Basic Guide to Color Calibration using a CMS (updated and enhanced) - Page 13

Depends on what your actual luminance is that corresponds to 0.63%. At >100 nits (plasmas, LCDS, etc.) 0.0063 x 100=0.63 nits is easily measured by the i1pro as long as you average multiple measurements and perform the dark current offset correction. For FP set-ups at half that value it's still doable but your error bars are increasing quickly so you're probably better off moving to a direct reading using the CA 813. Tom, have you ever done a head-to-head between the 813 and the i1pro, I'd like to know in absolute units where they start to significantly diverge, say around the 5% level below 1 nit (0.05 nits). Previous measurements posted by umr with a PR-670 as reference indicated that below about (0.1 nits) the error bars > 10%. I think the manufacturer rates it to meet error requirements down to 0.2 nits.

Hi,

Does anyone have or know where to get the SD color definitions xyY for PAL (EU). I think the SMPTE-C is for the NTSC system only.

Regards,

Kostas

PAL/SECAM (EBU)
R: 0.6400 0.3300 0.222021
G: 0.2900 0.6000 0.706645
B: 0.1500 0.0600 0.071334

Sorry if I have missed this somewhere here (I am pretty green on this calibration stuff) but when I use HCFR in xyY mode my Y calcuations don't even come close to the ref Y. I get numbers like 13.198 for red instead of around the 0.213 ref number. I am sure I am its something simple that I am missing.

The 0.213 reference is a ratio, not Y. You want to divide Yred by Ywhite and get 0.213

Ah that makes sense. Thanks!

Be sure to be consistent with the percent. In other words, if you are measuring 75 percent R, G, and B, use 75 percent white (75% amplitude/ IRE).

Here are my numbers on a Mits 73833 using a Spyder II (I know) and HCFR. This is from Toms 1994 sheet. Most of my dE variance seems to come from Saturation which the mits has no control of. Any ideas on correcting it?



Red 0.642 0.332 0.211 6.4% 30.3% 63.3% 0.9
Green 0.300 0.622 0.736 6.9% 71.5% 21.6% 2.2
Blue 0.152 0.065 0.071 2.1% 81.0% 17.0% 1.4
Cyan 0.203 0.327 0.800 4.0% 95.5% 0.5% 4.4
Magenta 0.323 0.140 0.290 2.3% 91.8% 5.9% 2.9
Yellow 0.428 0.522 0.935 1.8% 90.3% 7.9% 2.8

Thanks Zoyd. What about White(D65) xyY and the secondaries values? I know there is a relation between the primaries and the secondaries (ie a percentage or substraction) but I don't remember it

Regards,

kopa13

D65 is D65, it doesn't change with gamut and white Y=1.0. Secondary Y calculate from primary Y (C=B+G, etc.).

Is it right that we are most sensitive to +Green in an image. Or am I wrong here. It's worse to have some +green in the grayscale tracing than a over saturated green in the gamut?

I'm following your reasoning. The middle scale of the gray's is most importat to get right, escpecially with Red, next after is green n the lows. Last place high part of the grayscale.

Could you comment on this calibration report on a Sony A3000? If there is no CMS, then does Sony have some other approach? How was this done?

http://www.avsforum.com/avs-vb/showt...0#post12008400

umr also said in response to a question about how to duplicate the calibration,

"You would need my $24,000 spectroradiometer to do it. It is pointless for me to tell you how. Measuring colors on displays like this that support xvYCC color space is incredibly challenging and requires the highest level instrument to do a credible job."
__________________
W. Jeff Meier

Mr. Meier is held in high esteem over on the other thread, as are you on this one. Is there an explanation to what appears to be an absolute difference in your opinions?

It's also a deceptive statement because it implies he did move the primaries (because he thinks that's what you think) when as Tom pointed out all he did was balance the grayscale. If he had been able to adjust the primaries they would lie precisely on top of their targets. Grayscale tracking is not as demanding a calibration task for a probe as CMS calibration because white SPD's (spectral power distributions) are better behaved than isolated primaries.

btw, the improvement in primary color accuracy, which is not included in the report, went from
dE RGB:10.9,5.1,3.5 to 9.2,2.2,2.8 which is marginal at best.

Oh, I see. Very interesting. No magic involved. The primary colors for at least that particular set were pretty good to begin with.

Thank you.

I love this thread. It's kinda like reading Scientific American, fascinating but pretty much incomprehensible for us noobs.

Tom,

Great thread for us beginners. I have Sharp 12K MKII and Colorfacts 6.0 Pro. I'm still learning to us it. Anyway I have read through the whole thread and I get the jist of the procedures. I have one question. In the Sharp CMS you have Gain, Offset and Gamma. WHEN and HOW do I adjust the gamma setting. I'm coming over from CRTs and normally we would adjust our RGB Gain pots to try to achieve an even cutoff point for all three tubes. This was more or less setting Gamma. If Gamma pots was to low, you would lose shadow detail. If to high, you would wash out shadow detail and have a slight glow on screen. Then we went on into calibrating the gains and cuts. Is it the same for the DLP gamma setting on my Sharp?

Thanks

Andre

ALL,

Does anyone know anything about the TOPCON BM-7 Luminance Colorimeter? Any info would be great.

Thanks All

Andre

Tom, a naive question but what exactly do you mean by the terms "top end" and "bottom end"...for example, does "top end" mean "add white" while "bottom end" means "add black"?

Thanks.

Tom:

Appreciate the response but I am guess that I am a little "brain dead" this moring...must be result of too much travel lately.

While I understand that (i) gray, as in the colour, is the combinations of various amount of RGB and (ii) 0 IRE to 100 IRE is the same gray (read combination of RGB) with difference brightness/luminace I still am puzzled as to the meaning.understanding of:

-- brighter part of the gray scale; and
-- darker part of the gray scale.

With that is a possible nivices interpretation that "brighter part of the gray scale" means brighter combinations/portion of RGB while the "darker part of thh gray scale" means the darker combinations/portion of of RGB where brighter and darker refer to chromaticity (i.e. exclude luminace).

TIA.

Tom:

So, in other words, the gain/offset controls add and subtract the amount of gray with high luminance (i.e. 80-100 IRE*)/low luminance (i.e. 10-30 IRE*) to a given IRE window (say, 50 IRE) so that the combination (i.e. 50IRE to gain + offest) results in the desired color temperature (say, 6500K) for the given window....do I now get it?

* The only difference between IRE windows is the brightness/luminance...that is, all IRE windows have the same chromaticity but different brightness/luminance.

TIA.

This stuff is complicated and my heads feels as though it is going to explode...that said, here goes another kick at the can...


I understand that when performing a gray scale calibration that for each IRE window one want to achieve, among other things, a given chromaticity/color and a given luminance/brightness such that the resulting color temperature is 6,500K -- right?

I also understand that, asuming that the luminance/brightness is correct, that the chromaticity is adjusted by adding/subtracting amounts of red, green and blue -- right?

I do not undersatnd the use of the terms "bottom end" and "top end"...is it as simple as:

a) offsets are used to adjust the chromaticity for the lower IRE windows; and
b) gains are used to adjust the chromaticity for the higher IRE windows?

If yes, then where is the magic cutoff point and why?

If no, then I am still lost...not a comfortable feeling.



I wil tackle this one next...that said, I understand (or, at least think I do)that each level of 10-90 IRE should have a prescribed brightness as it is the combination of color/chromaticity and brightness/luminance that produce the 6,500K...that said, I am lost as to how gamma fits in!



Also an issue for another day....

Thanks for everything...much appreciated..