or Connect
AVS › AVS Forum › Display Devices › Display Calibration › Basic Guide to Color Calibration using a CMS (updated and enhanced)
New Posts  All Forums:Forum Nav:

Basic Guide to Color Calibration using a CMS (updated and enhanced) - Page 64

post #1891 of 1936
Not yet but I'll do it now. :-)
post #1892 of 1936
Quote:
Originally Posted by prme19 View Post

Do you mean the backlight?.. I will try that then.. Thanks

hello
I have not been posting here for many months New job, moving to new country. Also sold my ES model and bought an F.

In the past, I made posts about contrast and brightness setting. Contrast ratio and therefore minimum blackness level are important. In order to optimize this, in Movie mode I set contrast to 100 and even increased white balance which influences 100IRE as much as possible. This maximized contrast without clipping. Brightness setting was also minimized without clipping offblacks.

With Standard mode these settings would cause clipping. So I had to limit contrast setting to around 80 to avoid clipping. But a small amount of clipping could be tolerated if it would improve contrast ratio.

Then I used the 10p white balance setting to get a BT1186 gamma.

I adjusted CMS so that dE at the different saturations were not showing a large error. So I did not optimize for one saturation. Same method as 10k.

After all this, Movie mode and Standard mode showed almost the same result. (Green saturation points showed a twisty curve in Standard mode).

Backlight is set according to room light conditions. Going from 20 for a bright room to a very low value for a dark room.

All this is from calibrations done 6months ago, and written from memory. I hope to start soon with calibrations of the 55F8000.
post #1893 of 1936
Quote:
Originally Posted by turboman123 View Post

hello
I have not been posting here for many months New job, moving to new country. Also sold my ES model and bought an F.

In the past, I made posts about contrast and brightness setting. Contrast ratio and therefore minimum blackness level are important. In order to optimize this, in Movie mode I set contrast to 100 and even increased white balance which influences 100IRE as much as possible. This maximized contrast without clipping. Brightness setting was also minimized without clipping offblacks.

With Standard mode these settings would cause clipping. So I had to limit contrast setting to around 80 to avoid clipping. But a small amount of clipping could be tolerated if it would improve contrast ratio.

Then I used the 10p white balance setting to get a BT1186 gamma.

I adjusted CMS so that dE at the different saturations were not showing a large error. So I did not optimize for one saturation. Same method as 10k.

After all this, Movie mode and Standard mode showed almost the same result. (Green saturation points showed a twisty curve in Standard mode).

Backlight is set according to room light conditions. Going from 20 for a bright room to a very low value for a dark room.

All this is from calibrations done 6months ago, and written from memory. I hope to start soon with calibrations of the 55F8000.

Hey Turboman123,

Pls correct if I'm wrong, so you're saying I should start with 100 contrast in movie mode, rather than 95 as suggested by Zoyd? as also suggested by Zoyd, I then adjusted my backlight setting to achieve 30-40 ftL. with this, I got better dE readings compared to using just the contrast setting before. I can try starting with 100 to see which has better results.

In standard mode, doing the same process above, my contrast is at 73 with backlight at 8 and brightness at 52.
post #1894 of 1936
Quote:
Originally Posted by prme19 View Post

Hey Turboman123,

Pls correct if I'm wrong, so you're saying I should start with 100 contrast in movie mode, rather than 95 as suggested by Zoyd? as also suggested by Zoyd, I then adjusted my backlight setting to achieve 30-40 ftL. with this, I got better dE readings compared to using just the contrast setting before. I can try starting with 100 to see which has better results.

In standard mode, doing the same process above, my contrast is at 73 with backlight at 8 and brightness at 52.

hello prme19
Contrast 95 is maybe good as a general suggestion. I don't argue with zoyd: he is one of my calibration gods.
But on the ES8000 in Movie mode, you can go higher without clipping. You can even increase the RGB gain in the white balance (which is the same as increasing contrast).
All this will increase contrast ratio.
However, these settings are specific to the ES8000 in Movie mode. Therefore, it would be best discussed in the ES8000 thread in the LCD panel forum, rather than in this general calibration forum.
In the ES8000 forum , this has been discussed in the past by 10k, myself and some others.
Also brightness 52 is most likely too high, clipping offblacks (both Movie and Standard). Generally brightness 45 was found to be correct.
See my settings and results in my link below.
post #1895 of 1936
Use a black clipping pattern like the one on the AVS HD 709 disk. 16 invisible 17 barely visible. I would not use Standard mode. Movie Mode is considered the most accurate out of the box for samsungs. I had a D6500 plasma and started with 95 Contrast , it is not normally suggested to go higher then this. If you are calibrating with an accurate meter then set your light output to a comfortable level , no pinking of the whites, no clipping and no eye strain, use this guideline to calibrate your contrast. Go from there.
post #1896 of 1936
i am a noob when it comes to calibrating TV's admit it. althought i've had quite a bit of experience calibrating monitors. the HCFR guide has me a bit confused.

i was able to follow it all the way up to the CMS section, where it talked about adjusting Tint, Saturation, and CMS control. is the guide saying that if your TV has CMS control, you don't have to use Tint, and Saturation to calibrate the colors? OR is it saying if your TV does have CMS control, you'll have to use Tint, Saturation, AND CMS control. it's really confusing.

http://www.curtpalme.com/forum/viewtopic.php?p=117708&sid=c016b0c5c5d037f8a8caccf90fdb95aa#117708
post #1897 of 1936
If you don't have a cms, use general settings "color" control (in general, brightness of all colors - primaries and secondaries...) and general "tint" control (hue of all the secondaries...)

if you have a cms, use it (hue/saturation/luminance of primaries rgb and secondaries ycm) and don't touch general controls (color and tint). you can eventually use it if cms is broken...
Edited by realzven - 2/7/14 at 9:28am
post #1898 of 1936
Quote:
Originally Posted by realzven View Post

If you don't have a cms, use general settings "color" control (in general, brightness of all colors - primature and secondaries...) and general "tint" control (hue of all the secondaries...)

if you have a cms, use it (hue/saturation/luminance of primaries rgb and secondaries ycm) and don't touch general controls (color and tint). you can eventually use it if cms is broken...

this is the TV i have. http://reviews.lcdtvbuyingguide.com/samsung-lcd-tv/samsung-es8000-picturesettings.html

so are you saying that i should leave the first column on the left (see link) where it says "color" and "tint" unchanged and use the 2nd to the last column on the right under "Color Space" to adjust all the primary and secondary colors?

if so, is there a good work flow to do it because the guide for HCFR is really confusing, it tells you to use the general "color" and "tint", and then tells you to go into CMS,
post #1899 of 1936
Use the same guide but skip the color/tint part
post #1900 of 1936
Quote:
Originally Posted by realzven View Post

Use the same guide but skip the color/tint part

ha, once you said that it makes sense now. smile.gif thank you!
post #1901 of 1936
also, this might be a dumb question, since i've only calibrated monitors before, and never TV. wouldn't it be easier, since if our TV's are connected to HTPC and playback from an HTPC, wouldn't it be easier to just install the Xrite software on the HTPC and let the program itself do the calibration and let the video card driver do the work and create a color profile that loads everytime you boot up. it just seems might be easier than doing it manually by punching in the buttons on the remote and in different TV menus. as long as the backlight and contrast level is to the user's liking, and don't clip, wouldn't that be easier? sorry for the dumb question, again, i've never calibrated TV's before
post #1902 of 1936
Problem with htpc calibration is are you sure that your cg output is acurate ? Pc level 0-255 vs video level 16-235 ? YCbCr mode available ?

If you watch movies with a BLURAY player you have to check if anything is wrong (sanity check)

calibrate with pc and with the player and compare (attention with color space, mapping, cliping ...)
post #1903 of 1936
Quote:
Originally Posted by realzven View Post

Problem with htpc calibration is are you sure that your cg output is acurate ? Pc level 0-255 vs video level 16-235 ? YCbCr mode available ?

If you watch movies with a BLURAY player you have to check if anything is wrong (sanity check)

calibrate with pc and with the player and compare (attention with color space, mapping, cliping ...)

i use media center, and LAV Video/Audio splitter for all movies, blu-ray, and tv, so as far as output goes, it's pretty consistant, all through HDMI bitstream losless. since they're all through LAV Filter and Hardware aceleration at 0-255 color space RGB. and i never had any clipping issues on the gray scale. i am just curious that's all. of course i'll be calibrating through my TV, but i was just wondering.
post #1904 of 1936
Calibrating via changing graphic card's gamma curves is not a good idea, IMHO, especially if the TV is considerably off -target...you can get banding and other problems...
post #1905 of 1936
Hey you guys, i got the Colormunki meter today, thanks to amazon that got here in a day. i wasn't expecting that. i decided on the Colormunki because it's basically the same meter as i1Display Pro, but since i am using HCFR anyway for the TV.

anyway, here's my first attempt at calibrating the standard mode using the regular white balance (2 pt). the reason i only used 2pt white balance as oppose to 10pt, because my TV's micro dimming feature is only available in standard mode, and not available in Movie mode, which has the 10pt white balance option. anyway here are the results.









i accidentally deleted the BEFORE reading. so here is the AFTER. but the BEFORE was pretty bad, it was something like 10~20 Delta E.

but as you can see, i got it down to less than 3 Delta E, i was pretty proud of myself for being the first try.

one thing i am confused tho, the CIE gamma. i followed the HCFR's instruction and adjusted the CMS so all the primary and secondary colors line up as best as i could. but i used the same meter to calibrate my workstation apple monitor, everything well fine with the Xrite software calibration, but i ran HCFR and measured my monitor (not the tv), and the CIE graph shows that my primary colors and some secondary colors did not get lined up by the software to be within the reference boxes. i know that sRGB and REC 709 are suppose to be the same in terms of CIE gamma, but this makes me wonder, wouldn't it be better to have all the primary colors outside of the box? because earlier today when i was calibrating the TV, when i got to the CIE graph and CMS section, i remember HCFR measured my TV's GREEN and RED were beyond the reference box, but i followed the instruction in the HCFR guide and moved those color points to be within the reference box. i just don't know what to trust.
post #1906 of 1936
Monitor calibration as performed by profiling software such as X-rite or Argyll doesn't include or do CMS gamut calibration, just white balance/gamma...however, it creates a profile of the monitor that color managed applicatons can use to simulate the target colorspace (say, Rec709 or sRGB) i.e. transform the colors in software...you need CMS in your monitor as well to calibrate in the same way as TV...

as an experiment, I used Adobe After Effect's color management to employ my monitor's profile in doing colorspace transformartion then did the measurements with HCFR and primaries and secondaries ended up more-less spot-on when passed throuh the profile...
post #1907 of 1936
Quote:
Originally Posted by mytbyte View Post

Monitor calibration as performed by profiling software such as X-rite or Argyll doesn't include or do CMS gamut calibration, just white balance/gamma...however, it creates a profile of the monitor that color managed applicatons can use to simulate the target colorspace (say, Rec709 or sRGB) i.e. transform the colors in software...you need CMS in your monitor as well to calibrate in the same way as TV...

as an experiment, I used Adobe After Effect's color management to employ my monitor's profile in doing colorspace transformartion then did the measurements with HCFR and primaries and secondaries ended up more-less spot-on when passed throuh the profile...

thanks for the explanation, that makes sense now.

although, should i be worried that my Xrite software profile calibration software doesn't adjust the primary and secondary colors to fit inside those reference color boxes? the reason i ask is because the HCFR guide said when calibrating tv, we should use CMS to get those colors inside the box. i am just confused if i should have my monitor's colors inside those references boxes, or leave them outside, and if i should have my TV's colors outside as well? the HCFR guide made it seem that if you have your colors outside of the reference boxes your image colors will look over saturated, does this mean on my monitor colors will look "under" saturated than they should be since they're outside of the reference boxes?
post #1908 of 1936
...you should absolutely bring primaries and secondaries within reference on your TV because you can, but you obviously can't do that with a monitor if it doesn't have a CMS...so primaries and secondaries are what they are on your monitor but if they are not far from reference, it's not that big a problem, especially if they are only more saturated but the right hue - white balance is more important...if your monitor is really wide gamut like AdobeRGB, it probably has some kind of sRGB emulation built in that you should use...knowing the make and model of the monitor would help...
Edited by mytbyte - 2/9/14 at 7:56am
post #1909 of 1936
Quote:
Originally Posted by mytbyte View Post

...you should absolutely bring primaries and secondaries within reference on your TV because you can, but you obviously can't do that with a monitor if it doesn't have a CMS...so primaries and secondaries are what they are on your monitor but if they are not far from reference, it's not that big a problem, especially if they are only more saturated but the right hue - white balance is more important...if your monitor is really wide gamut like AdobeRGB, it probably has some kind of sRGB emulation built in that you should use...knowing the make and model of the monitor would help...

hey mytbyte, really thanks for the explanation. never knew color management was this confusing.

to answer your question, i have a apple cinema display 27". i think it's the White LED, same as my TV i believe. samsung 60ES8000.

here's a snap shot of my cinema display after the Xrite softtware profile calibration. i dont know if that's TOO far out of the reference point or not.


also, with HCFR, i picked "white LED" for my samsung 60ES8000 under display type, should i have left it at the default "non refresh display". because i kind of assumed that my LED tv is white LED.
post #1910 of 1936
that's not at all far off, it probably exceeds the gamut a bit so that the rest of color's saturation can track close to reference....you can do saturation sweeps to check where the 0%-75% points end up...the Samsung is white LED most likely, even if it's a top model...
post #1911 of 1936
Hey mytbyte, for my samsung 60ES8000 TV, what are your thoughts on adding a software profiler calibration on top of the calibration i've already done with HCFR. if you refer to the results i posted above for the TV, althought it looks quite good, less than 3 dE, do you think i should install the Xrite profiler software and do a software profile calibration on top of that ? since my setup is pretty much connected to a HTPC, and all sources are played back through hardware acceleration through Media Center (broadcast tv, blu-ray, mkv, and what not).

do you think there'll be any down side?
post #1912 of 1936
I understood you posted the Cinema display results and was very surprised that it's so close out-of-the-box...well...I'd avoid changing the graphics card tone curves for output to the TV (even though X-rite supports this) since it can introduce banding...Your TV's calibration controls are now the ultimate power in the Universe...I suggest you use them...exclusively...biggrin.gif
post #1913 of 1936
Quote:
Originally Posted by mytbyte View Post

that's not at all far off, it probably exceeds the gamut a bit so that the rest of color's saturation can track close to reference....you can do saturation sweeps to check where the 0%-75% points end up...the Samsung is white LED most likely, even if it's a top model...

do you think it's necessary to do a full primary/secondary saturation sweep and calibrate it with CMS? or is it enough that i just did primary/secondary 100% full saturation calibrations. the HCFR guide didn't mention if full scale sweep calibration is needed, only mentioned to use CMS to calibrate full 100% saturations.

i'll try to run a full saturation reading for all colors when i get home tonight, but i am "assuming" that if 100% primary saturation lines up to the 100% saturation reference point, the rest "should" pretty much line up right? or at least closer to reference than pre-calibration. oh... i am referring to the Samsung TV btw.

thanks so much for your valuable input. smile.gif
post #1914 of 1936
That's the catch...most TVs are not that linear and you can have perfect primaries/secondaries but other saturations/luminaces can be very off...that's why you should sometimes calibrate to 75% and recheck, it is more favourable to have 75% down close to reference than the 100% saturations because most of content, including skin tones, "happens" below 75% saturation...you need to make compromises, and the only way to know in which way the colors go when you make the changes, is to do and re-do full saturation sweeps...
post #1915 of 1936
Does anyone recall where the chart showing the various skin tones with their RGB values are located?
post #1916 of 1936
Quote:
Originally Posted by mytbyte View Post

That's the catch...most TVs are not that linear and you can have perfect primaries/secondaries but other saturations/luminaces can be very off...that's why you should sometimes calibrate to 75% and recheck, it is more favourable to have 75% down close to reference than the 100% saturations because most of content, including skin tones, "happens" below 75% saturation...you need to make compromises, and the only way to know in which way the colors go when you make the changes, is to do and re-do full saturation sweeps...

good info. i did a full sweep last night and sure enough. some of the 75%/50% were pretty off despite that their 100% were spot on. for example, Green 75% and below were way off, and so were Magenta. Red was a bit off, and so were yellow. but red and yellow weren't too bad. surprisingly although Blue's 100% was more off than the rest of the 100%, but its 75% and lower were the most spot on compare to the rest. so using CMS, again i went back in. i gotta say, it's a bit tricky trying to adjust secondary colors with RGB only lol. but i was able to get RED, Magenta, Yellow, and Blue all lined up preeetttyyy good. however, i wasn't able to get Green and Cyan any better. the thing with my Green, if i have 75% and below lined up to the reference point, not only 100% is way over to the right, it'll actually clip a large portion of the REC 709 CIE chart on the left, so i thought instead of clipping a large potion of the pie, i settled for a off green for 75% and below. here's the pic of the CIE after i adjusted all the primary and secondary 75% and below.




and just for kicks, i know that i am calibrating the "standard mode", which has the micro dimming, but i accidentally erased the pre-calibration data, but here's a chart of "movie" mode (without micro dimming), which is supposedly slightly more accurate than "standard" mode. but just to show you how off the original out of factory gamma coverage were.



i think i am going to settle for what i got for now. i might go back in later in the future to fine tune it some more, but for now, i am tired, and the wife is tired of me paying more attention to the TV than her.
post #1917 of 1936
so just while i thought i was done. i read somewhere on this forum from another user that adjusting Green in white balance might yield better overall accuracy on the gray scale. but correct me if i am wrong, but didn't the HCFR guide said Green is reference, and adjusting RGB GainGreen will alter the contrast level, and RGB Offset Green will alter the overall brightness, and therefore should not touch those.

as of right now, my blue is just a bit high from 50~100% IRE, and a bit low on 0~50% IRE. do you think i should venture into changing the green for both RGB Gain and RGB offset? what are some of the pros and cons if i do that. i just want to check before i start messing with it farther.
post #1918 of 1936
Thread Starter 
Quote:
Originally Posted by howzz1854 View Post

so just while i thought i was done. i read somewhere on this forum from another user that adjusting Green in white balance might yield better overall accuracy on the gray scale. but correct me if i am wrong, but didn't the HCFR guide said Green is reference, and adjusting RGB GainGreen will alter the contrast level, and RGB Offset Green will alter the overall brightness, and therefore should not touch those.

as of right now, my blue is just a bit high from 50~100% IRE, and a bit low on 0~50% IRE. do you think i should venture into changing the green for both RGB Gain and RGB offset? what are some of the pros and cons if i do that. i just want to check before i start messing with it farther.
You should always avoid adjusting green when working on white balance for the simple reason that it has a profound effect on gamma. It is best to stick with red and blue. Adjusting green is only necessary at 100%, where it may have to be lowered along with blue when, for example, red is too low and cannot be raised any further because it is already maxed out.
post #1919 of 1936
Speaking of white balance... I'm curious, where in dark films/normal content do you normally see 10% greyscale? As I can't seem to find any obvious blue tint thats in the 8000k range...
post #1920 of 1936
Quote:
Originally Posted by TomHuffman View Post

You should always avoid adjusting green when working on white balance for the simple reason that it has a profound effect on gamma. It is best to stick with red and blue. Adjusting green is only necessary at 100%, where it may have to be lowered along with blue when, for example, red is too low and cannot be raised any further because it is already maxed out.

this is good info to have. i think i fall into the latter category.

accordingly to my earlier results seen above, the RGB balance was slightly heavy on the blue side from 50~100% IRE, and the Blue in my RGB Gain was already lowered to "0", and couldn't bring it down more. so last night just for the hell of it, i raised Green and Red gain simultaneously, and that was enough to level out with the excess Blue on the top end. and to compensate, i lowered the overall contrast just a bit and rechecked all Black/White clipping APL patterns to ensure that nothing was clipped. as a result, i was shocked that i was able to flatten out my RGB Balance AND Gamma line even more. of course those settings shifted the CIE a bit, so i had to go back into CMS and adjust a bit. but overall, my delta has went down even more from less than 3 dE for Gray Scale to less than 2 dE, and Color delta roughly remained unchanged.

and funny thing is, the overall black level also went down from .013 cd/m2 to .012 cd/m2, that's the lowest i've measured at this back light setting. i am not sure why cool.gif but hey.. i am not complaining. smile.gif

here are the results. i did this last night after wife went to bed. lol.. feel like i am sneaking behind her back.








let me know what you think.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › Basic Guide to Color Calibration using a CMS (updated and enhanced)