Color HCFR Calibration Discussion (Post your calibration files here) - Page 122 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #3631 of 3872 Old 01-20-2011, 05:35 PM
AVS Special Member
 
spongebob's Avatar
 
Join Date: Dec 2001
Location: San Jose, Ca.
Posts: 2,849
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 5 Post(s)
Liked: 11
Quote:
Originally Posted by dlarsen View Post

Profiling is when you use a reference meter (master) to 'train' a 2nd meter (slave). Using matrix offset coefficients or 'fudge factors' applied to the raw data from the 2nd meter, the goal is to make the 2nd meter produce the same results as the reference meter does when presented with the same stimulus. Ideally simultaneously.

If this is done well, it can make the 2nd meter measure just like the reference meter for that particular display or SPD.

Dave

Thanks, Dave

I have the DTP-94 and the gray scale accuracy matched an eyeone calibration but the individual colors on the CIE are way off. Could I use offsets to match it?

bob
spongebob is offline  
Sponsored Links
Advertisement
 
post #3632 of 3872 Old 01-20-2011, 05:43 PM
Advanced Member
 
thomasl's Avatar
 
Join Date: Apr 2004
Posts: 592
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by WTS View Post

I have the latest version 2.1 and in the ini file I don't even see minrequest or maxreadloops listed. Last night I selected the eyeone probe the checked the box you mentioned and I could see the ini file get updated as soon as I clicked okay but when I open it I don't see those. Of course I don't have an eyeone connected but that shouldn't make any difference to the ini file. Can you attach a copy of your ini file.

Walter, I seem to recall now that starting with 2.1, they no longer showed up in the file but if you put them in there in the EyeOne section of the ini file, they still are being used by the program.

This is what I have right at the top of my ini file:

[Options]
Language=ENGLISH
[EyeOne]
DeviceCalibrationMode=1
MinRequestedY=80
MaxReadLoops=8

I inserted those bottom two lines - which used to be there by default in HCFR 2.0 but with MinRequestedY set to 10 instead of 80. Their affect on the dark readings averaging behavior is how I described in that post from a few years ago.

hope this helps,


--tom
thomasl is offline  
post #3633 of 3872 Old 01-20-2011, 05:53 PM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hi Tom,

What does your ini file look like when you select the HCFR probe.

Here's my ini file when I select Eyeone;

[Options]
Language=ENGLISH
[Defaults]
Sensor=Eye One
Training FileName=
Training Mode=1
Generator=DVD manual

and HCFR probe;

[Options]
Language=ENGLISH
[Defaults]
Sensor=HCFR Sensor
Training FileName=C:\\PROGRA~1\\COLORH~1\\Etalon_HCFR\\Zoyd-px600u.thc
Training Mode=0
Generator=DVD manual

I suppose I could insert those 2 lines after the HCFR sensor. I wonder if the program would even still be looking at those 2 lines or act on those lines. If they changed the code then those commands may not do anything in this new rev.

Thanks

Walter
WTS is offline  
post #3634 of 3872 Old 01-20-2011, 06:07 PM
Advanced Member
 
thomasl's Avatar
 
Join Date: Apr 2004
Posts: 592
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by WTS View Post

I suppose I could insert those 2 lines after the HCFR sensor. I wonder if the program would even still be looking at those 2 lines or act on those lines. If they changed the code then those commands may not do anything in this new rev.

It's been awhile since I experimented with this but I'm almost certain those settings still are active for 2.1 even if they don't show up in the default ini file when you first run it. I would have to give it another try to double check but I don't have an HCFR probe to test to see if those settings do anything when taking measures with it.

hope this helps,


--tom
thomasl is offline  
post #3635 of 3872 Old 01-20-2011, 06:12 PM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Thanks Tom, next time I do some playing with it I will insert those 2 lines and play with the numbers for the max/min.

Walter
WTS is offline  
post #3636 of 3872 Old 01-20-2011, 06:20 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by spongebob View Post

Thanks, Dave

I have the DTP-94 and the gray scale accuracy matched an eyeone calibration but the individual colors on the CIE are way off. Could I use offsets to match it?

bob

My guess is yes, quite likely you can match it or at least get very close.

Here's a tutorial on simultaneous probe training for HCFR. This is for the old rev but should give you the gist. See section 3.

http://members.aceweb.com/dlarsen/ge...tion_Files.pdf

Dave
dlarsen is offline  
post #3637 of 3872 Old 01-21-2011, 08:28 AM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Dave, In the pdf you attached where you set up the HCFR parameters, you made reference to the white measure and wrote use clear photo diodes. What does this "white measure" check box do exactly?

Thanks

Walter
WTS is offline  
post #3638 of 3872 Old 01-21-2011, 09:24 AM
AVS Special Member
 
nashou66's Avatar
 
Join Date: Oct 2006
Location: West Seneca, NY
Posts: 4,261
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 52
Quote:
Originally Posted by WTS View Post

Dave, In the pdf you attached where you set up the HCFR parameters, you made reference to the white measure and wrote use clear photo diodes. What does this "white measure" check box do exactly?

Thanks


Walt I am not sure exactly what happens here. But the new taos Chips I use in the HCFR probes I have made and the parts I have sent you have three arrays of diodes on the chip, RGB and Clear or White. I think selecting this option will only use the clear sensor to read the white field instead of using RGB together and calculating white from that. I have not tried this method yet. I have two more HCFR PCB's and will make a unit and test both methods using my i1 Pro.

For the file I sent you of my HD 145 colored lens equipped Marquee 8500 i did the simultaneous profile method.

I think this is the best method, but if you can get a better probe and use its numbers and manually enter them and then do a comparison run on the same display then you can see if there is much of a difference.

I wish HCFR would add more Pro probes like the PR 650,655 etc to the list
of supported probes. Id love to make a batch of Probes calibrated to those probes and sell them.

Once you build the probe Id like to hear the difference of the new chips to the older ones. I dont know anyone who has dont this yet. I was not able to locate the older chips so i only have the new ones to compare.

Athanasios
nashou66 is offline  
post #3639 of 3872 Old 01-21-2011, 11:11 AM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hi Athanasios,

That sounds logical about selecting the clear only sensor diodes with that check box.

Well I'm hoping I can talk a certain person into doing a profile of these 2 probes against his references. Although I'm not sure what he has for reference probes anymore.

The big difference between the new taos and old taos sensors is in sensitivity, the new ones are 80% as sensitive as the old ones they did this in order to measure brighter sources. Other than that I don't think there is any differences.

Walter
WTS is offline  
post #3640 of 3872 Old 01-21-2011, 12:05 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Walter,

As you and Athanasios both guess, I too believe the white' checkbox applies to the clear, unfiltered photo-diodes in the array but this is just a guess. The HCFR developers would have to clarify further.

From what I can see, the biggest difference between the old/new TOAS parts is that the new parts encode the info into a 16 bit word and pipe it via I2C to the micro. The old part provided a frequency or counts and wasn't tied to a specific bit-depth. For dynamic range, the old part had 1X,10X, and 100X gain settings and the new part gets dynamic range by having a 16bit word. As the new part must encode into a 16bit word, and have I2C overhead, it seems it must be slower in comparison but I imagine this is no real limitation for display calibration purposes.

From what I've read, both parts have similar results for display calibration purposes, it's just that the old parts are no longer available. You should note however that the cal files for the two parts aren't compatible or interchangeable. I believe the HCFR group make a distinction in the community cal files for old' and new'.

Dave
dlarsen is offline  
post #3641 of 3872 Old 01-21-2011, 03:39 PM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hi Dave,

I've attached a paper from Taos and it talks about the differences between the two. Basically the new ones are drop in replacements, no difference each for sensitivity.

The old files will work for the new one but the actual levels might be off because of the sensitivity differences but other than that there is should be no difference. I used Athanasios's file and it actually looks great but that was before I read this paper, hmmm.

 

TCS3200vsTCS2301.pdf 53.3701171875k . file

Walter
WTS is offline  
post #3642 of 3872 Old 01-21-2011, 03:52 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by WTS View Post

Hi Dave,

I've attached a paper from Taos and it talks about the differences between the two. Basically the new ones are drop in replacements, no difference each for sensitivity.

The old files will work for the new one but the actual levels might be off because of the sensitivity differences but other than that there is should be no difference. I used Athanasios's file and it actually looks great but that was before I read this paper, hmmm.

Walter,

Thanks for the info. I must have confused it with other new parts TAOS has recently introduced. They are incorporating I2C as the interface more and more and I suspect the fact that the I2C patents owned and enforced by Phillips expired in 2009 so it is now considered public domain with no licensing requirements or fees.

I actually have a meeting scheduled next week with the local TAOS rep. On a totally different subject from display calibration and colorimeters however.

Also, the last I looked, the HCFR team did make a distinction for old/new cal files. I can't understand why they would do that if they really are 100% compatible with eachother. I also see where you were using a file apparently made by Zoyd. If you have a plasma, I'd guess that would be a pretty good choice as he had it pretty much dialed to his eye1. (Although I do think the TAOS outperformed the eye1 at low light levels)

Dave
dlarsen is offline  
post #3643 of 3872 Old 01-21-2011, 04:08 PM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hi Dave,

Hey when you talk to the rep ask him about the filters in the chip and whether they degrade over time and why. I sent them an email but never received an answer - yet.

No I haven't tried Zoyd's file yet but I plan on testing it. Yeah that was a great thread to read, I'll have to read it a few more times. Too bad it hasn't continued on.

Walter
WTS is offline  
post #3644 of 3872 Old 01-21-2011, 04:27 PM
AVS Special Member
 
nashou66's Avatar
 
Join Date: Oct 2006
Location: West Seneca, NY
Posts: 4,261
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 52
Dave is right about the ability for the HCFR probe to outperform the i1 Pro down low.

What I found is not so much accuracy but accuracy in repeatability. The HCFr probe reads
low IRE's more consistently than the i1 Pro does. I have a PR 650 I am hoping PR will send me the file for. Then I'll be able to tell how close the low IRE's are on the Taos Chipped HCFR.

But that all depends on if this PR 650 is close in calibration, So I have to first test it against my known good i1 Pro.

Athanasios
nashou66 is offline  
post #3645 of 3872 Old 01-21-2011, 04:33 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Walter,

Will do, although I don't expect much technical information from a local rep. FWIW, (ancedotal) I can't say that I've seen any cause for aging concern with mine over the years. I believe the filters are well integrated into the package and close to hermetically sealed from the external environment.

I don't know what happened to Zoyd. I haven't seen him for a long time now but he knew his stuff and I had alot of respect for him. He sometimes locked horns with some of the 'pros' here with regards to a tri-stim just not being capable enough for display calibration or CMS. Folks like him (and Tom H) weren't afraid to go against the pro-flow and to argue otherwise. He designed hi-end spectros and knew color-science very well.


Dave
dlarsen is offline  
post #3646 of 3872 Old 01-22-2011, 09:25 AM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Athanasios,

I aways felt the HCFR in the low range was repeatable and just looking at the results I got down low it didn't look like it was out(no one colour push).

Dave,

Well I'm not a filter guy that's for sure but I tend to agree with you as it sounds logical. At least maybe this guy can get you an answer as I still haven't received an answer for Taos. In the electronics industry as I've always found that if they don't seem to want to answer or you can't find the answer in their specs/data sheets then chances are the device is not good in that respect.

Yes it did sound like Zoyd knew what he was talking about that's from my opinion, he surely had the background to be knowledgable about the subject at hand. It was very interesting reading from both sides.

Although I'm still some what confused when they say that if the probe is calibrated on a plasma or a certain plasma or any other light source reference then it's not good for any other plasma let alone other technology(DLP, CRT etc.). It would seem to me that if it has the proper coordinates(red/green/blue/D65) for REC709 calibrated into it for the CIE chart then after that it's only sensing colour and lum regardless of what the light source is. So why when it's point at another type/brand of light source or plasma why would it's inaccuracies increase? Isn't it like saying one pound of sand weights the same as one pound of feathers.

Walter
WTS is offline  
post #3647 of 3872 Old 01-23-2011, 03:44 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by WTS View Post

Although I'm still some what confused when they say that if the probe is calibrated on a plasma or a certain plasma or any other light source reference then it's not good for any other plasma let alone other technology(DLP, CRT etc.). It would seem to me that if it has the proper coordinates(red/green/blue/D65) for REC709 calibrated into it for the CIE chart then after that it's only sensing colour and lum regardless of what the light source is. So why when it's point at another type/brand of light source or plasma why would it's inaccuracies increase? Isn't it like saying one pound of sand weights the same as one pound of feathers.

Walter,

I believe it’s because the SPD of different illumination sources and different display technologies is, well, different. The filters in most tri-stimulus instrumentation are approximations for the weighted spectral response of the human visual system. The response of the HVS and the tr-stim filter approximations are based off of relatively broad-band sources. For sources that have a very different SPD makeup, these approximate fits and correlations can break down.

For instance, take your concerns about IR. You presumably are concerned about this because you believe you have a source that emits spectral power in this region, and you believe that other sources don’t. The HVS doesn’t respond here, and your TOAS sensor may. A little. You are concerned about this potential response mismatch between a SPD with an IR ‘spike’ introducing offsets or errors. It’s a similar issue with different technologies with different or less broad-band SPD.

It’s been shown by practical empirical methods and testing however, that the different SPD induced offsets produced by a tri-stimulus filter fit error or mismatch can be largely identified and mitigated in the transforms. Fudge-factored out by profiling to a technology specific SPD and reference measurement instrumentation. For practical empirical display measurement purposes anyway.

A 10nm spectro could have similar issues with narrow-band SPD too.

Dave
dlarsen is offline  
post #3648 of 3872 Old 01-24-2011, 06:20 AM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hi Dave,

Thanks for your insite into this. No I don't have concerns wirh IR I was only asking about it.

I tried the Zoyd file and it seemed like it was out at least to my eyes. Actually it and some of the others (random tries from other files including the other plasma files and CRT files) all placed the red as being too low on the white balance.

Where as the file I got from Athanasios showed the red to be just a hair low and according to my eyes the red was actually still a touch too much. I beleive Athanasios did his file from his just calibrated I1 pro. One problem with using the other files or any of these files is you have no idea what probe was used as a reference or how inaccurate it was to start with.

Comparing some of these to Athanasios's file gave readings that were way out. Even Zoyd's file, I chose a couple random greyscale points 40/70 and adjusted according to the Zoyd file and I could clearly see the white balance was way out. Which seems odd because he did his on a plasma with a I1 pro too which I assume was calibrated and still within calibration limits and his probe sensors came from the same batch of sensors as mine and Athanasios's were from the new series of sensors yet his file showed the best white balance at least to my eyes.

Walter
WTS is offline  
post #3649 of 3872 Old 01-24-2011, 08:23 AM
AVS Special Member
 
nashou66's Avatar
 
Join Date: Oct 2006
Location: West Seneca, NY
Posts: 4,261
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 52
The screen also plays a factor as some filter out some wavelengths I would think, no?

I did mine a unity gain Screen that looks perfectly white with no coloration.

The file was made with the i1 Pro facing the screen.

I did about 5 runs and saved the Values from the matrix and listed it. i then looked to see how far off they were and used the one in the middle range. I figured this would be the best way to make sure i did not have an erroneous readings on my first try.

Athanasios
nashou66 is offline  
post #3650 of 3872 Old 01-24-2011, 04:28 PM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Do either of you know what the wavelengths are for the RGB REC709 points. I'm guessing that blue is around 470nm, green is around 532nm and red is around 625nm, or are there a precise wavelengths for the 709 coordinates.

Walter
WTS is offline  
post #3651 of 3872 Old 01-24-2011, 05:39 PM
AVS Special Member
 
dlarsen's Avatar
 
Join Date: Mar 2001
Location: Beaverton, OR, USA
Posts: 1,908
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Walter,

AFAIK, the rec.709 chromaticities are based on the CIE1931 standard and this is represented by x,y points, not wavelength. There may be a way to get there, and the attached link does show the CIE lotus with wavelengths shown.

Dave

http://en.wikipedia.org/wiki/CIE_1931_color_space
dlarsen is offline  
post #3652 of 3872 Old 01-24-2011, 06:00 PM
WTS
AVS Special Member
 
WTS's Avatar
 
Join Date: Nov 2002
Location: Calgary
Posts: 1,769
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Thanks, From what I just read the colour primaries were originally based on the 1931 standard CIE but were changed in 1990. As far as I can understand this article the RGB colours are now B-470nm, G-550nm, R-600nm.

It's from Digital Video and HDTV: algorithms and interfaces by Charles Poynton. I tryied to paste a link but it wouldn't take but if you google that it's on page 243.

Walter
WTS is offline  
post #3653 of 3872 Old 01-26-2011, 09:23 PM
Member
 
CpuZapper333's Avatar
 
Join Date: Jan 2011
Posts: 43
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Hello all hope someone can provide some guidance, I have been attempting to calibrate my LG 50pk550. I have set the contrast and brightness using the % method and checked with plunge patterns, both are very close.
I have done a greyscale calibration and checked my primary and secondary colors. The cyan and magenta are off, magenta is the worst.
How do I correct this with the CMS in the LG. I have not been having much luck. My test files.chc for HCFR has been attached. By the way I was using the Spyder 3 with the HCFR software.

What do you think?

 

Test-Results.zip 1.6875k . file
CpuZapper333 is offline  
post #3654 of 3872 Old 01-27-2011, 07:43 AM
Member
 
hauptm's Avatar
 
Join Date: Mar 2007
Location: Winston Salem, NC
Posts: 106
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by CpuZapper333 View Post

Hello all hope someone can provide some guidance, I have been attempting to calibrate my LG 50pk550. I have set the contrast and brightness using the % method and checked with plunge patterns, both are very close.
I have done a greyscale calibration and checked my primary and secondary colors. The cyan and magenta are off, magenta is the worst.
How do I correct this with the CMS in the LG. I have not been having much luck. My test files.chc for HCFR has been attached. By the way I was using the Spyder 3 with the HCFR software.

What do you think?

I was having the same problem on a different TV but also using the Spyder 3.

I returned it and got the i1 LT. Works much better and am very happy with the results.
hauptm is offline  
post #3655 of 3872 Old 01-27-2011, 12:32 PM
Member
 
CpuZapper333's Avatar
 
Join Date: Jan 2011
Posts: 43
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I was able to pick up the spyder3 locally so I thought I would try it out. I do have access to a i1Pro (About 3 years old not re-certified) I have done a side by side greyscale comparison. In my geyscale test they were close. Both products seemed to have issues at the lower IRE levels, when using the IRE 15% or lower greyscale both would give different results on each test. The rest of the scale was X or Y readings were within .002, some were the same. I think I ran a quick primary and secondary on both and don't remember anything that stood out.

That said both testers could be out.

If the issue is the TV how can I compensate. I think I read something about picking a different luminance point and try recalibrating?
CpuZapper333 is offline  
post #3656 of 3872 Old 01-27-2011, 12:59 PM
AVS Special Member
 
alluringreality's Avatar
 
Join Date: Jul 2006
Posts: 3,120
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 33
Quote:
Originally Posted by CpuZapper333 View Post

The cyan and magenta are off, magenta is the worst.
How do I correct this with the CMS in the LG.

I am not familiar with the TV or brand. The measurements that can be found through http://www.testfreaks.com/televisions/lg-50pk550/ appears to also have the secondaries still off on the after measurements. Some TVs just have items that cannot be completely corrected with the controls the TV provides, and since I haven't looked at LG controls I cannot say if you could get any closer. Personally I would probably go with the i1 Pro measures, except for the low grayscale.
alluringreality is offline  
post #3657 of 3872 Old 01-27-2011, 01:28 PM
Senior Member
 
NxNW's Avatar
 
Join Date: Dec 2009
Posts: 225
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 13
Agreed - out of spec or not, I would tend to trust the i1pro a little more. The presumed improvement in accuracy would probably be more applicable to measuring gamut, not grayscale, but either way, a meter is better than no meter, and even an old i1 should give you reasonable results, even a some ways below 15% stim. [There has been a discussion here recently (about 20, 25 posts ago) on how to "profile" one meter using another - it might be instructive to try profiling your spyder against the i1.]

Assuming the spyder is giving at all reliable info, I wonder how you concluded cyan is off? Looking at the "delta E" row in the "Primaries and Secondaries" view, it looks like magenta indeed is the worst offender, but cyan doesn't strike me as obviously incorrect. I'd be more worried about green and blue.

Ignore the locations on the CIE chart for a moment- in the same Primaries and Secondaries datagrid view, the "delta luma" row (visible if you click the down triangle next to "Editable data") shows me green and blue are too bright. After the hue of magenta, I think that's where whatever error you have left is mostly coming from, and basically you need the CMS to make the necessary changes.

The hue of all your primaries seems reasonable. Your grayscale seems reasonable. Your white point seems reasonable. If you *still* want to get those dE's as low as possible, your priorities should probably be: making magenta less blue and/by reducing the brightness of the green and blue primaries. To do this you'll need to either experiment with the basic controls offered by the display (which can be an arsenal of very blunt weapons- you may introduce as much or more error than you solve) or start in with the CMS. No idea if that CMS is up to that task

That's the long way of saying what alluringreality said- you may already have calibrated this TV as well as can reasonably be expected with the tools and time available. If the picture doesn't look obviously wrong and it's better than when you started, then you may want to pause on this plateau awhile and enjoy it.
NxNW is offline  
post #3658 of 3872 Old 01-27-2011, 10:28 PM
Member
 
CpuZapper333's Avatar
 
Join Date: Jan 2011
Posts: 43
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Thanks for the info guys.

I have played with the CMS again tonight and attached the results.
I'm not sure why the delta E and Delta Luma drop when I move away from the ideal x and y for cyan and magenta. Magenta is still the highest delta E and delta luma, blue is not far behind.

What do you think of this calibration?

I'm wathcing the TV now and wondering how close it might be.

 

New-Color-test.zip 4.9951171875k . file
CpuZapper333 is offline  
post #3659 of 3872 Old 01-28-2011, 12:28 PM
Senior Member
 
NxNW's Avatar
 
Join Date: Dec 2009
Posts: 225
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 13
That looks a lot better at least on paper.

So you're saying the dE *increases* when you try to move the hue of magenta closer to (0.321,0.154)?

Your display was discussed here a few months ago- review this post:

http://www.avsforum.com/avs-vb/showt...1#post19216281

It contains a link to another helpful post:

http://www.avsforum.com/avs-vb/showt...0#post18868120

Basically, knd's conclusion was use 75% patterns when calibrating an LG CMS and you may get better results.

Not sure if this would help your magenta secondary or not.
NxNW is offline  
post #3660 of 3872 Old 01-29-2011, 09:12 AM
AVS Special Member
 
alluringreality's Avatar
 
Join Date: Jul 2006
Posts: 3,120
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 6 Post(s)
Liked: 33
Generally I don't think it's a good idea to try to minimize deltaE on color in ColorHCFR. The deltaE on color in ColorHCFR can do some odd things depending on how xyY is off, and the '94 calculation (not currently available in ColorHCFR) does not result in similar minimum deltaE values for the same xyY deviations. The primaries from CpuZapper333 seem to be close to Rec 709, so I'd consider that a reasonable target. If the primaries are very far off from 709 or SMPTE-C (601 in ColorHCFR) I tend to go with the color corrector tab from http://www.avsforum.com/avs-vb/showt...7#post15607437
alluringreality is offline  
Reply Display Calibration

User Tag List

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