Get an exclusive sneak peek into our new project. >>> Click Here
Our native mobile app has a new name: Fora Communities. Learn more.

Display Calibration Tools - New Site For LUT Analysis and Enhancement

4130 Views 40 Replies 11 Participants Last post by Iron Mike, Sep 26, 2015

buzzard767

Discussion Starter · Feb 18, 2014

DisplayCalibrationTools.com is a new Internet site dedicated to making LUT's better. The opening paragraph explains the purpose.

Professional Display Profiling & Display Calibration Tools

Display Calibration Tools offers dedicated profiling and calibration workflow tools that enhance and improve your 3D LUT calibration results. Our tools not only go beyond vendor based limitations but provide functionality that unleash the true potential of professional calibrations solutions such as Lightspace CMS and Argyll.

I've just begun to look at the possibilities. As a LightSpace user I'm particularly interested and the site should attract Argyll users as well.

See less See more

Reactions: 2

1 - 20 of 41 Posts

sillysally

· #2 · Feb 18, 2014

3

Just ran a 21^3 + 21 point GS total patches 9,324.

Used the Cycle dark/light as one of my options, seems to have less patch IR and not discolored.

Next time I need to add a little higher % for Weighted.

Took 3:13 to run the 9,324 points setting my K10-A at 3/4 of a second.

All in all the LUT looks good.

ss.

btw, I had a little problem oping my patch set because of the compression file type, downloaded 7-Zip and all is well.

See less See more

Reactions: 1

Cynephyle

· #3 · Feb 18, 2014

SillySally,

U can extract the downloads also via standard Winzip or Winrar.

The patch weight option is a uniform weight, you can see on the graphs how the weight affects the distribution of the saturation and brightness points.

Cynephyle

· #4 · Feb 18, 2014

Btw, did u profile a Plasma display ?

Gordon Fraser

· #5 · Feb 18, 2014

I had a play with these tools on Sunday and did a quick 9x9x9 profile on a JVC X500 at a dealers in UK. I used a 65% weighted custom patch sequence with, i think, 41 greyscale steps. I chose to run dark to light patch sequence with it though purely because it allowed me to judge when it was safe to nip out the door to get a coffee without any fear i'd contaminate the readings, hahahaha.....end result was pretty fantastic. I like the reporting tool as well. Going to try to test another custom patch sequence on my plasma today in the house

Reactions: 2

Cynephyle

· #6 · Feb 18, 2014

Quote:
Originally Posted by Gordon Fraser /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24373460

Going to try to test another custom patch sequence on my plasma today in the house

For Plasma profiling an additional concern is obviously ABL kicking in sooner or later (and temporary patch IR) during profiling.

Regarding ABL, the cycle dark/bright sort option is a nice approach to hinder/reduce ABL. Another approach is to sort all patches brightest to darkest, so the longer in the profile run the darker the patches, which might also reduce ABL.

I did a custom weighted 17^3 with custom 26 point Greyscale (patches arranged brightest to darkest) the other day and it resulted in the best LUT calibration ever on my Panny plasma...

Reactions: 1

Light Illusion

· #7 · Feb 18, 2014

2

Mike has done a great gob with this website - with a bit of help from Light Illusion

But the ability to generate reports and user patch sequences is a great addition, and will help improve calibration results further.

Keep an eye on the website as Mike will be adding new capabilities as time progresses.

Mike - congratulations!

See less See more

Reactions: 2

Cynephyle

· #8 · Feb 18, 2014

2

As always, Light Illusion support is 24/7/365.... even when it is not about Lightspace.

I guess the most accurate and most transparent solution out there just got a tiny bit better....

See less See more

Reactions: 1

ConnecTEDDD

· #9 · Feb 18, 2014

Incredible Work ! These new Display Calibration Tools in hands of LightSpace users are bringing the profiling to a new level.

Reactions: 1

Harry

· #10 · Feb 18, 2014

3

I'm really impressed ... this tools add valuable flexibility to LS!

Just have spent only a short time playing with ... already now I like the Custom Color Patch Sequence Generator (short: CCPSG). In a simple way it allows everyone to make individual profilings with other characteristics / weightings / sequence order / ... really good job, Mike!

Just one little note for everybody else "beginners" (I had a tiny problem ...

)

The CCPSG gives you also information about the number of Total Profile Points of your individual settings. Basically this number is consistent with LightSpace's indication of "Total Frames" under "Display Characterization" -> "CSV File" (to import own patch sequences) as long as you don't use a custom greyscale.

I want to note that, because first time I've used the CCPSG with option "Custom Greyscale Points", I've not notized the dynamically added word "app." next to "Total Profile Points", meaning approximately. So I was a little bit confused about the difference of "Total Profile Points" CCPSG told me, compared with the "Total Frames" number, LS has displayed ...

Mike explained that behaviour to me, so I want to share it:

The dynamic calculation of CCPSG is just approximate, since a custom greyscale will (possibly, very likely) generate duplicate patches that the LS code then later eliminates.

See less See more

Reactions: 1

sillysally

· #11 · Feb 19, 2014

Quote:
Originally Posted by Cynephyle /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24373425

Btw, did u profile a Plasma display ?

Yes the display is a Plasma 65VT60

ss

Reactions: 1

Cynephyle

· #12 · Feb 19, 2014

Quote:
Originally Posted by *Harry* /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24375305

The dynamic calculation of CCPSG is just approximate, since a custom greyscale will (possibly, very likely) generate duplicate patches that the LS code then later eliminates.

Yes, once a custom greyscale is chosen in the CPSG, the tool adds the corresponding R|G|B patches so that solutions like Lightspace can do a RGB Separation evaluation and populate corresponding graphs.

These additional patches might be duplicates to patches that are already generated by the grid sequence so the code later eliminates these duplicates.

You will always see the exact number of patches in a patch set once u display the patch set (before u save it) in the Profile Overview stats.

Reactions: 2

Harry

· #13 · Feb 20, 2014

2

Another question:

Custom Color Patch Sequence Generator:

What does the labeling on the X/Y axis on this 1D Grid Sequence graph represent ?

See less See more

Reactions: 1

Iron Mike

· #14 · Feb 25, 2014

Quote:
Originally Posted by *Harry* /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24383492

Another question:

Custom Color Patch Sequence Generator:

What does the labeling on the X/Y axis on this 1D Grid Sequence graph represent ?

Harry, I can answer this:

The x/y axis in the image you posted represent the RGB level, ergo in 8-bit 0-255.

A RGB patch that is linear would display on a graph similar to the one you refer to as, e.g 128/128 for a 50% grey.

A RGB patch that displays as 128/100 is weighted, it has a specific weight attached to it, so the brightness is lower as the linear version: 50% brightness vs. 39% brightness.

See less See more

Reactions: 1

sillysally

· #15 · Feb 28, 2014

Quote:
Originally Posted by Iron Mike /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24404821

Harry, I can answer this:

The x/y axis in the image you posted represent the RGB level, ergo in 8-bit 0-255.

A RGB patch that is linear would display on a graph similar to the one you refer to as, e.g 128/128 for a 50% grey.

A RGB patch that displays as 128/100 is weighted, it has a specific weight attached to it, so the brightness is lower as the linear version: 50% brightness vs. 39% brightness.

Yes N/Q you have a great tool Mike.

One thing your above graft has brought to mind is the difference between the weighted patch set or a lenear patch set.

I have been playing with the zoyd's posted patch set (2501 points). What I am thinking is that zoyd's posted patch set is lenear not weighted.

In this post using zoyd's posted patch set you will see how even the RGB balance and gamma is. http://www.avsforum.com/t/1517849/a-comparison-of-3dlut-solutions-for-the-eecolor-box/30#post_24367983

However the down side of that patch set is a bit unstable for Lum and Sat when compared to Steve's non beta patch sets.

I have ran that patch set 3 times, the results of the second time I ran it was not very good, the third time I ran it the results as far as dE's go's was very good after touching up the Grayscale.

When I say touching up the grayscale, I only use Red and Blue never Green or any Gamma adjustments.

So my question is would that be the difference between a weighted as opposed to a linear patch set. ?

ss

See less See more

Reactions: 1

Iron Mike

· #16 · Feb 28, 2014

Quote:
Originally Posted by sillysally /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24419859

Yes N/Q you have a great tool Mike.

One thing your above graft has brought to mind is the difference between the weighted patch set or a lenear patch set.

I have been playing with the zoyd's posted patch set (2501 points). What I am thinking is that zoyd's posted patch set is lenear not weighted.

In this post using zoyd's posted patch set you will see how even the RGB balance and gamma is. http://www.avsforum.com/t/1517849/a-comparison-of-3dlut-solutions-for-the-eecolor-box/30#post_24367983

However the down side of that patch set is a bit unstable for Lum and Sat when compared to Steve's non beta patch sets.

I have ran that patch set 3 times, the results of the second time I ran it was not very good, the third time I ran it the results as far as dE's go's was very good after touching up the Grayscale.

When I say touching up the grayscale, I only use Red and Blue never Green or any Gamma adjustments.

So my question is would that be the difference between a weighted as opposed to a linear patch set. ?

ss

the difference between linear and weighted patch sets is the difference in the 1D patch (1D patch == single R|G|B value) set that is then used to create the 3D patches (--> aka RGB triplets). The original LS 1D patch sequence is also weighted, with a custom weight - for a reason.

You can use any and play with any weight that u see fit, and just to give you an idea what the weight actually does, it's very simple: since it changed the single R|G|B vals of the 1D patch sequence it will now also change the values of the RGB triplets (--> aka 3D patches) that are being created from the 1D patches....

simply look out for how the hue, saturation and brightness coverage and distribution changes. Example: look for how many profile points are pushed into 0-25, 25-75 and 75-100%.

You want a distribution and coverage similar to what a common TV signal / BD disc content would represent.

- M

See less See more

Reactions: 1

sillysally

· #17 · Feb 28, 2014

Quote:
Originally Posted by Iron Mike /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24419932

the difference between linear and weighted patch sets is the difference in the 1D patch (1D patch == single R|G|B value) set that is then used to create the 3D patches (--> aka RGB triplets). The original LS 1D patch sequence is also weighted, with a custom weight - for a reason.

You can use any and play with any weight that u see fit, and just to give you an idea what the weight actually does, it's very simple: since it changed the single R|G|B vals of the 1D patch sequence it will now also change the values of the RGB triplets (--> aka 3D patches) that are being created from the 1D patches....

simply look out for how the hue, saturation and brightness coverage and distribution changes. Example: look for how many profile points are pushed into 0-25, 25-75 and 75-100%.

You want a distribution and coverage similar to what a common TV signal / BD disc content would represent.

- M

Thanks Mike for your very nice explanation.

What I was really driving at was comparing in real world time, how the different ways of going about making patch sets effects the outcome. In the above I am comparing the way Graeme Gill/zoyd writes there's for ArgyllCMS and how Steve wrights his. Weighing the plus and mines in real world calibration as opposed to theory.

ss

Reactions: 1

Iron Mike

· #18 · Feb 28, 2014

Quote:
Originally Posted by sillysally /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24420009

Thanks Mike for your very nice explanation.

What I was really driving at was comparing in real world time, how the different ways of going about making patch sets effects the outcome. In the above I am comparing the way Graeme Gill/zoyd writes there's for ArgyllCMS and how Steve wrights his. Weighing the plus and mines in real world calibration as opposed to theory.

ss

I can't speak for them, but the approach would be to create a patch set that "scans" the gamut as conclusive as possible, as in: report as much info about all parts of the gamut.

The more conclusive info about a display, the easier for the color engine to calculate appropriate compensations.

Reactions: 2

sillysally

· #19 · Feb 28, 2014

Quote:
Originally Posted by Iron Mike /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24420022

I can't speak for them, but the approach would be to create a patch set that "scans" the gamut as conclusive as possible, as in: report as much info about all parts of the gamut.

The more conclusive info about a display, the easier for the color engine to calculate appropriate compensations.

Yes Mike, what I would really like to see is Graeme come up with a larger patch set like 5000 points, Then compare the stability of the two patch sets using something like zoyd's 1000 point checker along with CM's checkers. .

ss

DrFaxe

· #20 · Feb 28, 2014

Quote:
Originally Posted by sillysally /t/1518452/display-calibration-tools-new-site-for-lut-analysis-and-enhancement#post_24421209

Yes Mike, what I would really like to see is Graeme come up with a larger patch set like 5000 points, Then compare the stability of the two patch sets using something like zoyd's 1000 point checker along with CM's checkers. .

ss

Why don't create Your own? All You need is Argyll. Zoyd showed all command line switches in the post here:
http://www.avsforum.com/t/1517849/a-comparison-of-3dlut-solutions-for-the-eecolor-box#post_24358174

If You like to precondition the set for the native or REC.709 gamut and don't want to do any measurements with argyll You can use a synthetic icc profile.

1 - 20 of 41 Posts

This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.