Yet another question on greyscale and gamma. - AVS Forum
Forum Jump: 
 
Thread Tools
Old 01-26-2014, 08:04 AM - Thread Starter
Senior Member
 
Jeroen1000's Avatar
 
Join Date: Nov 2008
Posts: 314
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
CalibrationGreyscale.jpg 123k .jpg file

I'm observing 2 things I can't really explain. Should I try to fix them or not?

1) Although all dE's are in the ballpark, the color temperature at 10% and 50% stimulus seem to be a bit off.
2) The gamma (the new word for this escapes me) dip betwen 60-90% looks pretty bad. Strangely, when I switch to contineous read, the gamma numbers goes up and up as the light outputs keep dropping. If I wait long enough, it goes beyond 2.22:-). What is up with that?

Thanks again for helping me out. I wouldn't get this far with all this help!
Attached Images
File Type: jpg CalibrationGreyscale.jpg (122.6 KB, 82 views)

My gear: Panasonic TH-42PF11EK pro plasma display. -- Iscan Duo video processor -- i1 display 3 colorimeter -- i1 pro 2 spectrometer
Jeroen1000 is offline  
Sponsored Links
Advertisement
 
Old 01-26-2014, 02:44 PM
AVS Special Member
 
Rolls-Royce's Avatar
 
Join Date: Apr 2003
Location: Victorville, CA
Posts: 2,063
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 37 Post(s)
Liked: 68
Quote:
Originally Posted by Jeroen1000 View Post

CalibrationGreyscale.jpg 123k .jpg file

I'm observing 2 things I can't really explain. Should I try to fix them or not?

1) Although all dE's are in the ballpark, the color temperature at 10% and 50% stimulus seem to be a bit off.
2) The gamma (the new word for this escapes me) dip betwen 60-90% looks pretty bad. Strangely, when I switch to contineous read, the gamma numbers goes up and up as the light outputs keep dropping. If I wait long enough, it goes beyond 2.22:-). What is up with that?

Thanks again for helping me out. I wouldn't get this far with all this help!

You are actually in very good shape. If the numbers bother you, just rerun the grayscale autocal if you are using a processor with that capability. Your setup escapes me at the moment.

That gamma glitch at 90% is not uncommon. Luminance wise, 90% is so close to 100% (which is used to calculate gamma for the whole range), that any small difference between actual and target values has a bigger impact than you might think. If you go back and remeasure 100% before remeasuring 90%, you
could find different numbers entirely. It's a moving target, and I honestly don't think it would make much of a difference to the final image.
Jeroen1000 likes this.

...Royce...

"I never drink...wine."
Bela Lugosi, DRACULA, 1931
Rolls-Royce is offline  
Old 01-26-2014, 04:52 PM
 
MonarchX's Avatar
 
Join Date: Jul 2012
Posts: 669
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 30
What Rolls said is spot-on. If you look into many calibration reports you will frequently find gamma past 80% tends to have inaccuracy spikes. At 90%, gamma is meaningless. It is best, however, if your spike goes up instead of down which will keep you from clipping whites and making bright scenes overly bright. I would also use 21steps to see exactly where the error is at. using 10% steps may make it look worse than it actually is.
MonarchX is offline  
Old 01-26-2014, 06:43 PM
AVS Special Member
 
Doug Blackburn's Avatar
 
Join Date: May 2008
Location: San Francisco - East Bay area
Posts: 3,457
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 25 Post(s)
Liked: 228
There are 2 big issues here...

Color Temperature is a HORRIBLE and nearly useless number to use to judge the accuracy of a video display. You can have way too much magenta or green in the image and you will still get a 6500K calculated color temperature.

What is important is D65...that point is defined so that there is only ONE accurate D65 point and that point is the basis for all dE calculations.

But there are problems with dE. The stated goal of dE is to give us an objective comparator (something calculated) that will result in good subjective (something evaluated by eye or taste, or touch, etc.) images. The problem with dE is that is has proven to be harder than HELL to calculate deviations from the D65 point (in our video world) so that every calculated result matches what we see when we look at images. Because of that, there are many formulas for calculating dE and it is surprisingly common for them to produce very different numbers for the same error. For that reason, I never look at just ONE dE number to know if my results are really good or not. I usually look at 3 different dE numbers for each point measured... usually the 1931 dE, 1994, and 2000.

If you low dE errors for all three of those methods of calculating dE, you've got a pretty darn good display calibration and there shouldn't be any significant issues with color temp if you are still looking at it. Part of the art of calibration is knowing where to hide errors. You may have calibration controls that allow you to have things good, say from 40%-60%, but that leaves a problem at 10%-30%. If you fix the problem at 10%-30%, you may find 40%-60% gets worse. That sort of thing is pretty common in TVs that have 2-point grayscale controls. You definitely want to avoid too much green as we will see green errors before any others. But hiding an error that causes a star field to have a red-tinted blacks versus too much blue between 40% and 60% can be tricky to deal with since the red-tinted star fields will be horribly distracting during Star Wars or anything else with "outer space" images, but being too blue through the midrange can cause you to think brighter parts of the screen are too yellow when they are actually accurate. Balancing the problems so all errors are as small as they can be is the goal. Sometimes you can't make 2 problems go away, you just have to find the controls adjustments that knock both problems down by half or so (if such settings exist) and live with those residual errors until you can add something like a Lumagen Radiance vide processor that can make grayscale, gamma, and color very accurate for most video displays even if the controls in the video display itself are not very good.
Jeroen1000 likes this.

"Movies is magic..." Van Dyke Parks
THX -- ISF -- HAA
Doug Blackburn is offline  
Old 01-27-2014, 01:25 AM - Thread Starter
Senior Member
 
Jeroen1000's Avatar
 
Join Date: Nov 2008
Posts: 314
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
I've updated my signature for the gear info. That will be useful and take out the gear guessing. Thanks for excellent explanation. Doug, as usual, is very thorough!

I've watched some TV and now I can visually tell what is wrong:

- With darker scenes (shadow, indoor, ...) the image is too dark. I've used power law gamma with 2.2 exponent and a 0 cd/m² luma target for black. My tv obviously does not go that black so the gamma at the low end should be 2.15 - 2.18 rather than 2.2 I guess. However, Calman does not give me a plain Y control like the Iscan Duo Control Panel software. So I have to adjust R, G and B and I've got no idea which one to set higher. The Y control would preserve the RGB ratio for me (dummy proof):-). Is there a good way to go about adjusthing Y?
Bright scenes look supberb!

- I think this one has to do with clipping withes: upon detecting the Iscan Duo, Calman presents a full screen pattern like this one (comes from another site, I'll take a picture later)
Now the portion to the utmost right has a pink hue. It does not show up on my grey scale, and I can't spot it in a 90-100% white pattern (10% of screen area) but it's there on that test pattern and quite clearly too. Does that indictate some sort of problem? Maybe I should get the AVS pattern which show the bars beyond 100% and check whether there is discoloration present beyond video level 235.


ps: I like autocal but I also like to do it myself because it is fun and I like to learn.

My gear: Panasonic TH-42PF11EK pro plasma display. -- Iscan Duo video processor -- i1 display 3 colorimeter -- i1 pro 2 spectrometer
Jeroen1000 is offline  
Old 01-27-2014, 06:44 PM
Advanced Member
 
Vic12345's Avatar
 
Join Date: Mar 2012
Posts: 776
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 20 Post(s)
Liked: 26
Good post Doug B.It helps having your posts.Check is in the mail.lol
Vic12345 is offline  
Old 01-28-2014, 03:01 PM
Advanced Member
 
hungro's Avatar
 
Join Date: Mar 2007
Posts: 628
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 27 Post(s)
Liked: 61
If you are sending him a Cheque then save that Money Vic and buy yourself a meter to calibrate rolleyes.gif
hungro is online now  
Old 01-31-2014, 08:53 AM
Member
 
randal_r's Avatar
 
Join Date: Mar 2010
Posts: 174
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 54 Post(s)
Liked: 67
...
randal_r is offline  
Old 01-31-2014, 12:31 PM
AVS Special Member
 
sotti's Avatar
 
Join Date: Aug 2004
Location: Seattle, WA
Posts: 6,640
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 44 Post(s)
Liked: 175
Quote:
Originally Posted by Jeroen1000 View Post

CalibrationGreyscale.jpg 123k .jpg file

I'm observing 2 things I can't really explain. Should I try to fix them or not?

1) Although all dE's are in the ballpark, the color temperature at 10% and 50% stimulus seem to be a bit off.
2) The gamma (the new word for this escapes me) dip betwen 60-90% looks pretty bad. Strangely, when I switch to contineous read, the gamma numbers goes up and up as the light outputs keep dropping. If I wait long enough, it goes beyond 2.22:-). What is up with that?

Thanks again for helping me out. I wouldn't get this far with all this help!

1) Your RGB balance is absolutely great for a 2pt control if you have a 10 point, than touching up 50% might make sense. Let dE be your guide for weather to worry more or not, CCT is a nearly worthless metric.

2) That's pretty normal in alot of displays. Turning contrast down a couple clicks might help, but look at the dE's and the color comparator. On the color comparator can you tell the difference between the target and the measured values at any point?

No?

Neither can I.

So since you can't see a difference between absolute perfection and your TV. Close CalMAN down and go enjoy some TV and Movies. The Superbowl is this weekend, should be good to watch. Go Hawks!

Joel Barsotti
SpectraCal
CalMAN Lead Developer
sotti is offline  
Old 01-31-2014, 12:41 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 943
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 120 Post(s)
Liked: 75
Ah I keep seeing "go hawks" all the time, and now I know what it means! I take it they're a football team.
spacediver is offline  
Old 01-31-2014, 01:10 PM
AVS Special Member
 
rahzel's Avatar
 
Join Date: Jan 2006
Location: Canada
Posts: 3,714
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 147
Seattle Seahawks are playing the Denver Broncos in the Superbowl.

Not a sports guy? smile.gif
rahzel is offline  
Old 01-31-2014, 01:25 PM
Advanced Member
 
spacediver's Avatar
 
Join Date: May 2013
Location: Toronto
Posts: 943
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 120 Post(s)
Liked: 75
thanks for the info. I love playing sports (tennis, ultimate, volleyball), but I only follow tennis and mma. I did gain an appreciation for (american) football after recently watching the tv show friday night lights smile.gif
spacediver is offline  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off