LightSpace 3D LUT Home Cinema Calibration Software - Page 74 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 1103Likes
Reply
 
Thread Tools
post #2191 of 2392 Old 04-04-2019, 06:12 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,110
Mentioned: 329 Post(s)
Tagged: 0 Thread(s)
Quoted: 5439 Post(s)
Liked: 5640
Quote:
Originally Posted by ConnecTEDDD View Post
Using your data Steve found the problems, we trust your measurements.

This specific gamut problem its a device issue, not only the specific model you have issue, but there a lot of other JVC which suffering from the same problem.

You have an American JVC projector in Europe, so seems that you have some serious contacts in USA, but you opened I see a new thread about that projector AutoCAL to another forum, not in AVSForum, something you where doing at past year models. (while AVSForum is the most popular forum in the world)

Thinking about the above, I find impossible to accept from your side that there is any fault being on JVC's part and always it will be a LightSpace (or other software problem).

The same problems you have find with CalMAN also, for that reason Lightning LUT provide you better results (since it has zero volumetric info...where the issues are starting when you generating volumetric correction, with lower saturation boost).

Anyway, LightSpace is working on solutions for such problematic devices gamut's, so stay tuned...
I opened the new calibration thread in a different forum mainly because I was offered more support there. I had not started a new thread on AVS since 2016, so it's not really a new thing.

I am still active on AVS and have reported many issues with the new JVCs (especially regarding the dynamic iris implementation in HDR and the colorspace bug with forced YCC422 in 12bits), on both forums, including in the post just above yours and recently in this owner's thread for the new models. I also mentioned them in the new calibration thread, in my first impressions and in the "suggestions to JVC for improvements". You clearly are not following the JVC threads very closely.

I never said that the issues are always with Lightspace. Just that in my experience, when there is an issue with Lightspace, it's the user or the display that is blamed, and the issue isn't fixed, the user is asked to perform convoluted manual steps as a workaround. This is what I am protesting against. Also you say that my measurements are trusted, well, not really. I was asked to produce far too many measurements the last time, and when I was asked to produce more, I refused as it was not needed. I was correct as the existing measurement were indeed enough for Steve to identify the issue *in Lightspace* that was causing posterization, and to create manually a LUT that resolved it using the convoluted LUT concatenation process. So yes, if my feedback was taken more seriously and if I wasn't asked to jump through hoops every time I report something, no doubt I would be less frustrated. I would also be happy to use email as a way to communicate.

You are correct, I do have some issues with Calman too and I have reported them many times, but at the moment I get excellent results with it on the rs2000 and as the interface is better for my needs, I use it more. If the outstanding issues with Lightspace were resolved, especially regarding madVR support, I would no doubt use LS more, especially for larger LUTs, at least until CM resolves its outstanding issues.

Anyway, as you said I've made all my points very clear. I just hope that attitude towards experienced JVC and madVR users will change in this thread.

I'm glad to hear that improvements are coming for undersaturated displays, though this isn't an issue I have at the moment with the rs2000.

By the way, you have my email, if you want to discuss any of this, you know how to reach me. It might be a good idea to clear the air.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders

Last edited by Manni01; 04-04-2019 at 06:40 AM.
Manni01 is online now  
Sponsored Links
Advertisement
 
post #2192 of 2392 Old 04-04-2019, 06:36 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by ConnecTEDDD View Post
Hi,

The gamma value is what you see when you are looking the DifGamma chart: https://www.lightillusion.com/profil...splay_difgamma

(You can see the gamma value for each point, as it is shows relative to the target color space gamma value.)
That is not easy to do when the target gamma is not a constant value (e.g. BT.1886).
Dominic Chan is online now  
post #2193 of 2392 Old 04-04-2019, 06:59 AM
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,377
Mentioned: 207 Post(s)
Tagged: 1 Thread(s)
Quoted: 3249 Post(s)
Liked: 4043
Quote:
Originally Posted by Dominic Chan View Post
That is not easy to do when the target gamma is not a constant value (e.g. BT.1886).
If you understand gamma, it is not so difficult, see:



The DifGamma plot shows the measured gamma value relative to the target gamma value.

So targeting a power law gamma 2.4 shows the BT.1886 values.

Targeting BT.1886 directly would show the difference between the compound BT.1886 gamma and the actual measured value.

This is probably more useful than knowing the actual gamma values per each point.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5

Last edited by ConnecTEDDD; 04-04-2019 at 07:09 AM.
ConnecTEDDD is online now  
Sponsored Links
Advertisement
 
post #2194 of 2392 Old 04-04-2019, 07:15 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
Quote:
Originally Posted by ConnecTEDDD View Post
If you understand gamma, it is not that difficult, see:



The DifGamma plot shows the measured gamma value relative to the target value.

So in this case, targeting a power law gamma 2.4 shows the BT.1886 values.

Targeting BT.1886 directly would show the difference between the compound BT.1886 gamma, and the actual measured value, which is probably more useful than knowing the actual gamma values per each point.
What graph are you showing here? According to the “LightSpace convention” in the link you provided, in a DifGamma graph a point with gamma 2.0 would be shown to be 0.4 above the 2.4 line, not below.

Quote:
The DifGamma display shows differential gamma values, with points above the line having a lower gamma value, so showing a lighter image display, and points below the line having a higher gamma value, so a darker image display.
That convention is totally illogical, but that’s beside the point here.

Last edited by Dominic Chan; 04-04-2019 at 07:44 AM.
Dominic Chan is online now  
post #2195 of 2392 Old 04-04-2019, 08:30 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
Good spot Dominic - I actually plotted a power law 2.4 gamma vs. BT1886 target for a display with the same black level.
Forgot I plotted it that way round.
Have inverted the graph to be the opposite now, so it now shows BT1886 vs. a target 2.4 gamma.

The graphs comes from here: https://www.lightillusion.com/error.html#bt1886_gamma

Steve
mrtickleuk likes this.

Steve Shaw
LIGHT ILLUSION

Light Illusion is online now  
post #2196 of 2392 Old 04-04-2019, 09:08 AM
AVS Forum Special Member
 
Join Date: Dec 2014
Location: Mississauga, ON, Canada
Posts: 6,946
Mentioned: 130 Post(s)
Tagged: 0 Thread(s)
Quoted: 4883 Post(s)
Liked: 1847
If “The first and last points have no valid gamma value” then they should not be plotted on the DifGamma graph. The graph shows them as having a gamma of 2.4.
Dominic Chan is online now  
post #2197 of 2392 Old 04-04-2019, 09:24 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
It's just a default 'zero' plot for no data.
But, I do understand the comment, and a change has been on the WIBNI list for some time.
However, other more important things keep bumping it down the list...


Steve

Steve Shaw
LIGHT ILLUSION

Light Illusion is online now  
post #2198 of 2392 Old 04-04-2019, 07:23 PM
Member
 
Join Date: Mar 2016
Posts: 56
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Liked: 30
Quote:
Originally Posted by ConnecTEDDD View Post
LightSpace is working on solutions for such problematic devices gamut's, so stay tuned...
Looking forward to that update. I hope the new solutions will apply to my Sony TV too.
Manni01 likes this.
omarank is online now  
post #2199 of 2392 Old 04-07-2019, 11:09 PM
Senior Member
 
vollans's Avatar
 
Join Date: Feb 2016
Location: Australia
Posts: 348
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 210 Post(s)
Liked: 153
Quote:
Originally Posted by ConnecTEDDD View Post
LightSpace CMS 10.0.0.2922 (21 March 2019) has been released.

Release Notes (incremental changes):

Add Sony ImageDirector 1D LUT format (for projectors).
Quote:
Originally Posted by vollans View Post
Fantastic - once I’m back on my feet and can physically get about again, I’ll have to give it a go and see how it goes with my Sony 760/885.
OK, I’m mobile, and been able to get to my theatre room, hook up the Sony to the network, and see how it goes. First impressions are great. It’s early days, but I’ve been mainly playing with using the new tools to tweak the output from the Panasonic UB820 so that it does an even better job of tone mapping in conjunction with the Sony. Using the Lightspace software I’ve been able to create a 1D LUT for the projector that tone maps the UB820 to a BT2390 curve which is helping enormously with shadow detail AND getting HDR moments in films that actually pop.

I’ve attached a very quick and dirty camera photo showing the same frame in the basic output from the UB820 with its own tone mapping, and a second with the Lightspace BT2390 curve applied. You should be able to see that some detail that was not at all visible before is now actually visible.

I’ve got a lot more to do, but first impressions are very positive.
Attached Thumbnails
Click image for larger version

Name:	37C9FEA7-0DE5-43CD-9EBA-BF9CEB439E12.jpg
Views:	73
Size:	44.1 KB
ID:	2550270   Click image for larger version

Name:	87B9BE9C-6C08-4D28-9CE5-26D370D218B7.jpg
Views:	74
Size:	46.5 KB
ID:	2550272  
wl1 and Anger.miki like this.

New Home Cinema: Sony VPL-VW760, Anthem MRX 1120, Q Acoustics 30x0 series speakers fronts, SVS Prime Elevation surround speakers, Proficient C791 Atmos, SVS PB2000 sub, Panasonic DMP-UB820, Apple TV, Pioneer CLD-D5V laserdisc, FetchTV, eeColor 3D LUT & Radiance XD-3D and ScreenTechnics CinemaSnap Hi Contrast 130" screen
vollans is online now  
post #2200 of 2392 Old 04-08-2019, 12:03 AM
aka jfinnie
 
Join Date: Aug 2015
Location: Norwich, UK
Posts: 3,270
Mentioned: 53 Post(s)
Tagged: 0 Thread(s)
Quoted: 2641 Post(s)
Liked: 1739
Quote:
Originally Posted by vollans View Post
OK, I’m mobile, and been able to get to my theatre room, hook up the Sony to the network, and see how it goes. First impressions are great. It’s early days, but I’ve been mainly playing with using the new tools to tweak the output from the Panasonic UB820 so that it does an even better job of tone mapping in conjunction with the Sony. Using the Lightspace software I’ve been able to create a 1D LUT for the projector that tone maps the UB820 to a BT2390 curve which is helping enormously with shadow detail AND getting HDR moments in films that actually pop.

I’ve attached a very quick and dirty camera photo showing the same frame in the basic output from the UB820 with its own tone mapping, and a second with the Lightspace BT2390 curve applied. You should be able to see that some detail that was not at all visible before is now actually visible.

I’ve got a lot more to do, but first impressions are very positive.
It's different for sure, but what's your reference? Very difficult to tell from the photo as ever, but it almost looks to lack both contrast and saturation. Should that background detail really be that prominent in the scene? Feels a little robbed of atmosphere.

What's the scene out of interest? I'm intrigued to know how my Lumagen dynamic tone mapping tone maps it through my JVC X7900 (not that it is any kind of reference either, I'm just interested).

Tone mapping is such a minefield...!
bobof is offline  
post #2201 of 2392 Old 04-08-2019, 12:14 AM
Senior Member
 
vollans's Avatar
 
Join Date: Feb 2016
Location: Australia
Posts: 348
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 210 Post(s)
Liked: 153
It’s in the first half of Wonder Woman, I didn’t make a note. I had to freeze it for my other half to go and check the barbie and thought I’d quickly check, and took a quick snap as it was the detail that I noticed, it didn’t look anything like as muddy in real life but that’s camera phones for you in a dark room. There’s a lot of smokey hazy scenes in that film. Over the whole film it really brought it to life. What was really noticeable was in the big “HDR scenes” you could see shadow detail and the HDR pop. That’s something that I’ve not really seen so far with just the Panny doing the hard lifting, or the Sony alone doing the hard lifting.

The 1D output for the Sony works, and with a bit of time I think I’ll be able to get something really good out of it.
bobof likes this.

New Home Cinema: Sony VPL-VW760, Anthem MRX 1120, Q Acoustics 30x0 series speakers fronts, SVS Prime Elevation surround speakers, Proficient C791 Atmos, SVS PB2000 sub, Panasonic DMP-UB820, Apple TV, Pioneer CLD-D5V laserdisc, FetchTV, eeColor 3D LUT & Radiance XD-3D and ScreenTechnics CinemaSnap Hi Contrast 130" screen
vollans is online now  
post #2202 of 2392 Old 04-18-2019, 11:15 AM
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,377
Mentioned: 207 Post(s)
Tagged: 1 Thread(s)
Quoted: 3249 Post(s)
Liked: 4043
LightSpace CMS 10.0.0.2938 (16 April 2019) has been released.

Release Notes:

Corrected issue with Stabilisation patches with connected hardware generators.

Corrected rounding error with JVC 1D LUT export format.

Added Sub Black option for VideoScale.

Corrected CSV Grey Only Quick Profile recognition.

Fix issue with Minolta CA-410 probe integration.

Removed Osee and Bon Upload options, as manufacturers never completed their work.

Added ‘Stop’ and ‘Pause’ capability to Pre-roll.

Added ability to use ‘Stabilisation’ frames with Pre-roll.

Fix issue with adjusting Pre-roll duration preventing it from working.

Added ‘Edge’ Quick Profile – a work-in-progress.

Download Link: https://www.lightillusion.com/downloads.html
hrc4u, chunon, T( )( )L and 2 others like this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5
ConnecTEDDD is online now  
post #2203 of 2392 Old 04-20-2019, 08:17 PM
Senior Member
 
vollans's Avatar
 
Join Date: Feb 2016
Location: Australia
Posts: 348
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 210 Post(s)
Liked: 153
Hmm. Not impressed. Just had Lightspace Connect upgrade to a new version. I’d already purchased the unlimited patches on the previous version, and now it won’t restore the purchase and says I need to pay again. Some warning would have been appreciated.

Edit: Even weirder, my second phone didn’t have a problem, but my main phone point blank is refusing to restore the purchase.
Attached Thumbnails
Click image for larger version

Name:	06742F9D-59B2-4B94-ABA2-D441C87C2693.jpg
Views:	50
Size:	44.2 KB
ID:	2556272  

New Home Cinema: Sony VPL-VW760, Anthem MRX 1120, Q Acoustics 30x0 series speakers fronts, SVS Prime Elevation surround speakers, Proficient C791 Atmos, SVS PB2000 sub, Panasonic DMP-UB820, Apple TV, Pioneer CLD-D5V laserdisc, FetchTV, eeColor 3D LUT & Radiance XD-3D and ScreenTechnics CinemaSnap Hi Contrast 130" screen

Last edited by vollans; 04-20-2019 at 08:42 PM.
vollans is online now  
post #2204 of 2392 Old 04-21-2019, 02:05 AM
AVS Forum Special Member
 
ConnecTEDDD's Avatar
 
Join Date: Sep 2010
Location: Athens, Greece
Posts: 8,377
Mentioned: 207 Post(s)
Tagged: 1 Thread(s)
Quoted: 3249 Post(s)
Liked: 4043
Alternate Patch Sets

LightSpace can use any user defined Patch Set, either as a Quick Profile or as a full Characterisation.

Example Patch Sets for both can be downloaded via the Downloads page, and information on their use can be found in the Display Profiling User Guide.

Reduced Gamut Profiling

However, there are further uses for User Patch Sets, such as deliberately reducing the Gamut Area to be profiled, which can be of use with displays that have a smaller gamut than the required colour space target, with internal colour management that cannot be disabled. Such displays will suffer gamut edge compression, such as JVC Wide Gamut projectors, that are colour managed to Rec2020 that cannot be disabled, as described in the Error! page, Pre Calibration Issues.

The first step is to define the actual gamut area the display is actively using within its target colour space. For example, using a JVC projector set to Rec2020 wide gamut, the R, G, and b peak xy values can be extracted, leaving the white point and gamma as per the Rec2020 colour space.

To define the actual gamut area of the display profile with a Grey Only, or Primary Only Quick Profile, and use the Points Info pop-up window to interrogate the max Red, Green, and Blue values, as shown.



The extracted colour space can then be used to define a gamut reduction LUT, using Convert Colour Space with Source as the Extracted colour space, and the Destination as the target colour space - Rec2020 for the JVC projector mentioned here.



The gamut reduction LUT can then be used as an 'Active LUT' within the Characterisation menu, and 'Export Colour List' used to export a CVS file with the reduced gamut patch set.



The patch set will not yet work as desired, as there are no 100% patches at all. So, 100% Red, Green, and Blue patches need to be added to the end of the list using Excel.



You will then have a reduced volumetric gamut profile set, with Peak RGB patches at 100%, so precisely locating the gamut area to be profiled.

The patch set will therefore not attempt to profile outside the display's available gamut, so will avoid any edge compression issues, such as suffered when a display is pre-calibrated to a larger gamut than it can actually attain.

But, by also having the three 100% RGB patches in the profile set means the calibration LUT will correctly target the desired colour space - Rec2020 in the above examples.

Note: When exporting the reduced gamut patch set from with the Characterisation menu it may be preferable to reduce the cube size - say 19^3, or 17^3, so when the final calibration LUT is being generated it will have approximately the same volumetric granularity as a standard 21^3 profile. However, a gamut reduced patch set can be made to any size, and even with a different percentage reduction on the different gamut edges, etc.

In some circumstances it may be helpful to slightly reduce the gamut further - say 5% or so, just to make sure the reduced gamut patch set is fully within the display's available gamut coverage.

The following example shows the extracted JVC gamut Green being reduced buy 5%, with the new xy values shown. The same would need to be done for Red and Blue, and a new 'Colour Space' saved using the new RGB xy values.



Note: The Gamut Reduction will be based on the relative distance of Red, Green, and Blue from the colour space white point.

While the above example only added back the three 100% Red, Green, and Blue patches, it may be preferable to add back 100% CMY as well, or all Primary Red, Green, and Blue patches, or even add back all the patches on the gamut edge!

(Any patch with a zero value for one of the RGB colour channels will be a gamut edge colour, so XXX,XXX,0 or XXX,0,XXX or 0,XXX,XXX...)

Additionally, it may be preferable to focus on getting the reduced gamut patch set best suited to the display first, finding the best Red, Green, and Blue gamut reduction values that on their own produce the best calibration with the best overall calibrated gamut.

(The calibrated gamut will be smaller than the actual max gamut of the display, as there are no 100% gamut patches.)

When the ideal reduced gamut set has been defined, the 100% gamut patches can be added back, to restore the the missing gamut.

Gamma Too!

From the above it should be obvious that altering just the Gamma value when generating a LUT to be used to export a User Patch Set will 'bias' the patches more towards the blacks, or the whites.



In the above example, the default Gamma of 2.4 for a Rec709 display will effectively be cancelled out, making the patches more equally spaced in linear light terms.
mikela and omarank like this.

Ted's LightSpace CMS Calibration Disk Free Version for Free Calibration Software: LightSpace DPS / CalMAN ColorChecker / HCFR
S/W: LightSpace CMS, SpaceMan ICC, SpaceMatch DCM, CalMAN 5, CalMAN RGB, ChromaPure, ControlCAL
V/P: eeColor 3D LUT Box - P/G: DVDO AVLab TPG
Meters: JETI Specbos 1211, Klein K-10A, i1PRO2, i1PRO, SpectraCAL C6, i1D3, C5

Last edited by ConnecTEDDD; 04-23-2019 at 11:50 PM.
ConnecTEDDD is online now  
post #2205 of 2392 Old 04-21-2019, 02:57 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
Quote:
Originally Posted by vollans View Post
Hmm. Not impressed. Just had Lightspace Connect upgrade to a new version. I’d already purchased the unlimited patches on the previous version, and now it won’t restore the purchase and says I need to pay again. Some warning would have been appreciated.

Edit: Even weirder, my second phone didn’t have a problem, but my main phone point blank is refusing to restore the purchase.
Please use the 'contact' details associated with the App (in the App store).
That seems to be a probable bug!

Steve
vollans likes this.

Steve Shaw
LIGHT ILLUSION

Light Illusion is online now  
post #2206 of 2392 Old 04-21-2019, 03:03 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
The 'Alternate Patch Set' info Ted has posted above (the original info is here: https://www.lightillusion.com/advanc...tml#patch_sets) will help with displays that suffer Gamut edge compression, as with the poor pre-set wide gamut modes some JVC & Sony projectors seem to have.

The attached file is an example Reduced Gamut patch set, with all 100% primary colours included.

Steve
Attached Files
File Type: zip 75%Gamut_All-Primary.zip (38.6 KB, 14 views)
mikela, Rolls-Royce and omarank like this.

Steve Shaw
LIGHT ILLUSION

Light Illusion is online now  
post #2207 of 2392 Old 04-21-2019, 09:03 AM
Advanced Member
 
jqmn's Avatar
 
Join Date: Apr 2005
Location: Houston, TX
Posts: 792
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 433 Post(s)
Liked: 155
Steve-- Is this the right patch set? When I look at these patches they show 28% in the 75% - 100% saturation range, 46% in the 75% - 100% brightness range and they have secondaries included.



Quote:
Originally Posted by Light Illusion View Post
The 'Alternate Patch Set' info Ted has posted above (the original info is here: https://www.lightillusion.com/advanc...tml#patch_sets) will help with displays that suffer Gamut edge compression, as with the poor pre-set wide gamut modes some JVC & Sony projectors seem to have.

The attached file is an example Reduced Gamut patch set, with all 100% primary colours included.

Steve
jqmn is offline  
post #2208 of 2392 Old 04-21-2019, 09:15 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
The patch set is a 25% reduction in volumetric saturation. That is what it is intended to be. With 100% RGB primary values added back.

It is an example of what is possible, and you will do better by making your own specific patch sets.

Hopefully the provided info explains how to do that?

Steve

Steve Shaw
LIGHT ILLUSION


Last edited by Light Illusion; 04-21-2019 at 12:45 PM.
Light Illusion is online now  
post #2209 of 2392 Old 04-21-2019, 09:45 AM
Advanced Member
 
jqmn's Avatar
 
Join Date: Apr 2005
Location: Houston, TX
Posts: 792
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 433 Post(s)
Liked: 155
Thanks/maybe as you know this an area of very limited knowledge for me! I was reading your initial post as a patch set that had patch saturation to 75% and lower only. I will go back and look at Ted's post above.



Quote:
Originally Posted by Light Illusion View Post
The patch set is a 25% reduction in volumetric saturation. That is what it is intended to be. With 100% RGB primary values added back.

It is an example of that is possible, and you will do better by making your own specific patch sets.

Hopefully the provided info explains how to do that?

Steve
jqmn is offline  
post #2210 of 2392 Old 04-24-2019, 02:19 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
The page - https://www.lightillusion.com/advanc...tml#patch_sets - has been updated with a much easier approach to understanding the use of, and the generation of, Gamut Reduced Patch Sets.

It should make a lot more sense now.

But you will likely need to make a patch set specific to your own display, using the outlined approach.

We have looked ta a number of different JVC/Sony projectors that have the issue of a fixed Rec2020 colour space being applied, and all have had very different actual gamut abilities.
A 'generic' reduced gamut set, as the 75% reduced saturation set I posted above, will not work correctly on most such displays.

Steve
mikela, alex_t, chunon and 2 others like this.

Steve Shaw
LIGHT ILLUSION

Light Illusion is online now  
post #2211 of 2392 Old 04-24-2019, 03:32 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,110
Mentioned: 329 Post(s)
Tagged: 0 Thread(s)
Quoted: 5439 Post(s)
Liked: 5640
Quote:
Originally Posted by Light Illusion View Post
The page - https://www.lightillusion.com/advanc...tml#patch_sets - has been updated with a much easier approach to understanding the use of, and the generation of, Gamut Reduced Patch Sets.

It should make a lot more sense now.

But you will likely need to make a patch set specific to your own display, using the outlined approach.

We have looked ta a number of different JVC/Sony projectors that have the issue of a fixed Rec2020 colour space being applied, and all have had very different actual gamut abilities.
A 'generic' reduced gamut set, as the 75% reduced saturation set I posted above, will not work correctly on most such displays.

Steve
Thanks for taking the time to investigate a few of these projectors.

A few remarks/questions:

1) The colour space on the JVCs isn't fixed to BT2020. The profile off mode (native gamut) might track BT2020 more or less, but it's not necessarily the best option to use that setting to calibrate the units. The various factory profiles (especially in the new models) are clearly labeled (rec-709, DCI, BT2020) or if not available can easily be created and uploaded. They track the target remarkably close out of the box once 100% is set to D65, up to the limit of the native gamut, with or without the filter applied (no filter with rec-709, with filter for DCI/BT2020). Many units don't need a LUT, especially a large one, to produce excellent results, especially after a JVC Autocal done properly (i.e. correcting the errors of the supported meters with a reference meter if necessary).

2) You have to be careful of the patterns you send. At least when using madTPG, if using 12bits (and possibly 10bits, I couldn't test with nVidia), you have to report BT2020 in the SDR WCG HDMI stream otherwise the correct colorspace isn't selected by the projector and you end up with a distorted gamut that needs a lot more correction than when the correct colorspace is selected. I don't see this issue when using 8bits, which is one of the reasons why I use 8bits, the other being the forced YCC422 colorspace in 12bits when using RGB Full or YCC444, which can also cause picture quality issues. If using 12bits patterns, YCC422 is recommended at the moment.

3) Most units will not show significant undersaturation or gamut compression when selecting one of the factory (or custom) color profiles. Using the JVC Autocal software if/when necessary, they will track the targets very closely within the gamut, and if necessary a small LUT will bring them close to reference when the target is rec-709 or DCI-P3. Those with a P3 filter will track BT2020 up to around 75% with the filter enabled (close to 100% of P3) when the BT2020 profile is selected, so should reproduce most content absolutely fine. With such a linear baseline, a larger LUT is not only unnecessary, it can in fact produce worse results, such as banding and posterization.

4) For units that do show a significant undersaturation using a factory color profile, then using profile off can help for rec-709 in conjunction with a large LUT, it won't help with DCI-P3 or BT2020 because profile off doesn't (yet) allow the use of the filter unless you use IP commands to flip it in the path every time it's needed, which is an option for some but not for most users. Hence using profile off will make the issue worse, not better, and will produce a poorer calibration in most cases. Hopefully JVC will correct this and offer an option for profile off with the filter in the future.

5) I still don't understand what prevents Lightspace from doing automatically the steps you suggest the user should do manually: measure x,y for each primary and if a primary is significantly undersaturated create a custom gamut based on these measurements and produce a LUT with the best possible results for that unit. Why isn't there a LUT option doing just that, given that x,y for each primary is measured in every single profile? In other words, which step can only be performed by a human and would produce worse results if done automatically by an improved LUT engine?

Thanks again for spending the time to investigate how to get the most out of these units with Lightspace and other available tools.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders

Last edited by Manni01; 04-24-2019 at 03:37 AM.
Manni01 is online now  
post #2212 of 2392 Old 04-27-2019, 12:45 PM
Member
 
Join Date: Mar 2016
Posts: 56
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 36 Post(s)
Liked: 30
Quote:
Originally Posted by Light Illusion View Post
The page - https://www.lightillusion.com/advanc...tml#patch_sets - has been updated with a much easier approach to understanding the use of, and the generation of, Gamut Reduced Patch Sets.
Thanks for providing a solution for displays exhibiting gamut edge compression.
This new approach is very well explained in the guide. But there are a lot of manual steps involved.

Given the process is iterative in nature, I think at least the part of creating a native or reduced gamut color space should be automated. If LS could extract the max Red, Green and Blue coordinates from a Quick Profile run and create a color space, it would make things a bit convenient. Even better would be if it was also possible to enter the gamut reduction % for each primary. So, for example, if I fed a Quick Profile measurement and entered 5%, 10% and 7% for R, G and B respectively, it would extract the max R, G and B coordinates from the bcs file first and then reduce the saturation of R by 5%, G by 10% and B by 7% to derive the reduced gamut coordinates and then create a colorspace for it. Then as per the process, we can use this colorspace to create a LUT which would be used to generate Reduced Gamut Profiling sequence.
Manni01 and mrtickleuk like this.
omarank is online now  
post #2213 of 2392 Old 05-13-2019, 12:27 PM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
Working with the help of Craig Rounds we have been spending a fair bit of time assessing the issues with the likes of JVC/Sony/SIM2 etc. projectors that have fixed pre-calibration to gamuts that are far larger than the projector's native gamut.
(Obviously true of any display that is pre-calibrated to a colour space that is larger than the gamut the display can actually achieve...)

The real answer is, and always will be, for the manufacturers to offer a mode that set the display to it's native gamut, but...

So, we have been working on way to overcome the pre-calibration issues, and the Light Illusion website now has updated information on the best approaches.

https://www.lightillusion.com/error....re-calibration
https://www.lightillusion.com/advanc...tml#patch_sets

As part of this we are adding new capabilities to LightSpace that will 'ease' the workflow for such pre-calibrated displays.
The first new addition will be in the next LightSpace release, and is contained within the 'Colour Sub Space' option of Display Characterisation.
It enables gamut reduced profiling, without the need to generate a User CSV Patch Set.

A Beta version is available via Ted as usual.

Steve

Steve Shaw
LIGHT ILLUSION

Light Illusion is online now  
post #2214 of 2392 Old 05-13-2019, 01:57 PM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,110
Mentioned: 329 Post(s)
Tagged: 0 Thread(s)
Quoted: 5439 Post(s)
Liked: 5640
Quote:
Originally Posted by Light Illusion View Post
Working with the help of Craig Rounds we have been spending a fair bit of time assessing the issues with the likes of JVC/Sony/SIM2 etc. projectors that have fixed pre-calibration to gamuts that are far larger than the projector's native gamut.
(Obviously true of any display that is pre-calibrated to a colour space that is larger than the gamut the display can actually achieve...)

The real answer is, and always will be, for the manufacturers to offer a mode that set the display to it's native gamut, but...

So, we have been working on way to overcome the pre-calibration issues, and the Light Illusion website now has updated information on the best approaches.

https://www.lightillusion.com/error....re-calibration
https://www.lightillusion.com/advanc...tml#patch_sets

As part of this we are adding new capabilities to LightSpace that will 'ease' the workflow for such pre-calibrated displays.
The first new addition will be in the next LightSpace release, and is contained within the 'Colour Sub Space' option of Display Characterisation.
It enables gamut reduced profiling, without the need to generate a User CSV Patch Set.

A Beta version is available via Ted as usual.

Steve
Hi Steve,

Thanks, it sounds like you are going in the right direction regarding making it easier for the user to deal with undersaturated gamuts. I'm looking forward to testing this when it comes out of beta.

It would be really useful if you could provide the settings you have used for the pattern generator and the projector in your tests, as well as the target gamut used for your measurements because without this information your data doesn't really make sense. I don't see anything wrong in the JVC chart you are showing. Are you really expecting the BT2020 mode in any consumer display/projector to not map to the limit of their native gamut? What matters is how well the saturations track within the native gamut, and if they track the target gamut (which is what the JVC does very well in rec-709, DCI-P3 and BT2020) I really don't see what's wrong or surprising in a consumer display or projector that isn't sold as covering BT2020 but DCI-P3, yet offers a BT2020 color profile to display consumer titles using this container, even if the content itself rarely goes beyond DCI-P3. What would you expect a display that only covers DCI-P3 to do when they offer a BT2020 profile in order to play correctly content using this container?

Again, the JVC projectors when configured properly do not force a BT2020 gamut unless you ask for one, so I'm not not sure why you keep repeating this. I don't understand that claim.

Although the BT2020 color profile doesn't cover beyond DCI-P3, which is what the projector covers natively, the BT2020 saturations are tracking well up to the limits of the native gamut, which produces very good results as a starting point.

You can get a very good rec-709 or DCI-P3 baseline (especially when using the P3 filter) without having a "forced" BT2020, so I'm really not sure what you are on about.

When using a BT2020 color profile, the errors beyond DCI-P3 don't matter because there is very little content, if any, beyond DCI-P3 from current consumer sources. This should only be used if you have a source that can only output BT2020 and are not using a VP and a 3D LUT.

With a projector such as the JVCs and with a VP such as the Radiance Pro or with MadVR, there is no advantage to using a BT2020 color profile/calibration. It only causes issues. You will get far better results using a DCI-P3 baseline and asking the source (madVR) or VP (Radiance Pro) to discard the BT2020 container and target a DCI-P3 gamut. This is especially the case if you plan to use a large 3D LUT.

What matters is that when creating a 3DLUT, the software doesn't add posterization errors at the limit of the gamut, when there are no such errors without a 3D LUT or with a smaller LUT.

Again, most units with a P3 filter are not "forced in BT2020" provided your pattern generator and the JVCs are configured properly, and they don't need a large 3D LUT to produce excellent results with rec-709 and DCI-P3 calibrations, especially after using the free JVC Autocal software.

So please produce the details of the settings used for your tests (JVC model, hours on the lamp/panels, user mode selected, color profile selected, colorspace & chroma upsampling used with the pattern generator etc) so that it becomes easier to understand the context for your observations.

The aim is to use a LUT as large as necessary, to correct errors when necessary. It isn't to use the widest gamut that produces the largest errors and requires the largest LUT, especially when there is little to no consumer content that need such a wide gamut, and no consumer display that support such a wide gamut anyway.

Making it sound like the JVCs are forcing a wider gamut than their native gamut as if it was a flaw unique to them is disingenuous. Please could you list the consumer displays or projectors under $10,000 that offer a BT2020 mode that reaches 100% of BT2020, hence would produce a better baseline than the JVCs for a BT2020 or a DCI-P3 calibration?

Amongst the other projectors you mention, where does the JVC (with the P3 filter) stand compared to the others? Why single the JVCs out? Hopefully, if you are able to configure each projector in its optimal state, you should be able to confirm that a recent JVC (with a P3 filter, so RS5xx/6xx or rs2000/3000) produces a wider gamut than say a recent Sony, given that these don't have a P3 filter, hence have a significantly smaller native gamut. Given that you mention a few projectors, it would have been more fair to show the results for all of them, instead of singling the JVCs out, as if they were worse than the others. They are not, far from it.

In fact, could you show any consumer display under $10,000 that offers a BT2020 and does a better job than the unspecified model in an unspecified mode using an unspecified source you have used in your linked paper?

Besides this, thank you for working on improving the workflow in Lightspace for the user, I'm glad I (and others) kept asking .

Hopefully it will reduce the amount of posterization caused by Lightspace when generating a large 3D LUT if the target exceeds the capability of the display/projector, and it will make it much easier for the user to get good results than having to use a LUT concatenation process manually.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders

Last edited by Manni01; 05-13-2019 at 04:38 PM.
Manni01 is online now  
post #2215 of 2392 Old 05-13-2019, 04:13 PM
AVS Forum Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Blair, NE
Posts: 3,780
Mentioned: 315 Post(s)
Tagged: 0 Thread(s)
Quoted: 1535 Post(s)
Liked: 1837
If using a 3D LUT with LightSpace, then why put the PJ in the REC2020 colorspace - especially a JVC? The JVC projectors can do >98% P3. Both the Lumagen and madVR convert to P3. I do a Display Characterization with a P3 Target Space. I then do a Color Space Conversion from the P3 Color Space to the measured color space of the PJ. For peak brightness, I use the native gamut of the PJ. For maximum gamut I use Reference color space in the PJ with the filter in place (almost the same as the hack but customer friendly). The signal chain is then LUT device > P3 > PJ's color space.

Ponderosa Theater
High Impact AV
- ISF Level III Video Calibrator, THX-HAA HT3 Audio Designer/Calibrator
desertdome is offline  
post #2216 of 2392 Old 05-13-2019, 04:27 PM
AVS Forum Special Member
 
desertdome's Avatar
 
Join Date: Aug 2004
Location: Blair, NE
Posts: 3,780
Mentioned: 315 Post(s)
Tagged: 0 Thread(s)
Quoted: 1535 Post(s)
Liked: 1837
Quote:
Originally Posted by Manni01 View Post
You can get a very rec-709 or DCI-P3 baseline (especially when using the P3 filter) without having a "forced" BT2020 . . .
In the past month I've calibrated the past five generations (considering the RS4500 as a generation) of JVC projectors. They all do extremely well when avoiding the BT2020 color space. Here is the final LUT for an RS500. It does just as well as any newer JVC projector.

Attached Thumbnails
Click image for larger version

Name:	JVC P3 LUT.PNG
Views:	135
Size:	88.3 KB
ID:	2566328  
mrtickleuk likes this.

Ponderosa Theater
High Impact AV
- ISF Level III Video Calibrator, THX-HAA HT3 Audio Designer/Calibrator
desertdome is offline  
post #2217 of 2392 Old 05-13-2019, 04:31 PM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,110
Mentioned: 329 Post(s)
Tagged: 0 Thread(s)
Quoted: 5439 Post(s)
Liked: 5640
Quote:
Originally Posted by desertdome View Post
In the past month I've calibrated the past five generations (considering the RS4500 as a generation) of JVC projectors. They all do extremely well when avoiding the BT2020 color space. Here is the final LUT for an RS500. It does just as well as any newer JVC projector.

I agree, I have the same experience with my rs2000, and my rs500 also did very well when new. It doesn’t make any sense to use a BT2020 calibration if you have a LUT holder able to discard the container and target DCI-P3. This is why I don’t understand Steve’s statements and am asking him to provide more information to offer some context for his data. I can’t make any suggestions until Steve provides details about the model and settings used to reach these weird conclusions.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders
Manni01 is online now  
post #2218 of 2392 Old 05-13-2019, 11:33 PM
aka jfinnie
 
Join Date: Aug 2015
Location: Norwich, UK
Posts: 3,270
Mentioned: 53 Post(s)
Tagged: 0 Thread(s)
Quoted: 2641 Post(s)
Liked: 1739
Quote:
Originally Posted by desertdome View Post
In the past month I've calibrated the past five generations (considering the RS4500 as a generation) of JVC projectors. They all do extremely well when avoiding the BT2020 color space. Here is the final LUT for an RS500. It does just as well as any newer JVC projector.

Can you provide some more info about what your workflow was here? It looks like you must have pre-calibrated it before the LUT somewhat? I guess from the screenshot that this must be a REC709 calibration? What is the mode on the JVC that you calibrated?
bobof is offline  
post #2219 of 2392 Old 05-14-2019, 01:29 AM
AVS Forum Special Member
 
Light Illusion's Avatar
 
Join Date: Aug 2010
Posts: 1,834
Mentioned: 28 Post(s)
Tagged: 0 Thread(s)
Quoted: 779 Post(s)
Liked: 1144
Just to reiterate, the issue is with ANY display that is pre-calibrated to a colour space that is beyond the actual gamut capabilities of the display, and that 'calibration' cannot be disabled.
Using Rec2020 as the example just makes the process easier to follow/understand.

If the pre-calibration target colour space is close to the display's actual capabilities there will be less of an issue.

Ideally, all displays should be able to have any/all internal calibration disabled, returning the display to its native gamut.
Why that is not possible with such projectors is just beyond me...

But, regardless, using a reduced gamut patch sequence to profile any display that is pre-calibrated to any given colour space will likely generate better end results - based on all the testing we have performed - as the profiling patches are all focused within the actual gamut of the display.

Beyond this, even a display that is not pre-calibrated, but cannot make the target gamut fully, will benefit from a reduced gamut profiling approach for the same reasons.

In reality, this is just an extension to the LUT Concatenation approach, but taking the concept further by ensuring there are no (or at least less) redundant measurements within the profile data, so maximising the volumetric information provided to the LightSpace Colour Engine.

As always, we are just providing options for LightSpace users, as well as providing additional calibration knowledge.
What you do with this info is really up to you.



Steve
mikela likes this.

Steve Shaw
LIGHT ILLUSION


Last edited by Light Illusion; 05-14-2019 at 01:50 AM.
Light Illusion is online now  
post #2220 of 2392 Old 05-14-2019, 02:37 AM
AVS Forum Special Member
 
Manni01's Avatar
 
Join Date: Sep 2008
Posts: 9,110
Mentioned: 329 Post(s)
Tagged: 0 Thread(s)
Quoted: 5439 Post(s)
Liked: 5640
Quote:
Originally Posted by Light Illusion View Post
Just to reiterate, the issue is with ANY display that is pre-calibrated to a colour space that is beyond the actual gamut capabilities of the display, and that 'calibration' cannot be disabled.
Using Rec2020 as the example just makes the process easier to follow/understand.

If the pre-calibration target colour space is close to the display's actual capabilities there will be less of an issue.

Ideally, all displays should be able to have any/all internal calibration disabled, returning the display to its native gamut.
Why that is not possible with such projectors is just beyond me...

But, regardless, using a reduced gamut patch sequence to profile any display that is pre-calibrated to any given colour space will likely generate better end results - based on all the testing we have performed - as the profiling patches are all focused within the actual gamut of the display.

Beyond this, even a display that is not pre-calibrated, but cannot make the target gamut fully, will benefit from a reduced gamut profiling approach for the same reasons.

In reality, this is just an extension to the LUT Concatenation approach, but taking the concept further by ensuring there are no (or at least less) redundant measurements within the profile data, so maximising the volumetric information provided to the LightSpace Colour Engine.

As always, we are just providing options for LightSpace users, as well as providing additional calibration knowledge.
What you do with this info is really up to you.



Steve
Thanks Steve, I agree with the above, I have myself requested to JVC that profile off is made available with the filter enabled, so we can get the native gamut (no processing) and the widest native gamut as well. And as I said in my earlier post, I applaud this new mode in Lightspace (why wouldn't I, I spent almost a year requesting a better way to deal with an undersaturated display than the convoluted LUT concatenation process!) and am looking forward to testing it when it comes out of beta, as it will be useful in some situations.

What I don't understand is your claim that the BT2020 calibration cannot be disabled on the JVCs, and why you are using this to show large errors when targeting BT2020.

The BT2020 calibration is only applied in the JVCs if you select the BT2020 color profile (with P3 filter) or the HDR color profile (without P3 filter).

If you select the DCI-P3 color profile with the filter in the new models (reference in older models), you get a calibration that covers close to 100% of DCI-P3, and wouldn't show the large errors in your paper. A large LUT with this baseline wouldn't show any of the large errors you show in your paper.

If you select the Rec-709 color profile without the filter in the new models (standard in older models) or profile off, you get a calibration that covers close to 100% of rec-709, without any of the large errors either. If it's undersaturated, you can usually get 100% of rec-709 by uploading and selecting a custom rec-709 color profile using the P3 filter (Rec-709-F). Again, a large LUT with this baseline wouldn't show any of the large errors you show in your paper.

So, to reiterate:

1) I don't understand your claim that "the JVCs use a BT2020 color profile that is forced". This simply isn't true.
2) I don't understand why you single the JVCs out as if they were producing some kind of horrible results, when they are actually producing excellent results, and the calibration you used in your paper would reproduce most UHD content without any issue if no 3D LUT is applied
3) Yes the issue is when applying a large 3D LUT that causes issues near the limit of the native gamut. It is good that Lightspace is making it easier to get good results in this situation, but please stop making it look like the JVCs are to be singled out.
4) If the above is incorrect, please post an example of a consumer display or projector under $10,000 that offers a BT2020 mode and does better than a recent JVC using the P3 filter.

I really suggest that you reword your paper, or at least provide more information about source, settings and models, as well as examples from other consumer projectors or consumer displays doing better in the same situation (I don't think you'll find any).

I think you should also illustrate that the JVCs, when targeting a gamut that doesn't exceed the target gamut, produce excellent results if the correct factory or custom color profile is selected. Otherwise, it means that you have an agenda or are not able to set them up properly.

Again, as it stands, this Lightspace paper is disingenuous and misleading, and technically incorrect. It seems to be designed to make the JVCs look bad, as if they all have large errors that need to be corrected. This is simply not true in most cases, provided you know how to set them up correctly, which a competent calibrator familiar with these units ( @Chad B , @Kris Deering and others) will be able to do.

Batch Utility V4.02 May 16 2019 to automate measurements files for madVR with support for BD Folders

Last edited by Manni01; 05-14-2019 at 03:05 AM.
Manni01 is online now  
Sponsored Links
Advertisement
 
Reply Display Calibration

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


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