or Connect
AVS › AVS Forum › Display Devices › Display Calibration › LightSpace CMS Now Supports Lumagen + eeColor 3D-LUT 4 All
New Posts  All Forums:Forum Nav:

LightSpace CMS Now Supports Lumagen + eeColor 3D-LUT 4 All - Page 20

post #571 of 710
Just out of a morbid curiosity, how long would it take with a ID3 pro?
post #572 of 710
It does depend, but probably 5 to 7 hours...
The variable is the display.
This is actually where the User Patch sequence capability helps, as you can vary the order of the patch sequence and randomly mix bright patches with dark.
That will help with the stability of the display over the extended calibration period.
We can (and probably will) change the standard patch set to benefit from this, but right now having the user patch capacity is a bit of fun tongue.gif

Steve
post #573 of 710
Quote:
Originally Posted by ConnecTEDDD View Post


21-Point Cube (9.261 Color Points) which requires 3 hours and 13 minutes (DIP Mode 1sec. per Patch) is a good idea to include it to future version of my disk...strictly for Klein K-10A users.

Hello Ted.

Are you using any extra delay with your K10-A for your Kuro.?

ss
post #574 of 710
Thread Starter 
Quote:
Originally Posted by sillysally View Post

Quote:
Originally Posted by ConnecTEDDD View Post


21-Point Cube (9.261 Color Points) which requires 3 hours and 13 minutes (DIP Mode 1sec. per Patch) is a good idea to include it to future version of my disk...strictly for Klein K-10A users.

Hello Ted.

Are you using any extra delay with your K10-A for your Kuro.?

ss

Hello SS,

When I'm using DIP Mode 1sec (with our known fixed 0.5sec internal LS delay before each meter read) I don't use extra delay.

Now Steve added 1.5sec DIP Mode so i will use 0.5sec extra delay from now wink.gif
post #575 of 710
Quote:
Originally Posted by Wouter73 View Post

Just out of a morbid curiosity, how long would it take with a ID3 pro?

For 17x17x17 points with a JVC X35 video projector with WHITE100 at 14 FL in my very dark room it takes 5H.

I really need a KLEIN K10A frown.gif
post #576 of 710
Quote:
Originally Posted by alex_t View Post

For 17x17x17 points with a JVC X35 video projector with WHITE100 at 14 FL in my very dark room it takes 5H.

I really need a KLEIN K10A frown.gif
Ask santa claus for a new k10A biggrin.gif
post #577 of 710
Hi Guys:

Merry Christmas and Happy New Year to all Users! I have an example of a real application for using Light Space 3D LUT mods with Photoshop, which I think you’ll find interesting. I took a screen grab of a nearly hidden space ship in the film Prometheus and lifted just the colors of the space ship’s hull so one can see it more easily. Pls refer to the images below. Image 1 is the nearly invisible space ship “Before” the Photoshop mod is applied. Image 2 is “After” the Photoshop mod is applied. Pls click on the Images to see them better. You'll need a dark environment to see the changes in Image 2. Notice the blacks are still black. tongue.gif

One point about the word “Images”. Images are just pictures, period. They can be pictures by themselves (e.g. .tif file), or pictures inserted into a LUT, or a Calibration LUT saved in Light Space as a .tif file by using the “Save As” command and appending .tif to its file name. A calibration LUT saved through Light Space as a .tif file is an image/picture … of a CLUT … and can, therefore, be modified by Photoshop like a picture. This is my understanding of a "LUT Image".

One other point I think you’ll find interesting. Light Space has 3D LUT math tools for the user. The 2 math tools I used in this mod were 3D LUT Addition and Apply Calibration LUT to Image (i.e. picture). Basically, you can add, subtract LUTs in 3D with Light Space. Very cool stuff.

Again, think of the possibilities of tweaking Calibration LUTs. Not only can you tweak the Calibration LUT for your display, you can also tweak a single image from a film to make the specific film look better on your display. One does this by tweaking your existing Calibration LUT with Photoshop. All you do is insert the film's screen grab into the Caibration LUT and use Photoshop to change HSL or brightness and contrast. Tweakers nirvana! looool. Because of the way the math is designed in Light Space, you do not have to make a tweak and re-run the calibration. These tweaks are all done post calibration. The math is done is seconds ... so math does matter. wink.gif I’m still learning this myself … it’s like Magic. If I can figure out how to do this on my own, anyone can. See this link for more information about Photoshop mods: http://www.lightillusion.com/lut_image.html

Kind Regards and good LUT’g and tweaking,

JJ wink.gif



Image 1. Before modification.


Image 2. After modification.
Edited by Carbon Ft Print - 12/26/13 at 4:31pm
post #578 of 710
I ran a couple of profiles tests on my home LCD screen with the i1D3.

21^3 profile in 3 Hours, 50 Minutes, 20 Seconds.

17^3 profile in 1 Hour, 29 Minutes, 57 Seconds.

I used Closed Loop mode, with 0.75 secs integration time, and no ALL.

What I found very interesting was the increase in calibration accuracy with the 21^3 profile.
When we have run such tests previously all calibrations suggest 17^3 was optimal.
However, that was some time back, and the quality of probes has increased since then, so it really does look like 21^3 is a seriously valid cube size for profiling.






Also, I ran test with the non-linear cube data sets linked to by Graeme - we do have a plot error in our graphs, but the generated LUTs were exactly as they should be, as should be the case with the colour engine within LightSpace not needing linear spaced patch data sets.

We'll get the graphs fixed ASAP biggrin.gif
post #579 of 710
Quote:
Originally Posted by Light Illusion View Post

I ran a couple of profiles tests on my home LCD screen with the i1D3.

21^3 profile in 3 Hours, 50 Minutes, 20 Seconds.



Typo - should be 2 hours, not 3. wink.gif
post #580 of 710
LOL - thanks for that correction biggrin.gif
post #581 of 710
Quote:
Originally Posted by Light Illusion View Post

I ran a couple of profiles tests on my home LCD screen with the i1D3.

I used Closed Loop mode, with 0.75 secs integration time, and no ALL.

What I found very interesting was the increase in calibration accuracy with the 21^3 profile.
When we have run such tests previously all calibrations suggest 17^3 was optimal.
However, that was some time back, and the quality of probes has increased since then, so it really does look like 21^3 is a seriously valid cube size for profiling.

Steve when are you going to release a tested version of the extended profile (21^3), I am chomping at the bit. smile.gif

ss
Edited by sillysally - 12/27/13 at 12:30pm
post #582 of 710
ss, you have mail!

biggrin.gif
post #583 of 710
Quote:
Originally Posted by Light Illusion View Post

ss, you have mail!

biggrin.gif

Oh darn the mail must be slow. frown.gif

Thanks for the very quick try, I will keep checking.smile.gif

ss
post #584 of 710
If anyone wants to try the beta version drop me a line.

ss, the link I sent you had an extra 'dot' at the end (before the 'zip' extension rolleyes.gif
post #585 of 710
So if i use the home version of LightSpace with an eecolor box and Ted's calibration bluray, i would need to calibrate grayscale manually correct?
post #586 of 710
Quote:
Originally Posted by Light Illusion View Post

If anyone wants to try the beta version drop me a line.

ss, the link I sent you had an extra 'dot' at the end (before the 'zip' extension rolleyes.gif

Thanks again for your great turn-around time Steve.
Quote:
Originally Posted by njfoses View Post

So if i use the home version of LightSpace with an eecolor box and Ted's calibration bluray, i would need to calibrate grayscale manually correct?

You may not need to do a grayscale, just setup your pre calibration as Steve suggests. Some of us do a grayscale mainly because of are displays, but even then i would suggest just for one or two points of your GS to start with. In any case I am sure you will love the outcome of your Profile/LUT.
Don't forget there are many of us here that will get you back on track if you have any questions, and of-course Steve is very fast to respond to any questions you may have. smile.gif

ss
post #587 of 710
Thread Starter 
Quote:
Originally Posted by njfoses View Post

So if i use the home version of LightSpace with an eecolor box and Ted's calibration bluray, i would need to calibrate grayscale manually correct?

Hello, for the best performance, it's highly recommended to make your 100% White as perfect as possible, in case that by tweaking your display controls wil not introduce non-linearity problems / problems in Color Ramps***) and make your RGB Separation Chart worse. You can download the Free DPS version of LightSpace to take some readings using your meters; use LightSpace's Quick Profiling Primary Colors measurements and show us the pictures to quide you.

*** This is the reason i have added so many color reproduction patterns to my disk (152), for users to check any possible display anomalies that internal controls are producing but meters can't report; becasue users are measing only a few color points.

Many of the users are reporting discolorations in grayscale ramps but their chart reports are showing errors below 0.5 dE, that's why anyone need to use some evaluation patterns to find out it's display's Safe Range of tweaking of it's internal controls wink.gif

You can download the LightSpace Demo from here, and then request a LightSpace DPS (Display Profiling System) Free Version License from here.
post #588 of 710
Quote:
Originally Posted by sillysally View Post

Thanks again for your great turn-around time Steve.
You may not need to do a grayscale, just setup your pre calibration as Steve suggests. Some of us do a grayscale mainly because of are displays, but even then i would suggest just for one or two points of your GS to start with. In any case I am sure you will love the outcome of your Profile/LUT.
Don't forget there are many of us here that will get you back on track if you have any questions, and of-course Steve is very fast to respond to any questions you may have. smile.gif

ss

Quote:
Originally Posted by ConnecTEDDD View Post

Hello, for the best performance, it's highly recommended to make your 100% White as perfect as possible, in case that by tweaking your display controls wil not introduce non-linearity problems / problems in Color Ramps***) and make your RGB Separation Chart worse. You can download the Free DPS version of LightSpace to take some readings using your meters; use LightSpace's Quick Profiling Primary Colors measurements and show us the pictures to quide you.

*** This is the reason i have added so many color reproduction patterns to my disk (152), for users to check any possible display anomalies that internal controls are producing but meters can't report; becasue users are measing only a few color points.

Many of the users are reporting discolorations in grayscale ramps but their chart reports are showing errors below 0.5 dE, that's why anyone need to use some evaluation patterns to find out it's display's Safe Range of tweaking of it's internal controls wink.gif

You can download the LightSpace Demo from here, and then request a LightSpace DPS (Display Profiling System) Free Version License from here.

Thanks. I have been a calman user for a while now, with an i1pro. I have a mits dlp that has full cms controls but the display is not even close to being linear at various saturation levels which makes the cms controls basically useless. I have been debating the lumagen with calman option vs the lightspace hcl, Ted's disc, and eecolor box option. I have zero use for the lumagen upconverting, scaling features etc and reading about LightSpace has intrigued me. I know regardless i will need a new meter (i1display pro most likely) since my i1pro needs the periodic dark readings. My thought is that i can then profile the new meter to get the best of both worlds. Just need to figure out the best way to go.
post #589 of 710
Thread Starter 
Quote:
Originally Posted by njfoses View Post

Thanks. I have been a calman user for a while now, with an i1pro. I have a mits dlp that has full cms controls but the display is not even close to being linear at various saturation levels which makes the cms controls basically useless. I have been debating the lumagen with calman option vs the lightspace hcl, Ted's disc, and eecolor box option. I have zero use for the lumagen upconverting, scaling features etc and reading about LightSpace has intrigued me. I know regardless i will need a new meter (i1display pro most likely) since my i1pro needs the periodic dark readings. My thought is that i can then profile the new meter to get the best of both worlds. Just need to figure out the best way to go.

Hello, your first step is to get a i1Display PRO (ID3) OEM Version that can work with CalMAN, LightSpace and other software solutions. LightSpace is not supporting the id3 retail meter.
post #590 of 710
Quote:
Originally Posted by ConnecTEDDD View Post

Hello, your first step is to get a i1Display PRO (ID3) OEM Version that can work with CalMAN, LightSpace and other software solutions. LightSpace is not supporting the id3 retail meter.

Yea i figured i needed the oem version of the meter. My dlp also has an auto-iris which makes gamma a pain to dial in. Hoping LS can help in this regard.
post #591 of 710
Quote:
Originally Posted by njfoses View Post


Thanks. I have been a calman user for a while now, with an i1pro. I have a mits dlp that has full cms controls but the display is not even close to being linear at various saturation levels which makes the cms controls basically useless. I have been debating the lumagen with calman option vs the lightspace hcl, Ted's disc, and eecolor box option. I have zero use for the lumagen upconverting, scaling features etc and reading about LightSpace has intrigued me. I know regardless i will need a new meter (i1display pro most likely) since my i1pro needs the periodic dark readings. My thought is that i can then profile the new meter to get the best of both worlds. Just need to figure out the best way to go.

I have a Lumagen 2041 and have compared CM's 21 point GS and 929 point LUT loaded into the 2041 to LS 4913 point profile/LUT loaded into the eecolor box.
My preference is LS 4913 doing a manual 10Pt GS using my VT60 controls, then loading the LUT into the eecolor box.

Two things to consider, eecolor box will not process a 3D movie and it is limited to what it can output for audio.

Your auto-iris may also be a problem no mater what software you use. Check it out, maybe ask Steve.

As Ted suggest, the i1Display PRO (ID3) OEM Version would be a nice addition to your meters. Use your I1Pro to profile your ID3, no matter what software you use.

Good luck,

ss
post #592 of 710
Quote:
Originally Posted by sillysally View Post

I have a Lumagen 2041 and have compared CM's 21 point GS and 929 point LUT loaded into the 2041 to LS 4913 point profile/LUT loaded into the eecolor box.
My preference is LS 4913 doing a manual 10Pt GS using my VT60 controls, then loading the LUT into the eecolor box.

Two things to consider, eecolor box will not process a 3D movie and it is limited to what it can output for audio.

Your auto-iris may also be a problem no mater what software you use. Check it out, maybe ask Steve.

As Ted suggest, the i1Display PRO (ID3) OEM Version would be a nice addition to your meters. Use your I1Pro to profile your ID3, no matter what software you use.

Good luck,

ss

Thanks for the info. I have an Integra DTR 80.3 avr that i use for all my audio needs and input switching and then run one hdmi to my display. I am assuming i can have the ee box as the final piece in the chain after the avr and before the display?
post #593 of 710
Quote:
Originally Posted by njfoses View Post


Thanks for the info. I have an Integra DTR 80.3 avr that i use for all my audio needs and input switching and then run one hdmi to my display. I am assuming i can have the ee box as the final piece in the chain after the avr and before the display?

That is correct.
post #594 of 710
Auto-iris is a lot like Plasma ABL, and you will need to treat the projector in the same way.
You will be able to calibrate accurately with a set of given patches, but when you then move on to 'real' images the calibration will adjust due to the auto-iris.

I would suggest running calibration a number of times with different patch sizes, and then select the best result when looking at your favourite real images.

I would also suggest having LS on a Laptop with HDMI out, and using the direct patch generation in closed-loop mode - and then verify via Ted's disc.
We use laptops with the Intel 4000 graphics chip set - they output perfect image levels (as actually do most modern graphics cards with no ICC profile rubbish going on).

I would suggest LightSpace HCL with the I1d3 OEM, and profiled/matched to your i1Pro.

Steve
post #595 of 710
Quote:
Originally Posted by Light Illusion View Post

ss, you have mail!

biggrin.gif

Ran the new LS Beta last night, 21^3. Solid as a rock, took only about 2:47 to complete a 21^3 profile, use .50 extra delay LLH off. Seemed to add a little bit more color detail (if that's possible) in flesh tones and over all detail color shading.

The 21^3 profile is now my new standard. smile.gif

Thanks Steve.

ss
post #596 of 710
Quote:
Originally Posted by sillysally View Post

Quote:
Originally Posted by Light Illusion View Post

ss, you have mail!

biggrin.gif

Ran the new LS Beta last night, 21^3. Solid as a rock, took only about 2:47 to complete a 21^3 profile, use .50 extra delay LLH off. Seemed to add a little bit more color detail (if that's possible) in flesh tones and over all detail color shading.

The 21^3 profile is now my new standard. smile.gif

Thanks Steve.

ss

If you get time, try it without the extra delay for a comparison. It is certain that different displays have different needs. The K10 itself doesn't require "extra" time which we all know. Which display, your VT60?
post #597 of 710
It has been a bit if a revelation adding the 21^3 capability (9261 points in total)... biggrin.gif

All previous tests had shown 17^3 to be optimal, but the last test we ran were a year or so back, and the improvements in display and probe technology in that time appear to have changed the rules.
All the tests we have now performed, with different displays and different probes, including the more common i1D3, have shown significant improvements in final calibration with 21^3 profile sizes.
Interestingly, the additional patch count has reduced the requirement on using 'average low light' measurements too, as the increased quantity of patches result in the same end effect.

We are now running a lot of tests on non-regular patch sequences (with a number of the beta testers doing the same) and are waiting on the final verdict compared to a standard 21^3 profile.

As for the 'extra delay' function, that is a potential requirement of the display, allowing time for the display to 'settle' after the patch colour changes.
We use patch sequences in a 'specific colour order' to minimise this need, but it is very possible that the same patches in a more random order may work better with some displays... specifically plasmas, as a random order of bright/dim/bright/dim... patches may help with ABL issues.

With the user generated patch sequence capability that would be easy to try.

Steve
post #598 of 710
Quote:
Originally Posted by Light Illusion View Post


We use patch sequences in a 'specific colour order' to minimise this need, but it is very possible that the same patches in a more random order may work better with some displays... specifically plasmas, as a random order of bright/dim/bright/dim... patches may help with ABL issues.

With the user generated patch sequence capability that would be easy to try.

Steve

Hi Steve:

So how's the user to do this?

Kind Regards,

JJ
post #599 of 710
Quote:
Originally Posted by buzzard767 View Post

If you get time, try it without the extra delay for a comparison. It is certain that different displays have different needs. The K10 itself doesn't require "extra" time which we all know. Which display, your VT60?

I ran a second 21^3 profile last night (2:43), with just setting the 100% white. Results are very good, on paper not quite as good as a full 10Pt but close.
Yes its my 65VT60.
The one ting that is impressing me with the 21^3 profile, is how close my WRGBCMY are lining up on a normal 1931 6 Pt Gamut chart and numbers. This is the case with a full 10Pt or just a 100% white 1pt grayscale.

So you don't think I need to use any extra delay with my K10-A when profiling a VT60.?

ss
post #600 of 710
Quote:
Originally Posted by Carbon Ft Print View Post

Hi Steve:

So how's the user to do this?

Kind Regards,

JJ
JJ, use Excel - export the existing patch sequence from LightSpace, and then re-order the patches within Excel as you desire.
You should be able to generate a Macro for that function.
Then load back into LightSpace as an .csv list.

biggrin.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Display Calibration
AVS › AVS Forum › Display Devices › Display Calibration › LightSpace CMS Now Supports Lumagen + eeColor 3D-LUT 4 All