HCFR - Open source projector and display calibration software - Page 67 - AVS Forum
Forum Jump: 
 92Likes
Reply
Thread Tools
post #1981 of 5008 Old 05-27-2013, 08:36 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 679
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 101 Post(s)
Liked: 88
Quote:
Originally Posted by zoyd View Post

Have you checked that you can get a good refresh sync with your probe using the ArgyllCMS 1.5.1 command line tools? The current documentation says that the colormunki display is not able to measure the refresh period but I don't know if that refers to the 1.4 or 1.5 distribution.
The Colormunki Display is not capable of measuring the refresh rate in any sane amount of time, due to the artificial measurement delays built into it. I've put the capability into the ArgyllCMS colinst library to set an integration time, so it's conceivable that HCFR could determine the refresh rate some other way (ie. the user enters a value or measures it with some other instrument like the i1pro or Colormunki spectro.) and then set that in the driver, and this should (in theory) give better readings for refresh type displays.

Note that the adaptive measurement algorithm for the Colormunki Display cannot be as sophisticated as the one used with the i1Pro Display, again due to the extra measurement delays, and this may be another source of some inaccuracy.
gwgill is offline  
Sponsored Links
Advertisement
 
post #1982 of 5008 Old 05-28-2013, 06:23 AM
Advanced Member
 
PeterLewis's Avatar
 
Join Date: Dec 2007
Posts: 656
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 25
Thanks Graeme for the thorough understanding...

For new munki/id3 users thinking how to get your colorimeter to sit flush on the screen with no gap,try using painters tape from edge to edge of the tv over the meter, as it keeps it flush and does not touch the screen and leaves no residue on parts of the bezel after use.

Samsung 60F8500
Panasonic 55GT30
Panasonic BDT310
PS3 (PS4 coming soon)
Sony STR-DG1000 AVR
Mirage Omnisat V2 7.1 set up with Energy S10 Sub
PeterLewis is online now  
post #1983 of 5008 Old 05-28-2013, 04:03 PM
Member
 
Leondavid's Avatar
 
Join Date: Mar 2013
Posts: 37
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

Good idea, profile in 3D mode with glasses as reference then and let us know how it goes,

zoyd hello how are you? in short, change my TV and now I'm enjoying a panasonic plasma, is incredible!
The model is the st50, 60 ".
I'm making him filming! a week ago that I have!
When you spend over 200 hours, I will do my first calibration.
I wanted to know who would you recommend as options in HCFR when do the first calibration? choose the spectral plasma sample? and if so, I have also put an array as correcion, to better measure values ​​or leave it as it comes, without default values​​?
I see that the controls have not 10pt in this model will be more difficult to set a level range, if so, should I use bt1886 and seek a curve as range?
Thank you very much as always.
Leondavid is offline  
post #1984 of 5008 Old 05-28-2013, 05:00 PM
Advanced Member
 
CheYC's Avatar
 
Join Date: Jan 2008
Location: Philadelphia, PA
Posts: 629
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 14 Post(s)
Liked: 51
Quote:
Originally Posted by Leondavid View Post

zoyd hello how are you? in short, change my TV and now I'm enjoying a panasonic plasma, is incredible!
The model is the st50, 60 ".
I'm making him filming! a week ago that I have!
When you spend over 200 hours, I will do my first calibration.
I wanted to know who would you recommend as options in HCFR when do the first calibration? choose the spectral plasma sample? and if so, I have also put an array as correcion, to better measure values ​​or leave it as it comes, without default values​​?
I see that the controls have not 10pt in this model will be more difficult to set a level range, if so, should I use bt1886 and seek a curve as range?
Thank you very much as always.

I'm pretty sure you should use either the spectral sample OR the correction matrix, but not both at the same time. Compare them. I found my results were very similar with the matrix as with the spectral sample on my 55ST50.
Leondavid likes this.

CheYC is online now  
post #1985 of 5008 Old 05-28-2013, 05:20 PM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
yes, use the spectral sample alone first. See how it looks and compare with generic plasma matrix (again alone). Without a 10 pt control you can't get a true bt.1886 but that's ok.
zoyd is offline  
post #1986 of 5008 Old 05-28-2013, 05:46 PM
Member
 
Leondavid's Avatar
 
Join Date: Mar 2013
Posts: 37
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

yes, use the spectral sample alone first. See how it looks and compare with generic plasma matrix (again alone). Without a 10 pt control you can't get a true bt.1886 but that's ok.
zoyd thanks! then when you first calibration, I do it with the spectral sample and then another, but with the matrix correcion! So then compare and see what results I have! but surely will like both!!
Thank you very much as always!
Leondavid is offline  
post #1987 of 5008 Old 05-28-2013, 05:50 PM
Member
 
Leondavid's Avatar
 
Join Date: Mar 2013
Posts: 37
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by CheYC View Post

I'm pretty sure you should use either the spectral sample OR the correction matrix, but not both at the same time. Compare them. I found my results were very similar with the matrix as with the spectral sample on my 55ST50.
thank you very much! then, if you can and if you got it, I pass the data file with your tv? ie save the file with the HCFR with the results you got to know well that I can get!
All I did was measure the values ​​of the box, they pass the 200 hours, I'll calibration.
The best way I found was true cinema mode! because the panasonic uses are those that bring a more so, it is the custom mode!
Leondavid is offline  
post #1988 of 5008 Old 05-28-2013, 08:52 PM
Senior Member
 
todd95008's Avatar
 
Join Date: Oct 2004
Posts: 231
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 19
I'm using the following test discs and want to know if the "round down level" (AVSHD, DVE) box should be checked for which discs ?

Get Gray
AVS HD709
GCD test disc
Mascior's test disc
Chad B test disc

I also have an old DVE disc (DVD version) but don't use it since the levels are in IRE steps not percent.

Thanks
Todd
todd95008 is online now  
post #1989 of 5008 Old 05-29-2013, 05:06 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
You should use round down for AVSHD and the DVE disks. Unchecked for all others as I'm not aware of any other test disk that rounds levels down.
zoyd is offline  
post #1990 of 5008 Old 05-29-2013, 07:17 AM
Advanced Member
 
PeterLewis's Avatar
 
Join Date: Dec 2007
Posts: 656
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 15 Post(s)
Liked: 25
Question in regards to making a compromise in greyscale adjustments...

I did a calibration where I got the 20,30,70 and 80 IRE all had a 0.5 delta E but the 60 IRE had the blue at 95% giving me a delta E of 4.1...

So I did a compromise and raised the blue drive up a couple of notches to get the 60 IRE deltaE under 3...But due to the compromise all my 0.5 readings are now in the 1.5-2.5 area.....

So should I have left my first calibration alone where most IRE where at 0.5 more or less but the 60 had a delta of 4.1 or was the compromise the way to go since now the greyscale from 20-80 are all under deltaE of 3 ?

Samsung 60F8500
Panasonic 55GT30
Panasonic BDT310
PS3 (PS4 coming soon)
Sony STR-DG1000 AVR
Mirage Omnisat V2 7.1 set up with Energy S10 Sub
PeterLewis is online now  
post #1991 of 5008 Old 05-29-2013, 07:34 AM
Senior Member
 
Make73's Avatar
 
Join Date: Jan 2010
Location: Finland
Posts: 209
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 21
Quote:
Originally Posted by PeterLewis View Post

...or was the compromise the way to go since now the greyscale from 20-80 are all under deltaE of 3 ?

I would choose compromise.
You can´t notice difference under dE3 with eyes.
Do you have reference meter?
However if you calibrate all under dE3 and measure again after couple of hundred hours of usage those are all changed, at least with plasma sets.
Make73 is offline  
post #1992 of 5008 Old 05-29-2013, 10:35 AM
10k
Advanced Member
 
10k's Avatar
 
Join Date: Sep 2005
Posts: 690
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 4 Post(s)
Liked: 38
Quote:
Originally Posted by PeterLewis View Post

Question in regards to making a compromise in greyscale adjustments...

I did a calibration where I got the 20,30,70 and 80 IRE all had a 0.5 delta E but the 60 IRE had the blue at 95% giving me a delta E of 4.1...

So I did a compromise and raised the blue drive up a couple of notches to get the 60 IRE deltaE under 3...But due to the compromise all my 0.5 readings are now in the 1.5-2.5 area.....

So should I have left my first calibration alone where most IRE where at 0.5 more or less but the 60 had a delta of 4.1 or was the compromise the way to go since now the greyscale from 20-80 are all under deltaE of 3 ?
In theory dE below 3 is not visible to human eye so your revised settings are probably superior. Ultimately the final test is looking at actual content.

You might find this post helpful - http://www.avsforum.com/t/1399319/official-samsung-unxxes8000-owners-thread/15270#post_23338059
In the comparison I posted there, I was absolutely able to see the extra green in darker areas of the picture, even though the IRE20 dE was just over 3. My revised settings are much better visually.
10k is offline  
post #1993 of 5008 Old 05-29-2013, 01:38 PM
Senior Member
 
todd95008's Avatar
 
Join Date: Oct 2004
Posts: 231
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 19
Quote:
Originally Posted by zoyd View Post

You should use round down for AVSHD and the DVE disks. Unchecked for all others as I'm not aware of any other test disk that rounds levels down.

Great, I have settled on using the GCD disk for most of the cal (best APL window patterns for plasma) plus some of the patterns in the other disks.

Thanks
Todd
todd95008 is online now  
post #1994 of 5008 Old 05-30-2013, 02:44 PM
Newbie
 
harry1985's Avatar
 
Join Date: Sep 2011
Posts: 9
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
quick question, is colormunki smile supported by HCFR right now?
If not, could I calibrate my projector(not bought yet, most likely it will be a Sony HW50ES) with this using ArgyII? It seems that the latest ArgyII already support the colormunki smile.

Thanks,
Harry
harry1985 is offline  
post #1995 of 5008 Old 05-30-2013, 03:58 PM
Member
 
Roland.Online's Avatar
 
Join Date: Feb 2010
Posts: 31
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 18
Hi all

First, just like to say thanks to JohnA, zoyd and all the other contributors to HCFR.

I have an installation of HCFR 3.0.4.0 working on my Win7 64 laptop working successfully with my i1Display Pro (i1Display 3).

However I'm struggling to get the same device working on my Win7 64 desktop machine. Here's a rundown:
1. both machines have the same OS: Windows 7 64-bit (Pro on laptop, Ultimate on desktop) with Service Pack 1 and all Windows updates.
2. no virus protection running (Microsoft Security Essentials installed but disabled for the purpose of testing)
3. installed i1Profiler 1.4.2 on both from: http://www.xrite.com/product_overview.aspx?ID=1454&Action=support&SoftwareID=1244
4. installed HCFR 3.0.4.0 on both from: http://sourceforge.net/projects/hcfr/files/
5. no other drivers explicitly installed.
6. the device shows up in Device Manager under Human Interface Devices as:
HID-compliant device
USB Input Device
7. the device has VendorID = VID_0765, ProductID = PID_5020
8. stopped "X-Rite Device Services Manager" service and terminated i1ProfilerTray utility.

On the laptop I did not need to install any other USB drivers (ArgyllCMS or otherwise) and I did not need to copy any X-Rite driver files to the HCFR program folder.

On the laptop, both iProfiler and HCFR recognise and read from the device perfectly. smile.gif

However on the desktop machine only iProfiler recognises and reads from the device: HCFR does not. frown.gif

Further testing and changes:
1. HCFR 3.0.4.0 fails to list any devices in the sensor list: it's completely empty.
2. there is no output in stderr.log (it's empty).
3. the spotread tool supplied with HCFR fails with:
Code:
Read Print Spot values, Version 1.4.0
        Author: Graeme W. Gill, licensed under the GPL Version 2 or later
        usage: spotread [-options] [logfile]
         -v                   Verbose mode
         -s                   Print spectrum for each reading
         -c listno            Set communication port from the following list (default 1)
        libusb:debug [libusb_init]
        libusb:debug [init_polling] Will use CancelIo for I/O cancellation
        libusb:debug [usbi_pipe] usbi_pipe() called
        libusb:debug [windows_clock_gettime_threaded] no hires timer available on this platform
        libusb:debug [usbi_pipe] filedes[1] = filedes[0] = 3
        libusb:debug [usbi_add_pollfd] add fd 3 events 1
        libusb:debug [libusb_init] created default context
        libusb:debug [libusb_get_device_list]
        libusb:debug [usb_enumerate_hub] busnum 0 devaddr 0 session_id 0
        libusb:debug [usb_enumerate_hub] allocating new device for session 0
        libusb:debug [initialize_device] active config: 1
        libusb:debug [usb_enumerate_hub] 4 ports Hub: \\.\ENUSB3#ROOT_HUB30#5&99F7FE2&0&0003#{F18A0E88-C30C-11D0-8815-00A0C906BED8}
        libusb:debug [usb_enumerate_hub] busnum 0 devaddr 1 session_id 1
        libusb:debug [usb_enumerate_hub] allocating new device for session 1
        libusb:debug [initialize_device] active config: 1
        libusb:error [cache_config_descriptors] could not access configuration descriptor (dummy): [50] The request is not supported.

the spotread process then crashes with:
Code:
Problem signature:
  Problem Event Name:   BEX
  Application Name:     spotread.exe
  Application Version:  0.0.0.0
  Application Timestamp:        510da25c
  Fault Module Name:    spotread.exe
  Fault Module Version: 0.0.0.0
  Fault Module Timestamp:       510da25c
  Exception Offset:     0005d981
  Exception Code:       c0000409
  Exception Data:       00000000
  OS Version:   6.1.7601.2.1.0.256.1
  Locale ID:    2057
  Additional Information 1:     3a01
  Additional Information 2:     3a01407e8dd8a067abbdc40e7754bfeb
  Additional Information 3:     ec3f
  Additional Information 4:     ec3f5fe6c70822331bfe1fac3cc48112

4. the spotread tool supplied with ArgyllCMS 1.5.1 recognises and reads from the device perfectly.
5. I removed all software and reinstalled the newer HCFR 3.0.4.2 from: https://sourceforge.net/projects/zoyd000.u/files/downloads/
6. HCFR 3.0.4.2 crashes before it shows the the sensor list:
Code:
Problem signature:
          Problem Event Name:   BEX
          Application Name:     ColorHCFR.exe
          Application Version:  3.0.4.1
          Application Timestamp:        51549ba5
          Fault Module Name:    ColorHCFR.exe
          Fault Module Version: 3.0.4.1
          Fault Module Timestamp:       51549ba5
          Exception Offset:     0011ef01
          Exception Code:       c0000409
          Exception Data:       00000000
          OS Version:   6.1.7601.2.1.0.256.1
          Locale ID:    2057
          Additional Information 1:     c5a0
          Additional Information 2:     c5a051f32f6218416b2b41e9e15b5971
          Additional Information 3:     c28a
          Additional Information 4:     c28a7116325629682b74468e2634125b

7. there is no output in stderr.log but a minidump.dmp was generated.
8. I replaced the default hidusb driver on the USB Input Device with the ArgyllCMS.inf driver supplied with ArgyllCMS 1.5.1 (libusb0 1.2.6.0) which installed successfully as "Eye-One Display 3 (Argyll)".
9. with the Argyll driver HCFR 3.0.4.2 crashes as before, the spotread tool supplied with HCFR fails as before, and the spotread tool supplied with ArgyllCMS 1.5.1 recognises and reads from the device perfectly.
10. I've also tried plugging the device into various USB 2.0 and 3.0 ports with and without intermediate hubs.

I've read as much of this forum thread as I can but unsure why the two systems would behave so differently, or how to resolve this problem. confused.gif

Any ideas?

Any help very much appreciated. I can post minidumps if that helps.

Many thanks
Roland
Roland.Online is offline  
post #1996 of 5008 Old 05-30-2013, 04:10 PM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
hmmm...the display pro does not need any driver installation since it uses whatever windows HID driver is available. Try uninstalling any other driver associated with it like libusb0
zoyd is offline  
post #1997 of 5008 Old 05-31-2013, 07:52 AM
Member
 
Roland.Online's Avatar
 
Join Date: Feb 2010
Posts: 31
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 18
spotread_profile.txt 32k .txt file
Quote:
Originally Posted by zoyd View Post

hmmm...the display pro does not need any driver installation since it uses whatever windows HID driver is available. Try uninstalling any other driver associated with it like libusb0

I figured that was the case, but there is no evidence of libusb0 on my system at all. I removed the ArgyllCMS.inf driver altogether, and all files like libusd0.*

However, does that imply that HCFR would not work even if a completely unrelated driver required libusb0?

Looks like a buffer overrrun. I profiled spotread using Dependency Walker, the log is attached.

I'm using the latest 3.0.4.2 fork of HCFR, but the version of spotread is 1.4.0, not the latest from ArgyllCMS, 1.5.1.

Thanks
Roland
Attached Files
File Type: txt spotread_profile.txt (32.1 KB, 11 views)
Roland.Online is offline  
post #1998 of 5008 Old 06-01-2013, 09:34 AM
Member
 
Roland.Online's Avatar
 
Join Date: Feb 2010
Posts: 31
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 18
So I tried rolling back to HCFR 3.0.4.0, ensuring no presence of libusb, but with similar results: HCFR shows an empty sensor list, and spotread crashes.

I tried with hidusb and with the Argyll drivers supplied with HCFR - same result. I also removed entirely an attached USB hub, and plugged my i1D3 directly.

I've attached a Dependency Walker profile log of the spotread app crash - different to the buffer overrun in my last post.

Hope this makes sense to someone?

Thanks
Roland

spotread_profile_hidusb.txt 58k .txt file
Attached Files
File Type: txt spotread_profile_hidusb.txt (57.6 KB, 7 views)
Roland.Online is offline  
post #1999 of 5008 Old 06-01-2013, 09:43 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
Quote:
Originally Posted by Roland.Online View Post

HCFR shows an empty sensor list, and spotread crashes.

This is the behavior I get when trying to load the ArgyllCMS 1.5.1 driver for my i1pro. Try reinstalling drivers from the HCFR drivers folder which contains the ArgyllCMS 1.4 version that work with HCFR.
zoyd is offline  
post #2000 of 5008 Old 06-02-2013, 09:28 AM
Member
 
Roland.Online's Avatar
 
Join Date: Feb 2010
Posts: 31
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 18
Problem solved. Well, at least worked around.

It seems that both the libusb and ArgyllDMC drivers conflict with a WD MyBook USB 3 external drive I had attached. Removing this allowed HCFR and spotread to successfully communicate with my i1D3.

Works fine now with both 3.0.4.0 and 3.0.4.2, with the standard hidusb drivers.

So the latest ArgyllCMS 1.5.2 works fine with that device attached, but HCFR and the distributed Argyll tools do not - are these older versions?

Cheers
Roland
Roland.Online is offline  
post #2001 of 5008 Old 06-02-2013, 09:48 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
yes, stated above^, still using ArgyllCMS 1.4 distro until John can sort out some issues in porting the new drivers.
zoyd is offline  
post #2002 of 5008 Old 06-02-2013, 10:23 AM
Advanced Member
 
janos666's Avatar
 
Join Date: Jan 2010
Location: Hungary
Posts: 599
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
How is this possible?



I experienced some very strange behavior when I used my i1d3 on some LCD TVs, so I started to do some sanity checks.
I put it on my laptop display (LED backlit LCD) and started to run continuous free measures to see if there are random false readings.
It ran for a while until I noticed that this color is on the edge of the sRGB gamut while I remembered to have narrower gamut on this display. So, I did a gamut reading and confirmed it: my i1d3 reads out-of-gamut colors which is impossible (well, there could be some scenarios when this would be possible but this laptop display does not have built-in gamut emulation, nor it ever had 100% sRGB gamut or non-additive behavior when measured with colprof and dispread in the past to calibrate and profile it).

So... did my i1d3 break or is HCFR 3.0.4.0 known to have serious bugs?

I just tried the same with a different probe placement. It shows the same problem. But why...?



I didn't profile the sensor against a reference spectro, nor I loaded any ccss files. The instrument works "raw". (So it's not a faulty correction matrix or insane spectrum based correction.)

"DIY certified hobby-calibrator" (based on ChadB's "warning signs" list
janos666 is offline  
post #2003 of 5008 Old 06-02-2013, 10:41 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
check it with spotread, I doubt it's an HCFR issue (which is just using a ported spotread anyway). Also, the d3 code changed a little bit between 3.0.4.0 and 3.0.4.2 so check that out too. you might need to profile it as well.
zoyd is offline  
post #2004 of 5008 Old 06-02-2013, 10:44 AM
Advanced Member
 
janos666's Avatar
 
Join Date: Jan 2010
Location: Hungary
Posts: 599
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

the d3 code changed a little bit between 3.0.4.0 and 3.0.4.2 so check that out too

I just did. It crashes on startup without any error notifications (Windows 8 x64 -> I ran both the setup and the executable with admin rights).
I will do some similar testing with CalMan.

Edit:

I managed to run the new version after cleaning the remains of the old and reinstalling the new.

But yet again, something is very wrong. This time I used the supplied "white LED" sensor profile (this should not make a difference in this matter, but still...). And I can easily measure impossible (out-of-gamut) colors again.

May be this is a limitation of the i1d3 device. May be it needs solid color patches or othervise the differently filtered sensors get different light samples from different parts of the display.
But this is not good if you consider that many displays use dithering (or even has uniformity problems too).


-> Just to make it clear: I am currently measuring the background image of the CIE chart graph (I thought it's a good idea to test the instrument if it looks OK or broken) and the gamut readings were took without moving the instrument on the display (100% filled overlay).

Otherwise, I like this new version with the saturation target boxes and tweaked integration times. smile.gif
But I still miss the ability to measure a color at different luminance levels (similarily to the saturation scale).

"DIY certified hobby-calibrator" (based on ChadB's "warning signs" list
janos666 is offline  
post #2005 of 5008 Old 06-02-2013, 08:24 PM
Advanced Member
 
gwgill's Avatar
 
Join Date: Jan 2013
Location: Melbourne, Australia
Posts: 679
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 101 Post(s)
Liked: 88
Quote:
Originally Posted by janos666 View Post

But yet again, something is very wrong. This time I used the supplied "white LED" sensor profile (this should not make a difference in this matter, but still...). And I can easily measure impossible (out-of-gamut) colors again.
It's not quite clear what you are doing, but such instruments are intended to measure spatially uniform, temporally stable patches.

While I know that the i1d3 has an intermediate diffusing screen in its optical path that is meant to equalize the composition of the light falling on its three sensors, it's not likely to be exactly 100% effective.

So if you deliberately try to throw different levels of light on the different sensors by exposing it to a non-uniform patch, then there is little surprise that you get odd readings. The solution is simple - don't do that.

Dithering patterns are by their nature spatially fine, and will not cause any issues with an otherwise uniform color patch.

Graeme Gill.
gwgill is offline  
post #2006 of 5008 Old 06-03-2013, 03:52 AM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
Quote:
Originally Posted by janos666 View Post


-> Just to make it clear: I am currently measuring the background image of the CIE chart graph (I thought it's a good idea to test the instrument if it looks OK or broken) and the gamut readings were took without moving the instrument on the display (100% filled overlay).

right, didn't realize you were measuring non-uniform image. Doesn't work with the D3 because the separate sensors have somewhat different FOVs so it needs a uniform source. I found this out one day when I was measuring real black and white content to test out my cable box white point and wondered why it was all over the place. As soon as I froze the image and found a uniform patch it was fine.
zoyd is offline  
post #2007 of 5008 Old 06-03-2013, 09:57 PM
Advanced Member
 
janos666's Avatar
 
Join Date: Jan 2010
Location: Hungary
Posts: 599
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 10
Quote:
Originally Posted by zoyd View Post

right, didn't realize you were measuring non-uniform image. Doesn't work with the D3 because the separate sensors have somewhat different FOVs so it needs a uniform source. I found this out one day when I was measuring real black and white content to test out my cable box white point and wondered why it was all over the place. As soon as I froze the image and found a uniform patch it was fine.

At least we know this now. I will be more careful with this.
Otherwise, it's frustrating that HCFR + my i1d3 work fine for hours when I test it at home, yet I had troubles with it last time I visited other TVs.
May be it's notebook power management issue. I guess I already actively measured when I connected the notebook to AC power, so may be the power save mode didn't disengage?frown.gif

"DIY certified hobby-calibrator" (based on ChadB's "warning signs" list
janos666 is offline  
post #2008 of 5008 Old 06-04-2013, 04:34 PM
Member
 
Some dillweed's Avatar
 
Join Date: Dec 2009
Posts: 46
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Okay, I've been messing around with this stuff for way too long because I can't seem to get it right. So, I have some questions and I'd appreciate answers if anyone can provide them.

I'm using my TV (LG 42LK450) only as a PC monitor, and it's being fed a full RGB signal. My MPC-HC install has also been set to full range (0-255) while trying to do all of this.

-Which patterns should I be using to calibrate this thing? If I knew what RGB values produced each 10% step, I'd probably just make static images for each 10% between full RGB black and full RGB white and use those as my steps in HCFR. I've tried both the GCD patterns and AVS HD patterns, and they result in slightly different colour balances in each gray and different gamma results, with greater differences as you approach the low end. Is there another option that will provide more consistent results for my specific use case? If I could avoid relying on video files for this then I would, because given the lack of consistency between the two sets that I've already tried, they don't seem like the right choice when my focus isn't just watching movies and whatever.
-The GCD has grayscale only going up to 100%, and the AVS HD patterns go up to 109%, but HCFR refers to each grayscale step in terms of IRE. Which patterns are you supposed to use when you're using full RGB instead of limited and you want your full range?
-Should I even be using HCFR or should I learn to use dispcalGUI instead? I'm playing PC games and doing general desktop stuff a lot more than I am watching movies or shows now, so I'm not trying to limit my grayscale range to the limited 16-235 video levels. I'm just not sure what I should be using for this.

I've just done the whole calibrate, check, re-calibrate thing over and over ad nauseum since I got my Colormunki Display. I guess I'm trying to find a way that possibly doesn't involve videos, since the two sets I've already tried don't seem to match up. Either that, or just a confirmation of how I should be doing this and which patterns to rely on when balancing each IRE on my LG's 10-point IRE white balance.
Some dillweed is offline  
post #2009 of 5008 Old 06-04-2013, 04:47 PM - Thread Starter
AVS Special Member
 
zoyd's Avatar
 
Join Date: Sep 2006
Location: Planet Dog
Posts: 5,018
Mentioned: 40 Post(s)
Tagged: 0 Thread(s)
Quoted: 441 Post(s)
Liked: 423
If your source is your HTPC then calibrate using the view images generator within HCFR. You should also make sure there are no calibration (gamma) tables loaded to the video card or if there are that they are used for all content you want to calibrate for, MPC-HC etc.

If your video card sends full range and MPC-HC is set for full range then set HCFR view images to 0-255.
Some dillweed likes this.
zoyd is offline  
post #2010 of 5008 Old 06-04-2013, 05:17 PM
Member
 
Some dillweed's Avatar
 
Join Date: Dec 2009
Posts: 46
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Ohhhh, man. Thank you for the heads-up about the built-in generator, zoyd. My OCD was getting the better of me and I had been driving myself crazy with all of this.
Some dillweed is offline  
Reply Display Calibration

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