eeColor Processor - CalMAN - 3D LUTs - Page 10

Yes the last LUT I ran with 21 point GS is using eecolor, however the two problems I am seeing is when using the Mini as a pattern generator and its 21 step grayscale pattern, there are some small hint's of color at a few of the steps. Also using S&M disc it seems to show color clipping, of-course this could be caused by my Oppo 95, however the color clipping never showed up when I used the Mini as the processor for a 125 LUT/21 GS.

ss

Come on Buzz you know what it means, 255.

All I am saying is comparing the results of a Lumagen Radiance 3D to eecolor and a LUT calibration. I am not saying there is a problem only trying to lessen the effect in those squares .

ss

u need to have the eeColor in your chain at all times... we found a very tiny distortion with brightness, so you need to setup your display (brightness / contrast / etc) with the eeColor in the chain (with a UNITY LUT)...

- M

Thanks Mike, as usual from you that is what I needed to know.

ss

If I remember correctly, on the VT50 it comes down to one click difference in brightness - that might differ on other TV sets.

- M

Yes I have dropped brightness down one click from non eecolor use. But still its something more than just that for my video chain, all be it part of the solution.

On the bright side, I am not really seeing any more motion jitter / stuttering on my VT50, after I started using eecolor and the larger LUT's.
I didn't reply to your post about this because I didn't do extensive testing, It just happen, and I use 24p direct in at 96Hz and source direct on my Oppo for Blu Rays. .

ss

yeah same here... need to do some testing to see if any of the LUT boxes introduce or boost motion jitter... I did experience it on the eeColor... will report back...

- M

When I used my Radiance Mini 3D and a 125 point LUT I did get a little jitter and motion blur, but not with a 3375 LUT using eecolor, at-least not yet. I calibrate the same way for both the Mini and eecolor, no enhancements and 96Hz. As you may know I use calman. and there 1:1 mapping software in the eecolor.
I would say it could be that are VT50's handle jitter/.motion blur differently except for what i saw using my Mini.

iow I don't have a clue to why the jitter/.motion blur seems to be gone now, I am just guessing.

btw, I only use the VT50 for Blu Ray's.

Looking forward to your report.

ss

You may already know, but the VT50 has a precision regulator for brightness in the service menu, named SUB-BRT. I always use it to fine tune the brightness setting on my TV.

That is something I have not seem much talked about. I also have access in the SM to Sub Contrast and Sub Bright, but never really knew how to use these to my advantage. On my set up now, I have brightness at the default of 50 as that gives me the best on screen material look compared to 51 when watching blu-rays. But this also does show that I am only seeing 19 on the black clipping screen ( possibly 18 a little )

If I raise my brightness to 51, I then can see 17 flash a bit, but the real world pictures look a little washed out and not as good. I think from what I am seeing on the AVSHD disk, that 51 is the correct setting, if I could make it a bit darker than it is, but not quite as dark as 50 would be.

Is there any how to's on what patterns to use and things to look for when adhusting the subs in the SM?

Just finished a 3375 LUT, took about 3 hours 20 min, 6600+ reads. Starting to get a feel for the larger LUT.



ss
Edited by sillysally - 4/28/13 at 2:11am

This is done using quick checker and about 24 hours after i ran the above calibration. The only difference is I forgot to turn on reference mode for my Mini when I ran QC, but my mini has no calibrations in it

.

ss

Thanks Buzz,

Yes that is what should happen with reset, but I am not sure about the patterns. And then there is the output settings to consider, but the output settings (4:4:4,:709 1080P@24), where the same for the calibration and the QA.
Anyway I have everything backed up on my HD just encase, using the Radiance's back up tool.

ss

I haven't posted here in a bit. I recently just performed another 3375pt cube on my Kuro. I used a c6 profiled against a colormunki spectro. I used the latest CalMan load. The results came out pretty good.....

^^^^^

On the charts it sure looks a lot better than the 17^3 (4913) profile I just ran using Lightspace.
btw your 4 hour time using your meters is pretty good.



ss

How does the ColorChecker SG measure on the LS 17^3 cube?

Not sure what you are asking.

I use CMB 5 (latest beta) to run the test, after I have installed the LUT in eecolor that I just made using LS. I use the Klein still at the same placement that it was at for the 17^3 profile.
Use the Mini 3D in reference mode for the patterns, then via a 6' HDMI cable out to the eecolor set at LUT 3 (That's where the newly made LUT was sent to from TurVue software).
So everything is done the same, except I used a different laptop to run CM's ColorChecker using Teds charts.

btw, I had my VT50 set to color space wide and used 2.4 for the Gamma setting, I did the 17^3 profile in LS and after the profile was done and I made the LUT I set it at 2.16 Gamma.

ss

Added.
If you want to see what QC looks like on a 17^3 LS LUT here is a link. http://www.avsforum.com/t/1437739/lightspace-cms-now-supports-lumagen-eecolor-3d-lut-4-all/180#post_23295202

The differences here are that when I made the LUT in LS I use a 2.2 Gamma setting, but I used the same 2.4 Gamma setting.
The other difference is I used a small window in that profile (link).
Edited by sillysally - 5/16/13 at 6:57am

when using ColorChecker in CalMAN change the points measured to SG with 100% sats and without black. This will measure over 100 points with 14 flesh tones.

He used C6 with 1 Sample per read whats why the total time is reduced.

When I was using C6 with my KURO, to get acceptable repeatability (not dE repeatability but xyY repeatability) I was using C6 with 5 Samples as a mimimum and LLH @ 10 Samples per read for measurements below 30 cd/m2, after tons of testing to my setup. I'm talking for November-December 2012 Builds of CalMAN.
Edited by ConnecTEDDD - 5/16/13 at 7:34am

you are correct Ted, i used 1 sample. But take a look at my screenshots below. Why would i use more samples? IMO, its near a point that it starts to become a waste of time. The difference between 1 sample and 20 samples is still WELL below the visual threshold(at least in my experiences) and i would never see the difference in moving content.

I will come back with more data's when I will have the time.

C6's response is not the same with 1 Sample measuring samething with 10cdm/2 vs. 20 cd/m2 vs. 90 cd/m2 etc....

The differencies are larger in xyY. Ok it small difference for our eyes but thats why we got the meters, to measure these small differencies better

More Stable Measurements will give you more accurate calculated corrections, we are talking about 3375 corrections.
Edited by ConnecTEDDD - 5/16/13 at 11:12am

hmmm ok, fair enough. Even if there are larger xyY differences between the samples, i still don't think the differences will be large enough to make any sort of visual difference with content. Yes, i'm sure more samples will give me better CalMan charts, but the added time just doesn't make sense to me. I can see if the dE differences between the points i measured were larger, i might try a different sampling. But it just doesn't make sense to be. All IMO, of course......

i would love to see your xyY differences when you have more time.

I have plans to make a C6 vs. iD3 OEM repeatability/speed comparison, but not have enought free time to measure 4-5 different technology displays since i have daily access to over 100 different models @ my work

@1forsnow
Seems to me you have got the best of both worlds, a great PQ verified by almost perfect dE Charts.



Not certain of the history but surely dE values were introduced as a measurement of the 'fit for purpose' use of the Display?

Certainly we can find the meters that are more and more accurate but have we got Displays that are good enough to constantly output that extremely accurate calibration?