Color HCFR Calibration Discussion (Post your calibration files here) - Page 119

I'm usually the one accused of being too picky . But, this seems fine to me. If the Blue were on target then the Cyan would be closer to the line. But it may be hard to get Blue closer.

I would think that a bigger problem would be that hue for Cyan is shifted towards Blue. That might be perceptible.

Is there some other problem with Cyan that you can see with your eyes. Like, possibly, that 75% saturation (which isn't measured here) is undersaturated?

Hi all,
I'm currently working on the CMS of my LG PK550. I'm using Color HCFR and a DTP94-B.
The PK550 provides a general color and tint control and additionally color & tint controls for all primaries and secondaries.

I already studies lots of guides, including the excellent guide written by Tom Huffman. Most guides explain how the correct luminance can be achieved and this works pretty well on my LG. But still I have found no answer what's the best approach to set the correct tint values using Color HCFR? Color HCFR provides no info about the deviation in saturation and hue. Only a combined error for xy.
When changing the tint controls for my primaries I can move the "dot" displayed in the CIE chart along the diagonals of the RGB triangle. But so far I didn't manange to change the distance of the RGB points from the D65 point. According to Tom's guide this is mainly controlled by the saturation. But here's exactly the problem: How can I change the saturation with color/tint controls only?

I'm thankful in advance for any advice. I've attached a screenshot of my Color HCFR measurements.


Best regards,
~nergalix

You may want to check out what I discovered about the the LG CMS in their LCD displays. The color function operates differently depending on which color you are using. It moves the saturation point for less than 100% saturation (as the 100% points for the primaries are fixed), but adjusts the lightness on the secondaries. I suspect that the CMS for their plasma's works the same way.

See here:http://www.avsforum.com/avs-vb/showt...0#post18868120

@Knd
Hi mate, that's a very interesting observation! So far I only did 100% Saturation measurements. I'll run a full sweep when I find some time and check if the behaviour of the LG plasma is similar to your LCD.

Have you found out what effect the main "tint" control has?

Best regards,
~Nergalix

It's been a couple of months since I last worked with the set, so I'm not exactly sure. I suspect that it may rotate the secondaries all together. I just kept the main tint control at 0.

hi,

i've attached a calibration file for my jvc lcos 56fn97, with calibration from a dvdo duo with full cms. the test patterns are on a hard drive and played back via a sageTV hd300 network media player over hdmi.

i set the gamma for 2.2 and, watching xyY in the hcfr software, adjusted x and y to best match (0.313, 0.329) and seem to have achieved a pretty good delta-e.

however, as a calibration novice, i really don't understand
1) why did my gamma come out closer to 1.7/1.8 than 2.2?
2) why is the gamma so bad at 90/100% when the delta-e looked so good?

i'd appreciate any help in understanding what is going on here.
(also attaching a before measurement)

 

after2.zip 4.30078125k . file

 

before.zip 2.2158203125k . file

I am not familiar with the dvdo duo to know what changes it can make. As far as the display goes, if it doesn't have a gamma control then brightness and maybe contrast are typically all you can use to affect gamma. I only looked at gamma, and the before gamma graph is much better than the final gamma. The final gamma graph drops off horribly near white. Your 90% Y and 100% Y are almost equal on the after measurements. I'm not sure what you did, but I'd suggest restarting. Although I'm not familiar with the processor, I'd guess it may be best to setup the display first before moving on to the processor.

my duo has a complete CMS so i was able to adjust x and y for each level from 10% to 100%

i really don't understand how the gamma could be off if the delta-e is under control.................all i can guess is the the Y needs to be adjusted for each level 10%-100% to match the gamma and somehow that got out of wack (i didn't touch Y at all, assuming the program would take care of that automatically since i set 2.2 as my desired gamma in preferences)

The x and y adjustments in your CMS are for the coordinates of the primary RGB and secondary YCM colors on the Rec709 chart.

The white balance settings are in your grayscale, which is the D65 white spot. You adjust R or B (usually leave G alone) to pull your white spot to the x,y of D65, ie .313, .329. Your grayscale looks very good actually.

Your gamma falls at the high end. Maybe you have your contrast setting too high?

My first impression is that you mis-understood the whole process and procedures of calibrating your display. Have you read the following calibration guide:

http://www.curtpalme.com/forum/viewtopic.php?t=10457

You adjust grayscale to D65 white point by balancing the RATIO of Red, Green, and Blue in the range of 10% - 100% white, using the settings related to grayscale adjustment (in most devices it is something like R/G/B contrast/gain, R/G/B brightness/offset, etc.) - but it's definitely not CMS, which is used to modify your color gamut.

You calibrate gamma by adjusting the AMOUNT of Red, Green, and Blue (or Y simply put) in the range of 10% - 90%. Before that, you need to set your Contrast (for 100% white) and Brightness (for 0% black) controls correctly.

When you mentioned "... i set 2.2 as my desired gamma in preferences", I assume you're referring to setting the gamma value of 2.2 into the HCFR software - but it doesn't have anything to do with the setting in your Duo. It only tells the HCFR software to calculate the desired gamma curve for you (shown as the white dotted line in the "Luminance" graph). The target Y values for 10% - 90% are also shown in the last row "gamma Y" (the row below "delta xy") in the "Measures" tab. What you need to do is to adjust the gamma setting in the Duo so that the measured Y values for 10% - 90% match the ones shown in the row "gamma Y". I have no experience on your Duo, so I can't tell you which setting in the Duo you need to tweak; but that's the concept of doing the gamma adjustment.

If you understand what I'm saying, you will easily know that your "before" gamma, although not ideal, is much better than your "after" gamma.

Please try to read the above calibration guide if you haven't.

Agree.. your blue is clipping on the luminance graph. you are most likely clipping at the higher end. Bring your contrast down, or just bring all RGB components at 80 and 90% down.

Your after GAMMA is too "bright". Looking at your Before/After RGB, I am guessing that you are increasing Blue and Red to bring it to 100%. Doing so will likely bring your GAMMA down as you are also increasing "Y", and probably caused clipping at the higher end. Bring your "Green" down first and then play around with Blue/red. Try using cutting as much as possible, instead of increasing. Good luck.

thanks for the feedback folks...i think i will converge here (hopefully soon).

i have read the guide, a couple of times actually, but i often find that reading is different that doing.

i guess that i have gotten the grey scale pretty well under control (by adjusting x, y to 0.313, 0.329, which correct balance of RGB) but have left gamma-Y uncontrolled (which corresponds to total RGB)

after i re-check my contrast (and lower it), my question is in which order should i modify grey scale and gamma-Y?

i have separate settings for xyY in my duo's CMS (color management system) but i was not using the Y control.
should i address xy first, then gamma-Y or the other order?
assuming they interact to some degree, i would expect some iterations to be needed to get both under control?

thanks

Well I don't know anything about your duo, but having two components that can both do various combinations of adjustments of grayscale and color hue, saturation, and presumably luminance seems like a recipe for confusion.

I would set one of them at "default" and do everything with the other one. Sounds like you're more into your duo.

yes, i'm using the duo, which let's me adjust both gamma-Y and x,y and i'm not sure which to do first..........or am i still confused/getting this wrong?

CMS is different than grayscale. CMS in Duo I think is using LHS (luminance, Hue and Saturation), so don't get that confused.

In your Duo, you should be able to either do the typical RGB, or the xyY.
If you use the xyY approach, then I think for each white%, just change xyY to match, from 100% down to 10%.

that is exactly what i am doing.

i think i've found the problem with my gamma and it seems to have nothing to do with what i had expected it to be.

when i adjust the x or y or Y with the duo unit, i entered the user menu, bring up the appropriate thing to adjust and use the up/down key to change the value of x or y or Y.

however, when in adjustment mode, the screen is modified to display, in addition to the window at the chosen IRE, a reading of input and output RGB (in upper left and right corners) as well as a blue and red rectangle showing the current value of the parameter i'm adjusted in the upper middle center of the screen. what i found is that when the menu is visible, it has a big impact on gamma-Y (seems not to have much of an impact on greyscale x,y though).

so the xyY readings i adjust too with the menu on/visible is not the reading i get when i go back and run a complete measurement (where the duo menu/display will be off)

i'm going to follow this up on the evdo duo forum and see if there is an option to adjust xyY without visible menu artifacts (or some other workaround).

Worst case, you might need to make the adjustment, exit the 'adjustment screen", take the measurement and see how it goes, and then re-enter adjustment screen to readjust... This is quite troublesome, but most display had different behaviors with different APL. I also have to exit my PJ menu screen as it somehow affect the reading a little bit as well.

I have the same problem when I calibrate my LG 47LH90. But, I've also found that using a full field (i.e. not a window) makes the problem go away. If your setup allows you to use full field patterns you might try that.

hmm.......i do have access to full field patterns, but i was under the impression that the results wouldn't be accurate unless the window was smaller than some size or another

can anyone comment on viability/accuracy of using a full field pattern for these calibrations?

On my LG 47LH90 measuring on a full field vs on a window gives the same values. I.E. without some menu panel. You could test this yourself. If the full field gives the same values as the windowed pattern then you need to see if the presence of a menu panel changes the values of those obtained on a full field.

sounds like a plan............top of my list of things to do

thanks

The Duo has an independent control for red, green, blue and luminance Y in the grayscale adjustment setting. Yes, you will probably get a different measurement with menu on and with the menu off.

I use the AVS disk as my pattern... get the TV close with the built in TV controls then use the adjusments in the Duo. You may want to do gamma adjustments first then do the RGB and then gamma. If you use a tool lke ChromaPure you can get gamma measurements at the same time as RGB.

In addition to individual controls for red, green and blue pm on the Duo, you have adjusments for x,y, and Y.

I use the apl window patterns for both grayscale and gamut, and I always measure with Duo menu off. It is tedious but it works. Make an adjusment exit the menu, then measure. On the Duo at least with my TV, the trickiest thing to set is the gamma at 90 IRE. Sometimes you have to lower Y at 90, sometimes you lower it at 100. You do need to go back and forth between 90 and 100 until you are certain the settings at either IRE are not effecting the other and you are at the gamma you want.

I am going through the process of calibrating my projector and have a simple question about colorHCFR.

Is it possible you can add the 0%,25%,75% saturation reference points to the CIE diagram? i.e. When I measure 0%-100% saturation for primaries/secondaries I would like to see how my measurements in the %'s other than 100% are lining up.

On checking my gamut on my projector using the % Saturation measures, I found the satuation luminance graph for both green and yellow to show being grossly high in the lower stimulus and sloping down to meet the target at 100% stim.

None of my other graphs show this error (Luminance, Gamma, RGB or CIE). Any ideas as to what the cause might be?

I'm using a Sony 46W3000 LCD TV, and the primaries and secondaries have good delta E and delta luma values with the TV at its Warm2 default setting.

Grayscale calibration gives me near perfect grayscale parameters with an average gamma of 2.20, but now the the delta luma of blue is 24.5%, whereas delta luma of green is negative. Is there anything I can do to fix this? The controls I have are color, hue, and white balance (gain/bias for each of RGB).

 

Luma.zip 5.0224609375k . file

Here are the results after Calibrating my LG 50PK550 last night using a X-rite i1DisplayLt & HCFR software. Let me know what you guys think.

 

HCFR Results.zip 4.583984375k . file

How can I get HCFR to display a specific color or gray level for use with the Unique Free Measure (or even Continuous measures) option?

I have HCFR connected over HDMI to my display and use the "View images" calibration mode, so that HCFR generates the patterns for the calibration.

This works great for grayscale and primary/secondary automated measurements. However there are times when I just want to work with a specific gray level which is not available under the Advance->Test patterns->Display Patterns options.

For instance, let's say I want to focus on calibrating my 70% level. When I use Unique Free Measure, I need HCFR to display a 70% gray level and measure that.

But when I take a free measure HCFR puts up a 0% full field and measures that. The help file says "If the 'display test colors during measures' is checked in General Preferences (which I verified it is) and if the selected generator is View Images (which it is), then the test color is displayed on the screen. However, it doesn't say anywhere how to test it WHAT test color to display.

Am I overlooking something? How do I specify the "test color" for a single measure?

Also under Display Image Parameters I use an Image Size % of 50%. This is because I am calibrating a plasma and if the image is full screen it saturates the diffuser. When I run a Measure Primary / Secondary Colors operation, this works fine. However I need to put up a red or green or blue pattern and just work with that pattern. I can use Display Patterns to put up the color I need to work with, but it puts it up as a full field, whereas I need it to be smaller.

Ideas? Thanks!

Just go to the next chapter until you get to the color or %gray that you want. Then you have 5 min. But if it advances on it's own, you can just go back, chapter by chapter. (Each color or gray% is a "chapter")

I gave it a look, and I couldn't figure it out either.

Are you referring to reading off of a DVD/blu-ray? I think so. I'm trying to use the self-generated patterns in HCFR.