is it mandatory to have contrast and brightness set to default before beginning the calibration process? I have a CPD-G520, which is a CR1 Chassis. I set the contrast and brightness to 50 and 50. (If someone can verify me the default contrast and brightness I would appreciate that. I'm guessing its Contrast: 85 and Brightness: 50? Since it changes to that value after calibration).

I don't believe the contrast brightness settings before calibration have an impact.
Problems during calibration:
For all 3: 9300, 6500, and 5300, I cannot hit a (Cont Min) minimum contrast for any of them. The slider is either very close to value of 255. Maxing out to 255 isn't able to reach the target given on Windas. This is also relevant for the last step SRGB. Using a white pattern, i'm unable to reach the proper luminance value even after having the slider maxed out.

In your first image, the target is 82 cd/m^2 (nits), yet your real time luminance value shows 104.77 nits, and the slider in WinDAS is at max. Why aren't you lowering the slider so you hit the target value?
 

NBAPwns13

n00b
Joined
Aug 22, 2020
Messages
8
I don't believe the contrast brightness settings before calibration have an impact.


In your first image, the target is 82 cd/m^2 (nits), yet your real time luminance value shows 104.77 nits, and the slider in WinDAS is at max. Why aren't you lowering the slider so you hit the target value?

If I'm not mistaken moving the slider to the right decreases luminance. If I decrease and move the slider to the left it actually increases the luminance. I'm going to have to double check again as I'm at work and report later tonight
 

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
If I'm not mistaken moving the slider to the right decreases luminance. If I decrease and move the slider to the left it actually increases the luminance. I'm going to have to double check again as I'm at work and report later tonight

Moving slider to right increases luminance.

If the opposite is happening, something is seriously wrong with your tube, or your copy of WinDAS, or your colorimeter, or HCFR.
 

NBAPwns13

n00b
Joined
Aug 22, 2020
Messages
8
So i'd like to report that I am still getting the same issue even with contrast and brightness set to their default values.
And Yes, for the final step, slider moving to the right decreases luminance.

On my first attempt on doing the WPB I was able to achieve an amazing grayscale. After being unsatisfied with the generated LUT i redid the whole WPB process and now im getting screwed values. (I remember changing around the settings in "Expert" color mode since it has access to gain/bias settings. Other than that after reverting back to an old .dat, I am unable to achieve a good grayscale as before).
I might just have to do an image restoration and see what happens from there.

EDIT: Image restoration didn't do anything.
Question: 1. In Windas, does anything save from the previous profile when loading up an older save file? Or will it be overwritten?(Gonna have to load my original dat and start from scratch again...)
2. Is ArgyllCMS driver required for the i1Display pro? Because i've read around and it is not required. Is it plug and play? or do i need to manually install the drivers? (I'm guessing it is installed correctly since It shows it's correct name under Displaycal and it has all its spectral corrections under the dropdown menu).


SRGB 0.PNG
SRGB 255.PNG


trash 2.PNG


Sharing my grayscale from my first WPB attempt.
Grayscale.png
 
Last edited:

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
So i'd like to report that I am still getting the same issue even with contrast and brightness set to their default values.
And Yes, for the final step, slider moving to the right decreases luminance.

that's kinda bizarre. And you can visibly tell that the screen is getting dimmer when the colorimeter reports a decrease in luminance, right?

It might not be an issue anyway, since I think the last step is for sRGB mode which we don't use (at least I don't).

Another question: Did you ensure your video LUT was reset before doing the WPB?

And have you been adjusting the reference targets in HCFR when going through the three main color temp modes in WinDAS? (so it starts with 9300k, and then moves to 6500 I think, and you need to set the reference targets accordingly in HCFR so that you can interpret the RGB color cylinders correctly.



Question: 1. In Windas, does anything save from the previous profile when loading up an older save file? Or will it be overwritten?(Gonna have to load my original dat and start from scratch again...)

Not sure, I don't think so. Any I've found that backing up my dat and then loading it doesn't restore the tube to the settings when I saved the dat. I think the FW900 has some buggy firmware around that.


2. Is ArgyllCMS driver required for the i1Display pro? Because i've read around and it is not required. Is it plug and play? or do i need to manually install the drivers? (I'm guessing it is installed correctly since It shows it's correct name under Displaycal and it has all its spectral corrections under the dropdown menu).[/quote\

I think it may be required depending upon the version of the colorimeter you have. But it's been a while since I went down that rabbit hole. I think HCFR uses the argyll drivers for the i1display pro, but not sure if it automatically uses those.
 

NBAPwns13

n00b
Joined
Aug 22, 2020
Messages
8
that's kinda bizarre. And you can visibly tell that the screen is getting dimmer when the colorimeter reports a decrease in luminance, right?

It might not be an issue anyway, since I think the last step is for sRGB mode which we don't use (at least I don't).

Another question: Did you ensure your video LUT was reset before doing the WPB?

And have you been adjusting the reference targets in HCFR when going through the three main color temp modes in WinDAS? (so it starts with 9300k, and then moves to 6500 I think, and you need to set the reference targets accordingly in HCFR so that you can interpret the RGB color cylinders correctly.





Not sure, I don't think so. Any I've found that backing up my dat and then loading it doesn't restore the tube to the settings when I saved the dat. I think the FW900 has some buggy firmware around that.

1. I found the solution. Displaying a full screen IRE pattern image, especially white causes the monitor to automatically decrease in brightness, screwing with the colorimeter readings.
Solution: I'm not sure if it was in the guide but changing the "Image area" under image parameters in HCFR to 10% will prevent the monitor from decreasing brightness and ensuring accurate readings.

2. Yes I went through the whole calibration from 93 to 65 and 50k.

3. Yes I made sure to disable/reset any LUTs before calibrating. And made sure default values were set in nvidia control panel.

4. In regards to the .dat. I think you have to finalize settings before saving the dat. So that after you load up that .dat your settings will be intact. I'll have to try again but when not finalizing settings before backing up a dat and loading the same dat again, my convergence settings i've set on the OSD gets messed up.
EDIT: Confirmed. You do have to finalize settings before saving a dat or else some settings wont save. Would need someone to confirm this as well just in case.

Well after trying out #1 from above, I was able to get it close to my first attempt.
Question: How can I correct 10 and 20 IRE? I'm not very knowledgeable about calibration but I would like those to be flat as well. Does that mean i'd have to rerun the WPB calibration again? Also for gamma. But i shouldn't have to worry about that since it will be corrected using a LUT.

Capture.PNG
 

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
Looks like you're honing in, gj.

See what the measurements look like after an argyll lut correction, you'll probably be able to get it dialed in nicely.

As for how to improve the low end... on the one hand, your delta Es for 10 is very good, despite the blue channel being out of balance, so it may not be an issue.

Surprised that you weren't able to get a reading for 0. Did you find that with low light readings, the colorimeter took a while to measure? (if so, that's a good thing, it means it's in adaptive mode - a small bell is ringing in my head telling me that you may need the argyll drivers to get adaptive mode to work, but I could be wrong).
 

NBAPwns13

n00b
Joined
Aug 22, 2020
Messages
8
Looks like you're honing in, gj.

See what the measurements look like after an argyll lut correction, you'll probably be able to get it dialed in nicely.

As for how to improve the low end... on the one hand, your delta Es for 10 is very good, despite the blue channel being out of balance, so it may not be an issue.

Surprised that you weren't able to get a reading for 0. Did you find that with low light readings, the colorimeter took a while to measure? (if so, that's a good thing, it means it's in adaptive mode - a small bell is ringing in my head telling me that you may need the argyll drivers to get adaptive mode to work, but I could be wrong).

Installing Argyll allowed me to use dispwin and dispcal command prompts.
Here is my final calibration. Took me a long time to figure all this out so, although not perfect I'm satisfied. Anything that I could potentially correct let me know. I chose 2.4 gamma as my target will try 2.2 and compare which i like.
Question: I've created a .cal file. Would I be able to create an ICC or icm profile using these command prompts? There is a program called idisplay profiler which i would like to use in conjunction with the .cal i created, since I have another profile (ICC or icm i think) under that program for another monitor.

1598383862300.png
 
Last edited:

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
Looking pretty good! How's the image quality?

I'm sure you could bundle the information in the cal file into the icc file, but not really understanding your use case.

Why not just stick with the cal file you've just created? I also sometimes create a separate cal file for 2.4 and 2.2, and switch between them depending on viewing material. I don't ever use icc files. The only relevant thing inside the icc/icm profile for our purposes is the vcgt tag which contains the same sort of information as that inside the cal file.
 

NBAPwns13

n00b
Joined
Aug 22, 2020
Messages
8
Looking pretty good! How's the image quality?

I'm sure you could bundle the information in the cal file into the icc file, but not really understanding your use case.

Why not just stick with the cal file you've just created? I also sometimes create a separate cal file for 2.4 and 2.2, and switch between them depending on viewing material. I don't ever use icc files. The only relevant thing inside the icc/icm profile for our purposes is the vcgt tag which contains the same sort of information as that inside the cal file.

theres something called DisplayCal profiler that I use. I created an ICC profile anyways so it was all good. Which is actually from the screenshot above
 

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
Ok that should provide similar results, though I always get much tighter results than you when I use the command line approach (not sure if this is because of the command line approach, or something else).
 
Joined
May 22, 2014
Messages
35
Hi everyone, I'm getting into doing the WPB on a DELL P1130 (aparently it's a rebadge of the Sony CPD-G520 mentioned in this thread) I've had in storage which i last used about 10 years ago.. If anyone has any tips or tricks in calibrating it, please let me know! :)
 
Last edited:
Joined
May 22, 2014
Messages
35
Ok so thanks to the OP for the amazing guide! i managed to figure out WinDAS and the instructions, thanks to the help of a few posts in here with updates to the original post, figuring out the final step with "view images" that i couldnt find..i did my first calibration but after the SRGB thing my settings got all borked so I'll be trying 2 calibrations back to back this time in hopes of it turning out alright, now that i have the hang of this. :)

Also near the end spacediver you mentioned that we should have Contrast at 90 and Brightness at 31?? is that what we Trinitron folks SHOULD have set it to before we started calibrating? I didn't see it mentioned before the calibration in the OP. Or is it supposed to be set to 85 and 50 like default? which one is the right one to calibrate at for my P1130 please? thanks!
 
Last edited:

spacediver

2[H]4U
Joined
Mar 14, 2013
Messages
2,621
Also near the end spacediver you mentioned that we should have Contrast at 90 and Brightness at 31?? is that what we Trinitron folks SHOULD have set it to before we started calibrating? I didn't see it mentioned before the calibration in the OP. Or is it supposed to be set to 85 and 50 like default? which one is the right one to calibrate at for my P1130 please? thanks!

I don't think it matters what you set those to before calibration. The 31/90 is what WinDAS will set the monitor to after calibration (that's what I meant by "should be"). But I believe these values are for the FW900, and that those values may differ depending on tube.
 

jbltecnicspro

Supreme [H]ardness
Joined
Aug 18, 2006
Messages
6,021
I don't think it matters what you set those to before calibration. The 31/90 is what WinDAS will set the monitor to after calibration (that's what I meant by "should be"). But I believe these values are for the FW900, and that those values may differ depending on tube.

F520 and Artisan are set to 85 Contrast and 50 Bright after calibration.
 
Joined
May 22, 2014
Messages
35
I don't think it matters what you set those to before calibration. The 31/90 is what WinDAS will set the monitor to after calibration (that's what I meant by "should be"). But I believe these values are for the FW900, and that those values may differ depending on tube.
ah, i understand. Thank you. By the way i rebooted the monitor and it seemed to forget all it's settings? :( Do i need to use the "Load to Set" function in WinDAS to load the settings onto the EEPROM? to save the data from RAM into the memory of the monitor?

-Also, I distinctly remember my P1130 remembering the OSD custom resizing i had done for all of my custom widescreen resolutions, but after doing the WPB.. now when i set a resolution- switch from it and then switch back- the CRT acts like it forgot the custom resizing i had done in the OSD? Anyone know about this, or know a fix? It's really quite frustrating.

And after re-calibrating, i ended up with a G2 of 108 for it to be pure it basically looked turned OFF when in a dark room now. :) and i have 18.5k hours on the P1130, is that a good G2 for that many tube hours? I believe initially it was set to 147
 
Last edited:
Top