| Forums | Polls | Register | Reply | Search | Statistics |
Display Calibration Light Illusion Forums / Display Calibration /   
 

3D LUTs for the home enthusiast

 
 
Page  Page 3 of 8:  « Previous  1  2  3  4  5  6  7  8  Next »

Author buzzard767
User - Mono 
#31 | Posted: 4 Oct 2012 20:31 
I must have a workflow error. Using the ee ColorBox, LUTs produced from CalMan are okay but those from LightSpace are terrible. Grayscale, no, Greenscale is so bad that there is no need looking further.

At first I thought it might be a meter problem so I switched from a D3 to an i1Pro but got the same results. LightSpace GS patches look normal with none of the coloration found in the finished product. To eliminate the possibility of a computer icc glitch using the LightSpace patches I did a run each of the Calman and LightSpace LUTs using internal patterns generated within ChromaPure software with no problems.

Here's the basics of what I did so please correct me:
1. Display Characterization - run 5p cube so as to approximate the Calman cube size
2. Color Space Conversion - pick a name, Color Space Rec.709, Destination - the 5p cube from #1, Peak Luma
3. Export the resulting LUT to the ee LUT holding folder in Windows
4. Change the name to match what the ee Box is expecting and load the ee Box

Author Steve

Mod - Scene 
#32 | Posted: 4 Oct 2012 20:44 
Hi Buzz,

Stick to Quick Profiling while your are evaluating the workflow.
Its quicker and easier.
And a 5 point LUT via Display Characterisation will only work on displays that have a good underlying setup.
Did you test the display via Quick Profiling first?
Talk to JJ as he has worked hard on understanding this bit

You can always test the LUTs in LightSpace (LUT Preview tool) to make sure the generated LUT is the problem, and not the eeColor box.

See the http://www.lightillusion.com/idiots_guide.html page for a good workflow guide to calibration.

And ask more questions!

Cheers,

Steve
Steve Shaw
Mob Boss at Light Illusion

Author buzzard767
User - Mono 
#33 | Posted: 6 Oct 2012 13:40 
In Quick Profile or Display Calibration there is a choice between DIP and Closed Loop timing. What is the reason NOT to use Closed Loop? It would seem this would be the fastest way to measure.

Author Steve

Mod - Scene 
#34 | Posted: 6 Oct 2012 16:48 
Hi Buzz,

DIP mode is for when the patches are being generated via a separate timeline or pre-built clip.
When using extended desktop mode there is no reason to use DIP - use Closed Loop mode.

The on-line user manuals do describe the different application, so a search for DIP mode via the website search facility should come up with more info.

Also, sometimes there may be USB interrupt issues, and selecting DIP mode can overcome those as it allows a fixed duration that can be set to be longer then the probe read time, so allowing data transmission headroom.

Closed Loop is the fastest mode, and if it can be used, it should be

Steve
Steve Shaw
Mob Boss at Light Illusion

Author JJ Johnson
User - Mono 
#35 | Posted: 7 Oct 2012 17:16 | Edited by: JJ Johnson 
Hi Steve,

My Sony 46” TV I’ve been testing does not have good color separation … which is needed for Quick Profile. I have to do an 11pt Display Characterizations instead. In the field, this would be the difference between an 8 min Quick Profile LUT cal and a 1 hour (11pt) Display Characterization LUT cal.

With my display there is an interesting direct relationship between Luminance level and Color Separation. Color Separation usually is good at the bottom continues somewhat linear as long as the luminance is high enough. By linear, I mean that the R, G or B is parallel to the Color Separation’s 45 degree black reference line. At 247 Nits (lol) the linearity range is 100%. At 100 Nits (~30 ftL), the range is linear until 40% where it acts very strange. Here is all the data I took. The table cannot be formated so I used : as a delimiter.


Luminance (Nits): Linear Range (0 – 100%): Comments:
247: 100%: Brightest display settings with good dynamic range:
206 (85% of 247): 50%:
178 (70% or 247): 45%:
100 (40% of 247): 40%:
44 (18% of 247): 20%:


This relationship exists if Sony's internal calibration engine is not turned on. For my display, this happens in Standard Pic mode (as opposed to Custom mode).

What do you think?

Kind Regards,

JJ

Author Steve

Mod - Scene 
#36 | Posted: 7 Oct 2012 17:50 
Hi JJ,

If the RGB Separation deviates from the 45deg target in any way, even if RGB are parallel/identical, the display is not suitable for Quick Profiling.

This is where full cube profiling works, and there really is no alternative.
(It is also why the inferior cube profiling of alternative systems with a lot less points will not work accurately).

Steve
Steve Shaw
Mob Boss at Light Illusion

Author buzzard767
User - Mono 
#37 | Posted: 8 Oct 2012 15:27 
Steve:
DIP mode is for when the patches are being generated via a separate timeline or pre-built clip.
When using extended desktop mode there is no reason to use DIP - use Closed Loop mode.

Closed Loop is the fastest mode, and if it can be used, it should be

Idiot's Guide:
The main likely option is 'Integration Time', which if available should be set to a longer time for accuracy, and shorter for speed. Short times can be used for quick checking of a display prior to final profiling with longer integration times. For example, with the i1 Display Pro use 0.25 secs for quick checking (with the understanding that accuracy is limited), while using 3.0 seconds for final accurate profiling.

A D3 doing a full 17 point profile at 3 seconds per read is about 4 hours. What about Probe Integration Time for final profiling with a K10-A?

Author Steve

Mod - Scene 
#38 | Posted: 8 Oct 2012 15:34 
Hi Buzz,

I have a K10-A here, and in Closed Loop mode it is a lot faster - easily half the time, although I have not done a measurement.
The display in use will also have an effect on time taken as the darker the display the longer it will take.

With the K10-A there is no user controlled Integration Time - the probe takes care of that internally.

Steve
Steve Shaw
Mob Boss at Light Illusion

Author buzzard767
User - Mono 
#39 | Posted: 8 Oct 2012 16:11 
Profiling a D3 from an i1Pro - is the following procedure correct?

1. Connect i1Pro
2, Select Measure All and enter a name in the Active Probe pop up box
3. Measure ALL
4. Disconnect i1Pro
5. Connect D3
6. Select i1Pro in the Reference Probe drop down box
7. Select Measure all and enter a name for the D3 in the Active Probe pop up box
8. Measure all

Anything else?

Author Steve

Mod - Scene 
#40 | Posted: 8 Oct 2012 16:17 
Hi Buzz,

Don't select a probe in the Reference Probe drop down box until after you have profiled all probes.
(It should make no difference, but just to be safe...)

Then, before doing real display profiling select the D3 as 'Active' probe, and the i1Pro as the 'Reference'.

Should all work fine then

Steve
Steve Shaw
Mob Boss at Light Illusion

Author buzzard767
User - Mono 
#41 | Posted: 9 Oct 2012 15:31 
Yes, I did do it that way - just didn't explain it correctly. Thanks.

Buzz

Author JJ Johnson
User - Mono 
#42 | Posted: 10 Oct 2012 16:36 
buzzard767:
I must have a workflow error. Using the ee ColorBox, LUTs produced from CalMan are okay but those from LightSpace are terrible. Grayscale, no, Greenscale is so bad that there is no need looking further.At first I thought it might be a meter problem so I switched from a D3 to an i1Pro but got the same results. LightSpace GS patches look normal with none of the coloration found in the finished product. To eliminate the possibility of a computer icc glitch using the LightSpace patches I did a run each of the Calman and LightSpace LUTs using internal patterns generated within ChromaPure software with no problems. Here's the basics of what I did so please correct me:1. Display Characterization - run 5p cube so as to approximate the Calman cube size2. Color Space Conversion - pick a name, Color Space Rec.709, Destination - the 5p cube from #1, Peak Luma3. Export the resulting LUT to the ee LUT holding folder in Windows4. Change the name to match what the ee Box is expecting and load the ee Box

Buzz:

Are you using Color Box or eeColor. I've used Color Box extensively for 6 weeks or so, but returned it when I could not determine how good the LUT cal quality was. Buying eeColor is much less expensive. I'm trying to determine LUT Cal Quality with LS and it's taking me a lot of time because the tool is very feature rich.

How do you determine where to upload the LUT information from with eecolor? I never got that far.

Kind Regards,

JJ

Author buzzard767
User - Mono 
#43 | Posted: 10 Oct 2012 20:26 
JJ Johnson:
Are you using Color Box or eeColor. I've used Color Box extensively for 6 weeks or so, but returned it when I could not determine how good the LUT cal quality was. Buying eeColor is much less expensive. I'm trying to determine LUT Cal Quality with LS and it's taking me a lot of time because the tool is very feature rich.

How do you determine where to upload the LUT information from with eecolor? I never got that far.

Hi JJ. I bought the first ee Color Processor from SpectraCal at CEDIA 2011. I didn't care for what it did which was basically adding a lot of luminance to everything except flesh tones. That Vivid mode stuff might appeal to some folks but it wasn't for me. I hung onto the box on the promise that 3D LUTs were going to become available for it "someday". When Spectracal Studio was announced they gave me a 30 day pass to try it out and it didn't work correctly at all. Eventually they loaded a better version of Studio on top of Calman 5 beta. Spectracal said I had to have the Spectracal ColorBox so I sent it to Seattle and it came back with that unusual purple face. Other than that it is my belief that the only things done to the box were setting the 6 LUT slots to Unity and adding their own OSD graphics. They may have done something else so that the LUTs would program on the fly but I don't know the answer to that one. I worked for two months and determined that the 125 point LUTs were about the same as those produced by Calman and ChromaPure for the Radiance video processors but once I started programming larger LUTs the processing went bonkers. See this: http://www.youtube.com/watch?v=YB1w9sGkSe8 LUT1 was a 5x5x5 and LUT2 was (I think) a 9 per side, 3 or 7 deep. Pretty, eh? Anyway, two months of measurements wasted. Spectracal support hardly pays any attention because I'm pretty sure they can't get it to work either or they should have said something by now.

I've been loading LightSpace LUTs with the ee software with limited success but that has been with two displays neither of which can reach the Rec.709 gamut points. I'm moving to Florida next week where I have three displays to work with, and I've also just yesterday learned a little about using the Axis and Smoothing tools in LightSpace so I expect some degree of success soon. Those tool, I think, are very important if the display isn't doing it's part.

Is your uploading question about getting LightSpace LUTs into the eeBox? If so, download the newest ee software if version 2 is on the web site. If it's still v1 it won't work and I can email you a copy of v2. When you get v2 I'll give you the specific steps for loading a LUT onto the Box.

Buzz

Author JJ Johnson
User - Mono 
#44 | Posted: 11 Oct 2012 08:48 
buzzard767
Hi Buzz,

Can you upload a Light Space LUT into an eecolor LUT processor purchased from eecolor or do you need the Color Box w. unity cube presets for uploading?

I'm trying to select my next external LUT processor, now that I've returned the Color Box.

The HDlink Pro price point is more attractive ... but I don't understand the level of hardware integration with Light Space. I’m looking for a LUT processor under $300US to my customers.

Kind Regards,

JJ

Author Steve

Mod - Scene 
#45 | Posted: 11 Oct 2012 08:55 
Hi JJ, with the V2 control software you can upload directly to the eeColor box - you do not need the SpectraCal version (actually, I would NOT recommend that version as we don't know what has been played with internally).

You can manually reset all internal LUT/Settings to unity your self.

Please read the thread on the HDlink Pro: http://www.lightillusion.com/forums/index.php?action=vthread&forum=13&topic=34

Drop me a mail for V2 control software and unity LUT to load.
Steve Shaw
Mob Boss at Light Illusion

Page  Page 3 of 8:  « Previous  1  2  3  4  5  6  7  8  Next » 
Display Calibration Light Illusion Forums / Display Calibration /
 3D LUTs for the home enthusiast

Your Reply Click this icon to move up to the quoted message

 

 (?)
Only registered users are allowed to post here. Please, enter your username/password details upon posting a message, or register first.

 

 
 
Online now: Guests - 4
Members - 0
Max. ever online: 264 [6 Jun 2020 12:02]
Guests - 264 / Members - 0