AVS Forum banner

1 - 20 of 63 Posts

·
Registered
Joined
·
1,033 Posts
Discussion Starter · #1 ·
Was going to post in the calibration section but thought it might be more valuable here. Mods if this should be moved please do.


Well I finally got Calman 5 Autocal to work on a JVC X55R. IMHO the results are nothing short of amazing.


It was quite the adventure into unknown territory as the Calman documentation on Autocal is out of date and missing tons of critical information.

I won't bore you with all the details but I have invested more hours than I can count into this only to repeatedly hit the proverbial brick wall.

Many times I gave up but I just didn't want to believe I couldn't get this to work.

Through tons of testing, guessing, experimenting and some help from Manni on the Spectracal forums and some limited help from Spectracal I was able to to get everything to work.

It was well worth the effort.

Or if you want to save yourself the time hire a professional calibrator



Check out the stellar result below.

Watched Gravity last night and it was stunning!

Note: This is on high lamp on my testing screen-10'.5 diag 16:9 StudioTek130. I used the Anime Color Profile as it was the only mode that did not have an under saturated green.


Grayscale-23 points




CMS results-Note I manuly tweaked CMS, Autocal was close.




Gamma




Dynamic Range

 

·
Registered
Joined
·
1,033 Posts
Discussion Starter · #3 ·

Quote:
Originally Posted by DavidHir  /t/1524873/stellar-results-using-calman-5-autocal-on-a-jvc-x55r#post_24545974


Looks great! What kind of meter?


I've got Chad B coming out next week to calibrate my RS4810 (identical to X55).

Thanks I was amazed by the results.


I tried both the xrite i1Display Pro and the Spectracal C6-essentially a modified i1Display. I compared both meters against a friends Klein and the i1display pro was closer. That's what I used.


Chad is the man I'm sure you will be thrilled. I'm sure he knows this but the color mode Anime was the only one without a grossly under saturated green.
 

·
Registered
Joined
·
14,563 Posts
Chad has calibrated four different displays of mine over the last 9 years including an audio calibration (which was a big improvement over the receiver auto-calibration). He'll be doing both the projector and new audio system next week - will post the results.
 

·
Registered
Joined
·
1,033 Posts
Discussion Starter · #6 ·
I was stunned by the performance of this projector after calibration . It's got to be one of the best values out there.

The black level and shadow details are the best I have seen.
 

·
Registered
Joined
·
14,563 Posts
Thanks and that is great to hear. My room is also painted black with black carpet, so that should help contrast from what I am told. This is my first projector (coming from Panasonic 65" VT60 plasma).
 

·
Premium Member
Joined
·
10,069 Posts
Hi trans-lux,


I'm glad you got it to work in end



However, I would urge you to check how saturations track when using the anime profile with the JVC CMs, and also to run a colorchecker.


With such a wide color profile to start with, and if you target rec709 with the internal CMS, you will likely see a strong undersaturation at 75% sat and below, which leads to significant undersaturation on real content even if the graphs look great at 100%.


You might want to compare these calibrated results with a non corrected standard profile (which will be not as close at 100% sat but closer below) for an overall more accurate picture. Put a content with saturated red like the clowns in The Dark Knight at the beginning, or the background in Kung Fu Panda during the escape of the baddy, you should be able to see if red is undersaturated or not with actual content.


It is likely that the biggest improvement in the picture is on gamma anyway.


I might be wrong but the only way to know is to look at the way saturations (and luminance) track below 100%. There is a linearity tracking layout in Calman that will show you this.


In my experience in order to get a better gamut than standard on the JVCs you have to target a custom, slightly oversaturated gamut at 100% of rec709 or maybe rec709 at 75% sat to get an acceptable balance at all saturation levels.


Let us know if yours is an exception or not!


This being said, the gamma autocal on the JVCs works great, and the fact that you can use 21 steps is a great addition compared to the manual controls. I love the JVC autocal on Calman
 

·
Registered
Joined
·
1,033 Posts
Discussion Starter · #10 ·
Manni

Thanks for the tips.

No matter what color profile I chose green was significantly under saturated.

I thought Anime was the solution but I see your points.

I will check with some content as well as color checker and some sweeps.

Will report back.
 

·
Premium Member
Joined
·
10,069 Posts
Ideally you want a colour profile only slightly oversaturated as a starting point, as the more correction you apply with the JVC CMS to target rec709 at 100% sat, the more undersaturation it creates at 75% sat and below.


Most of the times it means using standard and living with the slight green undersaturation at 100% sat, which is far less visible with real content than the undersaturation at lower levels when targeting rec-709 exactly.


You can also try to raise the color control by a few clicks to compensate the slight undersaturation by a small raise in luminance.


In the past I found these gave better results with real content than using the JVC CMS.


Of course with a 3D LUT you don't have these issues anymore, but that means using an external processor.


Looking forward to your feedback on this.
 

·
Registered
Joined
·
1,033 Posts
Discussion Starter · #13 ·
I think I'm going to loose my mind!

Before I ran colorchecker and the sweeps I wanted to insure that the work I did prior was still intact.

Well it looks like its not


I have touched nothing in the factory menu other than turning on/off e-****.

I did also set the lamp mode to low and high.

I can't image either of these would throw everything off?


I looked at some live content and it looked good

I watched the opening to the Dark Knight and colors were nicely saturated and well balanced..


The one item that stands out is that when I did the Autocal yesterday I see above white values.

Today I don't and it appears that the JVC does not have the ability to generate these patters

I'm very confused....


Below are the measurements from today-notice no above white values present????







 

·
Premium Member
Joined
·
1,789 Posts
Someone else was complaining about Calman on a new X500 wouldn't save the results. Sounds like you have the same issue?
 

·
Premium Member
Joined
·
11,978 Posts

Quote:
Originally Posted by trans_lux  /t/1524873/stellar-results-using-calman-5-autocal-on-a-jvc-x55r#post_24548283


I think I'm going to loose my mind!

Before I ran colorchecker and the sweeps I wanted to insure that the work I did prior was still intact.

Well it looks like its not


I have touched nothing in the factory menu other than turning on/off e-****.

I did also set the lamp mode to low and high.

I can't image either of these would throw everything off?


I looked at some live content and it looked good

I watched the opening to the Dark Knight and colors were nicely saturated and well balanced..


The one item that stands out is that when I did the Autocal yesterday I see above white values.

Today I don't and it appears that the JVC does not have the ability to generate these patters

I'm very confused....


Below are the measurements from today-notice no above white values present????








As far as I know Calman doesn't calibrate for above white levels. There is a clipping checker that does above white in some of the workflows, but it will not be represented in the normal calibration part. It is only there to ensure you're not clipping. I don't even think the 3D LUTs do above white.
 

·
Premium Member
Joined
·
1,789 Posts

Quote:
Originally Posted by Kris Deering  /t/1524873/stellar-results-using-calman-5-autocal-on-a-jvc-x55r/10_10#post_24549547


You can't do an auto cal of a X500 with Calman, the software doesn't support it yet.

My mistake. The post I was remembering was in the X500 thread but he was talking about last year's 4810. Maybe his solution will help trans_lux?
 

·
Registered
Joined
·
1,033 Posts
Discussion Starter · #18 ·
So after countless additional hours I really think I have finally got a handle on this-I think.

IT's clear that there is a small difference between the intarnal JVC patterns and an external source like the AVSHD BD.

Was able to get 14.6ftl post calibration


Some comparisons below.

I also used the Standard color profile vs Anime.

I will try the Anime profile with the same config when I have some time in the future.
I would appreciate any and all feedback on the Saturation and Lumiance sweeps



Using one of the user memories I set up the following;
  1. low lamp
  2. standard color profile
  3. custom color temp
  4. custom gamma
  5. HDMI input set to standard-no above white
  6. Calman ISF calibration for auto calibration
  7. Calman set to ITU BT.1886 gamma target
  8. Used the internal JVC patterns

Saturation sweeps using external patterns from Ted's Lightspace disc


Luminance sweeps using external patterns from Ted's Lightspace disc



Autocal grayscale results using internal JVC patterns


Autocal grayscale results using external AVSHD patterns


Autocal color results using internal JVC patterns


Autocal color results using external AVSHD patterns


Autocal gamma results using internal JVC patterns


Autocal gamma results using external AVSHD patterns
 

·
Premium Member
Joined
·
10,069 Posts
Yes there is a significant difference between the JVC internal patterns (or the Lumagen patterns) and the bluray. As my main source is my HTPC, this is why I use PC Client as a source, as the results are much closer to the actual Bluray output. The most significant difference is at 5%, using the PC Client as a source really helps fighting the black crush on the JVC.


Re saturations, as you can see red after correction targeting rec709 at 100% sat is very significantly undersaturated at 75% and below. This is unfortunately the way the JVC CMS behaves. It has been reported many times over the last few years, and they either don't understand the issue or are unable to fix it.


If you measure anime corrected to rec709 at 100% sat, it should be even worse as more correction in the CMS is needed, as the starting point is more oversaturated.


The wonky gamut on green is another trade mark of many units, especially the two entry-level models, and it also leads to uncorrectable green and cyan.


The problem is that most people - including many reviewers - target rec709 at 100% sat, get a great looking graph without checking the saturations and so are happy with the CMS, despite the fact that when used like this it delivers an undersaturated picture. Not having a go at you by the way, I did it myself until I realized something wasn't right (it's visible, but it's not obvious, and the improvements due to greyscale/gamma make the calibration look better than uncalibrated, so not easy to diagnose).


Bar using an external processor with a 3D LUT to mitigate the issue (I use a Radiance Mini 3D with my rs45), you have to either target a slightly oversaturated primary at 100% sat, so in effect create a custom gamut, or try calibrating at 75% sat of rec709 (75% of rec709) to find a better balance between the points. There is very little content at 100% sat, except maybe some animation. You have much more real content at 75% and below, so to get an accurate picture you need to make sure that 75% is on target. This is best done manually.


If you compare the calibrated gamut with uncorrected standard (which is slightly oversaturated at 100% sat but tracks much better at 75% and below) you should be able to see that the calibration is undersaturated using the examples I suggested earlier.


Of course the only way to compare fairly is to calibrate gamma for both (so that you still get the improvements in the greyscale when watching the uncorrected standard).


So in one user profile, set standard and your custom gamut calibration as well as your gamma settings, and in another select standard with no custom corrections and the same greyscale/gamma corrections.


This way you'll be able to compare the results visually with the only difference being the gamut.


Before doing so, check that the profile with uncorrected standard has a correct greyscale. And measure its saturations, you should see that it tracks much better below 100% sat.
 

·
Registered
Joined
·
1,033 Posts
Discussion Starter · #20 ·
Manni

Thanks so much for the information.

I will follow your suggestions and report back.
Quote:
As far as I know Calman doesn't calibrate for above white levels. There is a clipping checker that does above white in some of the workflows, but it will not be represented in the normal calibration part. It is only there to ensure you're not clipping. I don't even think the 3D LUTs do above white.

Kris

I thought that was odd as well.

If you set the projector to Super White 16 – 255 one of the options that becomes available for the Autocal is 23 points for calibration adding 105 and 109.
 
1 - 20 of 63 Posts
Top