The Official ChromaPure thread - Page 8

0.6 is a level of error that is essentially invisible. You can't do better than this.

You display test patterns on the screen. You measure those test patterns with a color analyzer and software. The software tells you the level of error the color analyzer sees relative to what it is supposed to see. You use the TV's controls to make those errors smaller.

This is a very general answer to a very general question.

It is likely that the large difference in Y is only because one is set to read in fL and the other is set to read in cd/m2 (270 cd/m2 = 79 fL). This still shows a discrepancy, but a much, much smaller one.

I think that the ChromaPure 100% stim (x0.299, y0.307) should be
R98
G99
B121 (not B110). Is this is typo?

The 90% reading (x0.296, y0.307) is almost the same, and you have this essentially correct at
R93
G100
B120

Since the white point target is the same for SMPTE-C, Rec. 709, and EBU, any differences in the RGB analysis between the gamuts will be very small. For example, x0.299, y0.307 in SMPTE-C is
R98
G98
B120

and in Rec. 709 it is
R98
G99
B121

which is negligible.

Your chromapure Y values are in cd/m^2 and the calman Y values are in fL. Your contrast is likely too high and is causing a color shift which may be why your 90 and 100 IRE x,y are off. I'd suggest lowering contrast to the 45-50fL range for daytime viewing.

Again, both products have the ability to produce reports. Pick a set of adjustments, run a calibration with the same settings for each product and post them. Otherwise the prudent assumption about differences is operator error.

Checked my notes and I wrote B=110 at 100IRE and B=120 at 90IRE.

Will have to find that setting and change lumination setting to be the same.

Did another calibration today and at 100IRE - Red hit the max adjust point of 50. Turned the BackLight off today since the ChromaPure is not setup for the Enhanced Chroma5 - someone else brought up this point. Still can't understand my getting the Yellow tinge on Calman and Red with ChromaPure. Nothing is changed except switching between Calibration software.. Results of todays calibration: CIE94=0.5 CCT=6494 Gamma=2.2
100IRE R50 G-4 B-40 90IRE R24 G-8 B-39 80IRE R26 G-8 B-41
70IRE R15 G-21 B-38 60IRE R12 G-16 B-34 50IRE R3 G-24 B-34
40IRE R-3 G-22 B-30 30IRE R-6 G-23 B-28 20IRE R-8 G-18 B-23
10IRE R-7 G-12 B-16
One of the other Forum members mentioned turning the colour setting right down to 0 before starting - I haven't heard of this one before.
Backlight is set at 69 (Turned off for this calibration)
Contrast is set at 90 Brightness is at 52 Colour is set at 48 and Tint at 0
Used to have Contrast set at 84 but recent checks on this put it now at 90

It is impossible for x0.299, y0.307 to read B110 and x0.296, y0.307 to read B120. The difference between these two readings is in the red channel. Blue is virtually the same. The level of stimulus doesn't matter.


I really don't know what this means. If you use the backlight during normal viewing, then you should calibrate with it on. It is the same with the Color control.

I agree with a previous poster. Please run calibration reports or take screen shots of the grayscale run and then post them. For the purpose of diagnosing this discrepancy, dE, CCT, and gamma are not relevant. All we need to see is the xy and RGB data.

ChromaPure defaults to CD/m2. The first change I made in options is to set it to FL for easier comparison with most reports which use FL.

Tom, when using ChromaPure with a Spectracal LED modified C5, would Chromapure even see that, or if it did would it make a measurable difference.

The C5 you sent me is not modified but the one p5browne uses is. When he uses it with CalMan vs ChromaPure could that account for the differences he sees ?

You'll have to ask the CalMAN guys to be sure, but I believe that there is no modification to the meter itself. I think that the modifications are at the software level. A vendor is free to create any software-based mode for a meter they like. We are working on precisely this for version 1.3 as a way of improving the Chroma 5's accuracy, which is quite good already, but certainly has room for improvement.

The LED tables are loaded into the C-5 itself. I purchased one recently and use it with an existing CalMAN installation. No other software or data came with the probe.

In the Calman software, when it detects the meter, it comes up as a Chroma5 Enhanced and you select `LCD with LED' instead of `LCD' for monitor selection. So I turned Off the Local Dimming when calibrating with the ChromaPure since it's LCD only.

As Tom said, you should set your display up for calibration in the same manner you intend to use it. Why turn off local dimming since it should improve black level performance ? Whether local dimming is on or off, it is still an LED backlit LCD display. Turning off local dimming is not the same as turning off the backlight.

Even with LD off, it is still an LED lit set so it would not matter.

The reason for having LD off is the effect it would have when using window patterns. I now calibrate with LD on but use full screen patterns so LD never gets an opportunity to mess with the windowed patterns due to the adjacent dimmed areas. This seems to work well for me but rests on the unproven theory that an LD set should be calibrated with LD on.

I just looked at the SDK for the Chroma 5 and, yes, I had forgotten that the C5 includes the ability to load custom calibrations into the EEPROM. However, you would still need software support to call those calibrations. Without that they would be unused.

I'll try later with full screen patterns. But, with field patterns and local dimming on, what I've seen is that luminance INCREASES when a calibration subpanel (e.g. the Red adjustment for one of the IRE levels) is being displayed. The luminance decreases when the calibration subpanel is removed. Toggling local dimming on and off has no effect if there is no calibration subpanel.

Under typical circumstances window patterns are for plasma displays. Any device that can deliver full output should use fields or constant APL windows.

Some displays are just annoying. That's one reason why people buy external video processors.

True that, Tom. I had thought that by "software", you meant an external table or file loaded by the program when the C5 was used.

Yes the Chroma5 has the ability to store 4 internal calibration tables. From the factory they only use two of them. In our calibrations lab we add support for LCD/LED and UHP lamped projectors. We called it the Enhanced Chroma5.

CalMAN has three types of calibration tables for the meter itself. Profiling where a user uses a reference and target meter to create a calibration matrix for a specific display. External for meters like the Spyder3, i1Pro and ColorMunki where they don't have the ability to store internal tables. The external tables are downloaded from our website on demand as needed. And finally internal for meters that have onboard storage of tables like the Chroma5, K-10 and Hubble.

To be able to compare CalMAN and ChromaPure numbers you need to be using the same internal tables within the Chroma5 itself. So with CalMAN you need to select 'LCD Direct View' so both applications are using the same table. Otherwise you are comparing apples to oranges or in this case LCD' vs. LCD/LED'.

I had some concerns about this so I fiddled a bit. You also need to have CalMAN in the default state. There are options that will cause CalMAN and ChromaPure to engage in spirited debate.

For the record, I can think of no reason--so long as both are using the same modes for the Chroma 5--why the two programs would report different numbers for the same meter on the same display using the same test patterns.

They should be the same, within reasonable tolerances.

That is why I have asked p5browne to post calibration reports. I am particularly skeptical of some of the RGB values that were reported.

Agreed, both CalMAN and ChromaPure use the same XYZ/xyY numbers from the meter and then run various formulas on the raw meter data to display the xyY, RGB, DeltaE, etc... So those numbers should be very close if not the same.

Just in case it wasn't clear, there are a lot of knobs on CalMAN. Some of those effect the presentation of data. If you can't readily compare the raw meter output you need to be sure the cooked numbers are subject to identical transforms. I think it's sufficient to use the defaults and select the same dE calculation.

Being an old fart, will see what I can do in regards to last night's calibration report. I still much prefer the flatlined Green PQ (100IRE to 10IRE - set to Gamma 2.22 and the Reds and Blues left at Zero - despite the critics out there that I don't do things by the book) to either of the Calman (Yellow Tinge) and ChromaPure (Red Tinge) full calibrations. Trying to compare Green only between Calman and ChromaPure - Calman will calibrate the Green only - ChromaPure, you have to do the Complete individual calibration on every IRE from 100IRE down to 10IRE - won't let you do any individual IRE. You also have to do the startup part of 10IRE to 100IRE or it won't let you calibrate anything. With Calman I can do it any way I feel like. The Green settings were not even close to each other between Calman and ChromaPure.

Oh yeah now I remember my methods are Unorthodox! Guys, I'm just trying to get a full Calibration that looks great. I'll get on the reports tonight when I get home. It will be my New thing us seniors have to learn everyday!

Sorry, I really don't comprehend what you are trying to say here. Once you initialize the meter, you can begin with any module you like and calibrate whatever level of stimulus you like. For example, just open the White Balance module, display the desired level of stimulus, and then you can calibrate one level, multiple levels, or all of the levels. It is entirely up to you.

The Pre and Post-Calibration Grayscale modules are not for the calibration process at all--that is, changing values to meet targets. They are for reporting the initial state of the display and then the success of your efforts once the calibration is complete.

Sorry, the Before is the After - BackLight 69 Contrast 90 Brightness 52 Colour 48 Tint 0 White Balance Warm Gamma Low BL Low

ChromaPure Calibration Report

Technician: ChromaPure User
Company: ChromaPure Video Calibration Software


Greyscale Before Greyscale After
% Stimulous X Y CCT dE % Stimulous X Y CCT dE
20 % 0.313 0.331 6,491 0.9 20 % 0.000 0.000 2,021 NaN
30 % 0.313 0.329 6,506 0.1 30 % 0.000 0.000 2,021 NaN
40 % 0.312 0.330 6,518 0.8 40 % 0.000 0.000 2,021 NaN
50 % 0.313 0.330 6,462 0.5 50 % 0.000 0.000 2,021 NaN
60 % 0.313 0.330 6,485 0.4 60 % 0.000 0.000 2,021 NaN
70 % 0.313 0.329 6,503 0.3 70 % 0.000 0.000 2,021 NaN
80 % 0.313 0.329 6,461 0.4 80 % 0.000 0.000 2,021 NaN
90 % 0.313 0.329 6,476 0.2 90 % 0.000 0.000 2,021 NaN
100 % 0.312 0.329 6,548 0.4 100 % 0.000 0.000 2,021 NaN

2/8/2010 7:11:27 PM Page 1 of 2


ChromaPure Calibration Report

Technician: ChromaPure User
Company: ChromaPure Video Calibration Software


Color Gamut Before Color Gamut Before
Color X Y Yy Delta E Color X Y Yy Delta E
Red 0.000 0.000 0.000 NaN Red 0.000 0.000 0.000 NaN
Green 0.000 0.000 0.000 NaN Green 0.000 0.000 0.000 NaN
Blue 0.000 0.000 0.000 NaN Blue 0.000 0.000 0.000 NaN
Yellow 0.000 0.000 0.000 NaN Yellow 0.000 0.000 0.000 NaN
Cyan 0.000 0.000 0.000 NaN Cyan 0.000 0.000 0.000 NaN
Magenta 0.000 0.000 0.000 NaN Magenta 0.000 0.000 0.000 NaN
White 0.000 0.000 0.000 NaN White 0.000 0.000 0.000 NaN

2/8/2010 7:11:27 PM Page 2 of 2

ChromaPure Calibration Report

Technician: ChromaPure User
Company: ChromaPure Video Calibration Software

Before gives the values with Red, green and Blue IREs set at Zero - The After Green Only Calibration

Greyscale Before Greyscale After
% Stimulous X Y CCT dE % Stimulous X Y CCT dE
20 % 0.297 0.321 7,520 5.5 20 % 0.299 0.297 7,832 9.9
30 % 0.296 0.321 7,582 5.8 30 % 0.297 0.304 7,844 8.4
40 % 0.297 0.320 7,561 5.7 40 % 0.298 0.311 7,642 6.8
50 % 0.296 0.321 7,636 6.0 50 % 0.296 0.315 7,673 6.2
60 % 0.297 0.321 7,527 5.6 60 % 0.297 0.322 7,512 5.6
70 % 0.297 0.319 7,564 5.7 70 % 0.297 0.323 7,475 5.5
80 % 0.297 0.317 7,585 5.8 80 % 0.297 0.325 7,473 5.8
90 % 0.299 0.317 7,450 5.5 90 % 0.299 0.330 7,258 5.9
100 % 0.302 0.319 7,261 4.8 100 % 0.302 0.330 7,087 5.1

I am still not sure what this data refers to. You have posted two sets of grayscale data. The first set has one grayscale run and a bunch of zeros. The second set has two grayscale runs--I assume that this is Before/After or After/Before.

Stimulus	x	y	R	G	B	CIE94
20	0.313	0.331	99%	100%	99%	1.2
30	0.313	0.329	100%	100%	100%	0.2
40	0.312	0.330	99%	100%	100%	0.9
50	0.313	0.330	100%	100%	99%	0.6
60	0.313	0.330	100%	100%	99%	0.6
70	0.313	0.329	100%	100%	100%	0.2
80	0.313	0.329	100%	100%	100%	0.2
90	0.313	0.329	100%	100%	100%	0.2
100	0.312	0.329	99%	100%	100%	0.4

% Stimulus	x	y	R	G	B	CIE94	x	y	R	G	B	CIE94
20	0.297	0.321	87%	103%	111%	7.5	0.299	0.297	105%	96%	129%	18.2
30	0.296	0.321	86%	103%	111%	8.0	0.297	0.304	97%	98%	124%	13.6
40	0.297	0.320	87%	103%	111%	7.5	0.298	0.311	94%	100%	118%	9.9
50	0.296	0.321	86%	103%	111%	8.0	0.296	0.315	89%	102%	115%	8.8
60	0.297	0.321	87%	103%	111%	7.5	0.297	0.322	86%	103%	110%	7.5
70	0.297	0.319	88%	102%	112%	7.6	0.297	0.323	86%	103%	109%	7.6
80	0.297	0.317	89%	102%	114%	7.9	0.297	0.325	84%	104%	108%	7.9
90	0.299	0.317	92%	101%	113%	7.3	0.299	0.330	84%	104%	103%	8.2
100	0.302	0.319	94%	101%	110%	5.9	0.302	0.330	87%	104%	103%	6.5

Other than some change on the low end, the second set of data shows very little change between the two grayscale runs. They both show too much blue and too little red. The first single set is as close to perfect as a grayscale run is going to get.

Not sure what conclusions, if any, should be drawn from this.

The second set of tables were to show the Before, what the system looked like with with all IREs and and CMS Zeroed that was missing in the first table.
I know the results look A-OK, but, I've ended the exact same with the Calman calibration as well. I'm just trying to figure out why, when nothing has changed, my PQ is Red tinged with ChromaPure and Yellow with Calman? But the values all point to the ideal PQ? Will have to find what setting is affecting both calibration final out comes.

Because you are using two different calibration tables within the Chroma5, you will get two different results. Within CalMAN use the LCD Direct View display type and you will get the same results as ChromaPure.