Basic Guide to Color Calibration using a CMS (updated and enhanced) - Page 10

I sure get different results when I put in 100% gray compared to 75% gray when I use that calculator:

Rec.709
GrayY=75%
Yr=0.159
Yg=0.536
Yb=0.054
Yy=0.696
Yc=0.591
Ym=0.214

Shouldn't we use these values when comparing and adjusting to the 75% gray pattern?

Thanks for the clarification. Sorry for the confusioin.

Tom - Actually, the two methods report different errors. The 1994 method is not uniformly lower. This is why we included two different flavors of dE in v2 of CalMAN. The advantage of the 1994 method is that it gives you a dC* and a dH* component so you have an idea of where the error lies (saturation or hue). The 2000 method then expands this to include an interaction effect between saturation and hue. Interpreting that one, as I've said before, can be a challenge.

Bill

We think its pretty cool, too. Combine this data with a "bullseye" chart, and it is pretty easy to set-up a CMS or color decoder in a certain version of a certain software product that just had its next release go into open beta.

Bill

Tom -

With the MitsDiamond series, specific to **833 series (I have the 65")..Does one use the Perfect Color alone to set the decoder (using your method) and Perfect Tint to set Primary and secondaries points on the CIE (using your method)?..In addition and a bit off top but still related, does Mits have only GGL, GRL, GBL to calibrate greyscale (gains?), is there not 2 sets of Red, Green, and Blue of Bias,Gains to set it?? I did not see them in the SM..

That must be Colorfacts?

you have no idea how much THANK YOUS I could say now...

As a follow-up to using this guide with the Mitsubishi Diamond XX833 Series, I was wondering if there are certain settings (like Picture Mode, Color Temp, Deep Field Imager, Video Noise, Video Mute, Film Mode, and SharpEdge) that routinely should be at a particular setting versus personal preference?

And, while we are discussing the Mits, would anyone with calibration experience care to comment on the relative merits of the CMS system in the Mits Diamond WD-XX833 vs. the Sony KDS-XXA3000. Is one CMS system inherently better than the other (e.g., more controls, higher accuracy, ease of use, etc.)? Thanks much.

Another reason to take the Sony off my list....

But this does leave me with another question... Can the Sony KDS-A3000 be "calilbrated" in regard to gray scale and color gamut by some other means?

Oh, you Swedes and your humor!

Bill, are you trying to cause an international incident by calling a Norwegian a Swede???

How do you get a Swede to laugh on New Year's Eve?
Answer: Tell a joke on Christmas Eve!

Does anybody here have experience on using Sony's RCP? I tried it and I can move all colors on the CIE diagram using:
- range
- position
- color
- tint

Is it a Color Replacement System or what? Shoud I try to use it to move all RGBCYM points to the correct xy coordinates using HCRF, any benefits/drawbacks here?

So I understand that with the XX833 we leave Color and Tint at their default level and adjust using only Perfect Color and Perfect Tint. Should we adjust those after adjusting for Gray Scale (I understand we might have to go through all adjustments multiple times but I presume it's best for the first go-through to adjust one versus the other initially)? And is there a preference to adjusting Perfect Color before/after Perfect Tint.

Finally, can someone explain in a little more detail adjusting the Diamond's GGL, GRL, GBL, CMG, CMR, and CMB.

Sorry for the newb questions but, well, that's what I am.

You may find this thread on our forum of interest (free registration required):
http://www.datapopuli.com/forums/viewtopic.php?t=194

Bill

Thanks for the link. Explains a bit. Confuses a bit I'm guessing that GGL, GRL, GBL, CMG, CMR, and CMB all affect Grey Scale and that I'm just going to have to play around with them until I figure things out.

yes you can leave them at default of 31 for Color and Tint and use PERFECT COLOR for color decoder and Perfect Tint for adjusting CIE points for primary/secondary..

Tom is correct, the GGL, GRL GBL are for the bright end of the greyscale and CMG CMR and CMB are the dark end of the greyscale..I was able to adjust these for LOW temperature settings..the weird thing is the GAINS parameters has alot more flexibility than the BIAS..but that should not matter as really the sets are not drastically off - so its more than enough (at least not for mine)..I am not sure but I do always hop in the SM while in low temperature mode - not sure if going in to SM while in HIGH temperature mode will give you diff parameters..

Honestly I did COLOR decoding first and also followed by CIE points and readjusted after my greyscale..going back and forth..I adjusted numerous times..trying to get my secondary and primaries running through the D65 point using the CIE u'v' view mode.

Your grayscale changes the relative balance of your primaries to make your secondaries. Once you change the grayscale, you are going to have to go back and validate the color decoder, so you might as well skip that step and do it all at once later.

Bill

It looks like the spreadsheet is only valid for Rec709. What about SMPTE? It looks like the reference Lab would be required.

I went into the spreadsheet and used your formulas to convert xyY to Lab for the 6 colors. I got the following:

SMPTE Lab??

L, a, b
53.168, 74.186, 67.638
87.046, -80.772, 83.958
35.400, 72.058, -103.412
91.144, -43.199, -14.202
61.568, 90.385, -59.075
96.575, -21.510, 95.642


Rec709 Lab (original in Tom's spreadsheet)
L, a, b,
53.168, 80.008, 67.134
87.631, -85.610, 82.631
32.258, 79.124, -107.765
91.098, -47.970, -13.986
60.338, 98.456, -60.922
97.146, -21.598, 94.221

So my plan is to put these into the spreadsheet and also insert the xyY of SMPTE for the colors. Does that make sense or I am I being too clever by half?

Excellent. Thanks for all your hard work. I like the bar graphs, too.

Most people use Luv for emissive and Lab for absorptive (e.g., textiles, paper). Beyond that, they work in conceptually similar fashions.