Color HCFR Calibration Discussion (Post your calibration files here) - Page 125

hi all,

I use Eye One iDisplay2.
I tried iMatch that comes with iDisplay2, but it isn't as powerful as HCFR.
So I tweaked TV (XBR 909) controls to for HCFR to get better PQ.
I used default full RGB profile for the display in Win7.
I know that LCD TV isn't perfect, so thought about compensate LCD anomalies with display profile
So I run iMatch to generate appropriate profile on top of already calibrated by HCFR LCD.
When I use iMatch the Color Temp wasn't expected 6500k, but lower in 6400k range and it asked me to change RGB controls of LCD - so I recorded current RGB (White balance gain and bias settings) and change it to satisfy iMatch and as result I got new profile that was applied to Win7 Display class.
After I run HCFR to validate what is resulted in worse charts than I got without iMatch and Color Temp want off 6500k to 6600k as result.

I did the same for Cinema and Photo TV modes that i can tweak independently on this TV and results of Color Temp were different as well.

Why there is difference? to What S/W readings / calculations to trust more?
Yes I calibrated sensor each time, but by different S/W.
Is there difference in calibration or calculations between those S/W?

thanks for you thoughts.

I've calibrated 3 plasmas, 1 LCD, and a projector using DVD manual and the AVS test pattern disk.

So I thought I'd try to use view images just to take the grayscale readings a little quicker, but I can't get it (view images) to work.

What happens is that I'll hit the go button and the PC and display screen will imediately go to black - I'm guessing it is starting at 0 IRE. But then nothing else happens. It doesn't advance to 10 IRE, 20 IRE. . . or anything. It just stays on black. I've left it on for 30 minutes and still no movement.

Would appreciate any suggestions as to how to fix. I'd really like to try this feature and see how much variance there is between the AVS disk and the HCFR generated images.

Probably your screen turned off due to no input at 0 IRE and HCFR is stuck trying to read the 0% patch. Just pull the probe off the screen and let a little ambient light onto it to get it to advance to the 10% window, then ignore the 0% reading.

Thanks for the tip rpauls.

What I finally ended up doing was to hit the left mouse key after about 2 minutes at 0 IRE. Then a message came up on the screen that said the application (ColorHCFR) wasn't responding,. But just after that messsage, the 10 IRE screen came up, and from there it cycled (very quickly) through the remaining 9 steps to 100 IRE.

Using view images is a big time saver as it cycles faster than using a DVD. It'll be interesting to compare gray scale readings to those using the AVS disk.

I'm using a 37% screen so it's pretty much like the fields on the AVS disk.

I wonder if Windows 7 has anything to do with this quirk.

I got into the same issues
I enabled "debug" for sensor and somehow it works now - just need to calibrate sensor for each iteration

im still learning here but after i second try i can tell you I will never go without a calibrated picture again , the picture is simply amazing the brown tinge is gone , and my plasma is hitting some amazing blacks . . . last suggestions helped alot guys , im still falling out of my IRE at 90 , and my gamma just a bit low from 10 to 60 IRE . . . tell me what you guy think

--Picture menu
Mode: Movie
Cell Light: 15
Contrast: 85
Brightness: 54
Sharpness: 10
Color: 50
Tint: G50/R50

Eco Solution submenu:
Energy Saving: Off
Eco Sensor: Off
No Signal Power Off: Off

Auto Adjustment and Screen submenus: [Grayed out]

Advanced settings submenu
Black tone: Off
Dynamic contrast: Off
Gamma: + 1
Expert Pattern: Off
RGB Only Mode: Off
Color space: Custom
White Balance [see below]
10p White Balance: On
Flesh tone: 0
Edge enhancement: Off
xvYCC: Off

Color Space Auto

White Balance submenu
R-Offset: 24
G-Offset: 24
B-Offset: 25
R-Gain: 24
G-Gain: 26
B-Gain: 25

10p White Balance submenu
Interval 1: Red 0, Green 0, Blue 0
Interval 2: Red+1, Green 0, Blue 0
Interval 3: Red 0, Green 0, Blue 0
Interval 4: Red 0, Green 0, Blue 0
Interval 5: Red-2, Green 0, Blue-3
Interval 6: Red 0, Green 0, Blue 0
Interval 7: Red-2,Green 0, Blue-6
Interval 8: Red-3, Green 0, Blue+1
Interval 9: Red 0, Green 0, Blue 0
Interval 10:Red 0, Green 0, Blue 0

Picture options submenu
Color tone: Warm2
Size: Screen Fit
Digital Noise Filter: low
MPEG Noise Filter: low
HDMI black level: low
Film mode: Off
Motion Judder Canceller: Off
Auto Protection Time: 10min

 

Color movie tweaking 2.zip 0.826171875k . file

I bought an i1 LT recently, looks to be reasonably new stock since it mentions Windows 7 on the box. Combined with ColorHCFR and the AVSHD709 disk I had my first attempt at a basic initial calibration last night on an Australian Panasonic P50V20A. As I understand it, this TV lies somewhere between the US G25 and VT25 - it's our top of the range 2D model with identical specs and features to our top of the range 3D set (VT25) except for the 3D functionality. It has 5m:1 contrast ratio, 24p/96Hz playback, THX (True Cinema) mode, etc.

In particular this TV features Professional 1 and 2 calibration modes which are based on the THX preset and include individual RGB Gain and RGB Cutoff controls in the White Balance menus, along with a single Gamma preset, and Hue and Saturation Color Management controls - no need to enter the service menu. The primary picture controls include Contrast, Color and Sharpness (0-100% sliders), and Brightness which ranges from -50 to +50, with 0 as the midpoint of the slider.

I've noticed something odd while calibrating the TV - my gamma seems to act strangely, even at default settings. I've disabled Eco Mode and all forms of additional image processing "enhancements", and using the defaults of 50% Contrast, 0 (50%) Brightness, and checking 2.2, 2.4 and 2.6 Gamma presets, my results all show the same shaped curve. For example, here is the 2.4 Gamma curve at default settings on Professional 2 (THX-based) preset:



I've attached the .chc file for 2.2, 2.4 and 2.6 Gamma at default settings (P50V20A_DefaultGammas) and they all show a similar curve. There's nothing to indicate while viewing this set that Gamma is behaving this oddly.

In any case, using Professional 2 (THX-based) mode, I then set the following and conducted my calibration:

Contrast: 62%
Brightness: -2 (i.e. 48%)
Color: 39%
Sharpness: 10%
Gamma is 2.6

I'm very pleased with the resulting image as I like a rich punchy picture and I watch movies in relatively low-light conditions, so I don't mind sacrificing some black detail. Despite the handicap of having a darker low-end with my first attempt I managed to get colors to the point where Delta E is relatively low for all but 10% IRE - to the eye the colors look pretty much perfect:



The full .chc file is attached for reference as P50V20A_InitialCalibration.

I'm also happy with a 2.3 average Gamma, as that is my preferred target (I find 2.2 a bit too washed out). The problem is that the Gamma ramp still shows the same odd shape which was found at default settings:



I did allow the TV and the sensor to warm up for at least 30 minutes, I did the calibration in a darkish room with regular calibrations of the sensor (though unfortunately probably more like every 30-45 minutes rather than every 10-15 minutes). Is there something inherent to the 2010 Panasonics that gives them this type of Gamma curve? Any setting I'm overlooking which could cause this?

I'll have another go at it tonight and try to do it in a pitch black room with 10 minute sensor calibrations, and I've also now ticked the "Average many reads on dark measurements" option in ColorHCFR to ensure a more accurate low-light reading. Any feedback would be useful.

 

P50V20A_InitialCalibration.zip 0.748046875k . file

 

P50V20A_DefaultGammas.zip 2.2119140625k . file

maybe contrast is set too high? My Panasonic TXP42G20 also shows this kind of gamma only less dramatic at 90 IRE than yours, see pic.

Interesting, thanks for that. I'm sure my curve is more exaggerated because I've deliberately tried to get a higher contrast effect (punch, pop, whatever you want to call it), which is why it flares out so much at the end and goes deeper at the start. Your curve seems to confirm though that overall it's a normal shape for the Panasonics.

What kind of test patterns are you using?

Got my i1 a couple months ago, have been slowly getting the hang of doing this myself. I have a couple observations and questions:

Is there a discrepancy between setting black level using DVE HD Basics vs. AVSHD709? Maybe I'm crazy?

On an LCD set, what difference does a 2.1 gamma equate to compared to 2.2? I mean, its LCD so blacks already look pretty crappy watching at night (at least on the 3 sets I have calibrated myself), and I've tried numerous tricks to try and get the gamma to go up but it seems like it may be pointless.

Also, accurate colors on two LCDs I have calibrated recently are crazy! I used the 21% of the Y @ 100 ire method and for both sets I actually had to turn the color control down significantly. It brought the red luminance in line on one set and lowered the delta e on the colors quite a lot; I can see how the image definitely looks more accurate/natural when watching cg animated movies or newer live action movies, but watching older movies on blu-ray (circa 1980s) or some hd cable content, the colors look washed out, especially at night. I have a feeling this is just a limitation with LCD displays, and I am dreaming of the day when i pull the plug on a plasma in the next couple months.

Hi! This is my first post in this forum, it will be a little long! Sorry for that!
Finally I could put my hands on an Eyeone Display 2. I wasted no time to start the calibration of my screens (and read tons of articles in the meanwhile).
So I calibrated my Panasonic 42G20 plasma using HCFR. The calibration was successful, I could reach a setting which seems to be good, compared to other users post. However, I use my PC as a source, so I thought, I could use the Lacie BluEye Pro software to make things even better using ICC profiles. (I used that software to calibrate my monitor). So I carried out the calibration process with Lacie Blueye pro. I set the target gamma to 2.2 here also. I discovered some interesting things after the calibration:

1) the initial measurement with Lacie showed a gamma of 2.5 altough in HCFR it is close to 2.2
2) after the calibration it confirmed that it could set Gamma to 2.2. As a result the picture became brighter, especially in the darker tones. So I started a measurement with HCFR
3) with the new settings HCFR measured a gamma of 2.0 or even higher, and the luminance and gamma graphs were off

I also tried eyeone's software the eyeone match 3, and the results were almost the same (off gamma values)

this is a huge difference in my opinion. lacie blueye pro is said to be one of the best monitor calibration software by professional computer-monitor reviewing sites, so it should be able to set correct values.

can someone explain the difference in gamma and luminance curve measured by the two software?

ps: the same difference in the curve exists on my computer screen. it is said to be calibrated to 2.2 but HCFR measures around 1.9

edit: color temperature is also measure 500K higher by HCFR in every measurement

Have you tested the difference in gamut measurement between the Lacie software and ColourHCFR also?

Can you please post some data (e.g. tables, graphs, or screendumps) showing the differences in gamma, grayscale and gamut for the two software solutions?

Better yet. Can you post the .chc files. You'll need to post them inside of a .zip file.

Lacie works with sRGB. I used REC 709 in HCFR. Also tried sRGB there, but the results are almost identical.
I didn't save the "wrong" results. I will measure it again and post it all.

A couple thoughts regarding gamma measurement. 1. Make sure you use the same test patterns for each test including having the same size for each. 2. Try to keep the light level in the room the same for each test.

I cannot effect how Lacie measures, it only has a start/stop button, gives some intrusctions during the process, and gives back an ICC profile in the end.
The room light level in the room is constant (totally dark)

here are the files:
Color_calibrated.chc: default ICC profile, settings done only on TV - this is what I use now
Color_calibrated_with_ICC_lacie_22.chc : TV settings as above, and calibrated with Lacie software

I started one more calibration, where I set the target gamma in Lacie to 2.4. it resulted in the worst picture from the all, but the gamma came back to 2.2! interesting!

Color_calibrated_with_ICC_lacie_24.chc : TV settings as above, calibrated with Lacie, target gamma 2.4

 

reports.zip 2.42578125k . file

Can you also post images or measurement data (gamma, grayscale and possibly gamut) from the Lacie software for the setting not involving a custom ICC profile?

E.g. use Lacie's software to measure gamma, grayscale and maybe gamut on the setup on which you have measured gamma=2.2 ("Color_calibrated.chc")) so it is ppossible to compare the measurement results.

here it is. gamut is almost identical to sRGB (in case you cannot see it on this bad quality jpg )

 

report2.zip 51.9482421875k . file

Interesting.

What test patterns and test generator/player are you using for calibrating with:

(1) ColorHCFR?
(2) Lacie software?

1) AVS HD 709 - Blu-ray & MP4 Calibration - AVCHD version
windowed greyscale patterns
2) it shows patterns for itself. it is a windowed application which has a circle in the middle (around the sensor)
there are some pictures in the following article:
http://www.tftcentral.co.uk/reviews/lacie_blue-eye.htm

Gamma measurements on plasma will be affected by the pattern or image displayed for measurements. For example if you measure the Windows patterns and compare it to one of the APL series from AVS HD 709 you will probably get different graphs, although the actual video levels match.

I agree with alluringreality: it will not make sense to compare measurements if the test patterns and test generator/player is not the same used for both software packages.

Can the Lacie software use the AVCHD patterns? If so you would probably use the APL-patterns (large or small) for your plasma TV.

it can't unfortunately. it is designed for photographic purposes, not home-cinema, I guess.
can you check the luminance graph in the .chc file for the ICC profile switched on? why is it incorrect?

ps: if I measure my LCD monitor with HCFR, it shows 2.0 gamma, although Lacie measures 2.2 which was set as calibration target.

Thanks to this forum, I have successfully calibrated my new TV. This is my first plasma TV and also the first time I had ever calibrated one. I come from photography world (not professional, just serious enthusiast), so I do have basic knowledge of color and display calibration. I do calibrate my computer monitor regularly but TV calibration is obviously not the same. TV calibration needs a lot of work going through the menu and a lot of adjustments, unlike monitor calibration, where the software automatically does most of the work for you. However, thanks to my basic understanding of color calibration, my transition into TV calibration was much easier after spending some time reading the calibration guides here. I think it would be helpful to somebody who is trying to do it yourself like I did by sharing my experience, especially for those who have the same or similar TVs as mine.

TV: LG 42PW350 720p 3D plasma
Sensor: Spyder3 colorimeter
Software: ColorHCFR
Pattern disc: AVS HD709

My TV is an entry level, one of the least, if not the least expensive 3D TVs out there. However, I bought this TV with plan for proper calibration in mind so this TV fits me (and my budget) well as it was said to have ISF calibration ready modes. As I expected, it has 2 advanced picture modes called "expert 1" and "expert 2" with 20-point gray scale adjustment and CMS (saturation and hue controls for primary and secondary colors, I guess you call it 2-D CMS instead of full featured 3-D CMS). These advanced features allow me to do a full calibration and fine-tune the picture quality.

I was able to use my Spyder3 (Express version) colorimeter which I bought to used with a third party calibration software package for my computer monitors. It works well with ColorHCFR software. However, I had some problems with "spyder3.dll" file as many did. The file came in the driver CD was the new version and it didn't work with ColorHCFR, I had to find the older one.

The Spyder colorimeters are not recommended by many people here because of its accuracy and consistency. However, it has been reported that Datacolor changed their quality control process more than a year ago and the newer Spyder3 (made in the last year or so) has been much better. The folks at Dry Creek Photo recently tested and their findings seem to confirm this. They also found that the Spyder3 is now better than the i1Display2, based on their samples (17 units of i1Display2 and 12 Spyder3). Please check this link for more information about their test:
(The forum doesn't allow me to post URL to the test because I'm a newbie here, please do a Google search with "Display Calibration Hardware Capabilities", you'll find the test done by Dry Creek Photo)

So if anyone wants to calibrate but doesn't find spending 450 bucks for the cheapest Spectro (Colomunki) or nearly a grand for the i1Pro justified, getting a new copy of Spyder3Express for about 70 bucks might be a good idea. The Spyder3Express sensor are practically the same as those of more expensive Pro and Elite versions except the ambient light sensor which we don't use anyway. I have checked with Datacolor to confirm this. I got the Spyder3Express, just to get the sensor, and use it with a third party calibration software package because you can't do much with the Express software. But 70 bucks for the sensor, I don't think you can find any better option than that.

Working with 20-point gray scale was obviously not easy, especially for the first time DIY like me. However, after a series of trial-and-error attempts, I ended up with a workflow that I found efficient. I switched to 100% white and adjusted RGB channels first. When all of them were (or close to) 100% each, I recorded my luminance value and put this into an Excel spreadsheet. Then In Color HCFR, I went to Measure menu->Measure (or something like that, I don't remember exactly) to turn on the 20-point gray scale measurement mode (the default mode is 10 point gray scale). I then opened my luminance graph to see the reference curve (you won't see anything else because you haven't performed any measurement). It put mouse cursor over each point on the luminance reference curve at 0, 5, 10.....95% gray level to check the gray percentage. then record those numbers and put them into the same spreadsheet as the 100% luminance. Use these reference percentages and the 100% white luminance to calculate the reference luminance for each gray level of the gray scale. You'll need these number to do adjustment for a smoother gray scale curve.

After calculating all these luminance values, shifting through the 20-point gray scale to adjust one by one for RGB levels and the luminance. Try to adjust the luminance at each level (by adjusting all 3 channels up or down and retaining them at 100% level) close to the corresponding calculated values so your adjusted curve will be as close to the reference curve as possible. You may find adjusting at 0, 5%, or 10% pretty hard because the sensor can't really provide reliable measurements at these very dark levels. If you see the numbers jumping all over the place, you can skip and go to higher levels. By calculating the reference luminance based on the 100% white, I was able to make the gray scale curve of my TV smooth and bring it very close to the reference one.

Regarding color gamut, I switch it to "wide" instead of "standard" because I found the standard color gamut does not cover 100% of the HD709 color space. The wide gamut on my TV is actually much wider than the HD 709 standard so if you don't use CMS to shrink it and match HD 709, you'll see the much more saturated colors.

After spending a few hours with my new TV, the results were great. Although it's not perfect but I'm very happy with what I got and my inexpensive TV performs very well after my calibration effort. Please check the attached file for my calibration results.
If you have any idea to make my TV look better, I would appreciate that.

 

LG_42PW350_calibration.pdf 403.7880859375k . file

@Andries:

BTW: The calibration report you attached earlier as a zipped bitmap seems a bit strange: the table at the bottom of the report seems to be ordered incorrectly. Is it supposed to look like that? It is difficult to see where the measurement data belongs.

It seems though that the gamma you refer to is just the point gamma for 75% white? The average gamma seems to be calculated to 1.78 - but it is difficult to understand how this value is achieved from the measurement data.

Can you change anything in the software to make the report present the data correctly?

I am not familiar with ICC profiles or the Lacie software, but here are my thoughts:

1) In my skimming I'm not sure how you know if both measurements are using similar signals. My impression was that ICC profiles are not necessarily used by all programs. How have you determined that the HCFR measurements are using the ICC profile?

2) There are different ways to calculate gamma. Since the Lacie result seems to be consistently higher, maybe it's using a different formula. There are a few options in HCFR, but the default is usually how gamma is discussed in this forum.


Some LCDs, or certain LCD settings, may also vary light output depending on the image displayed. Other LCDs or settings would measure the windows and APL series without much variation.

about the bitmap report, it is a screenshot from a PDF file. unfortunately it looks like the last cell in row of "17% heavy dark" is missing, and that shifts the whole table by one. So from there:
25% grey: 6523K 2.14chroma 2.35 gamma
50% gray: 6182K 2.62chroma 2.41 gamma
75% light gray: 6146K 3.51chroma 2.37 gamma
100% white: 6102K 3.57chroma (no gamma here)

I dont know how it calculates the gamma, maybe it uses a different method. that is what I'm trying to discover :/

have you checked the luminance graph in the Color_calibrated_with_ICC_lacie_22.chc file?

answers in the quote...
thanks to all of you for sharing your opinion, it is an interesting discussion!

ps: I made a new measurement on my monitor, and it shows correct values in HCFR and in Lacie also. now I'm confused. I will measure the TV again.