- This topic has 31 replies, 5 voices, and was last updated 4 years, 5 months ago by
.
-
Posts
-
This Mac has 3.0 USB ports. But it’s strange that ArgyllCMS is able to read a lot of colors and only at the end it wouldn’t work anymore.
Regarding the first log I may have disconnected manually the device as nothing was going on.
And for the second log I’m sure that even though the calibration finished just fine in the log the GUI was still showing a progress bar.I’ll give it another try Monday and check again which process are still running at the end.
This Mac has 3.0 USB ports. But it’s strange that ArgyllCMS is able to read a lot of colors and only at the end it wouldn’t work anymore.
I’m almost entirely positive that it is an USB 3.0 issue now. X-Rite has some information on it:
https://www.xrite.com/colormunki-display/support/kb5605They suggest the use of an externally powered USB 2.0 hub as a potential fix to the issue.
I just gave it another try (on the Mac Mini), this time the device connected through a USB 2.0 hub.
And I have the same problem, the dispcal process is still running (or more precisely sleeping) for more than 10 minutes now.I looked at the current temporary directory (using lsof for dispcal) and retrieved the log file below :
https://transfer.sh/fcDn7/e2313h-22062015.logI may try tomorrow to run it again with debug activated.
2015-06-22 09:35:13,934 The instrument can be removed from the screen.
It seems dispcal is hanging somewhere between closing down (which is the above line “Instrument can be removed from the screen”) and before writing the output file (the message “Written calibration file XYZ” never appears, so it doesn’t get that far). A debug run would be nice, although I doubt it’ll have more information. I’ll see if I can compile the information I have now and ask Graeme for help.
I’m wondering if there’s been any update on this problem? I’ve run into it recently too with the most recent downloads of Argyll and DisplayCAL. I’m using a ColorMunki Photo, USB2.0 powered hub, Dell U2713HM monitor, Mac OS X Sierra on a late 2012 Mac Mini. I let it run for 12 hours with no change (stuck at the “you may unplug the instrument” part), and dispcal is still running in the background.
I made the display profile using the command line tools afterwards, so I know they work.
The other thing that would be useful is for dry-run to show all of the commands it’s going to run. Right it only shows the dispcal command. I just thought of that randomly when I was trying to debug the problem.
Calibrite ColorChecker Studio on Amazon
Disclosure: As an Amazon Associate I earn from qualifying purchases.There’s nothing I can do here – there’s no easy way to detect if dispcal hangs (short of ugly hacks like automatically killing a subprocess after a time limit is reached).
I made the display profile using the command line tools afterwards, so I know they work.
I’m pretty sure the failure is sporadic, so I don’t think whether or not dispcal is attached to an actual tty or a pseudo-tty plays a role.
The other thing that would be useful is for dry-run to show all of the commands it’s going to run.
Subsequent commands depend on actual output (as in, generated files) from previous commands in multiple occasions. “Faking” that would be more work than it would be worth.
For the first time I got an error while calibrating for madvr… here is the end of the log
2018-02-27 18:21:23,875 DisplayCAL: Starting interaction with subprocess
2018-02-27 18:21:23,878 Number of patches = 1519
2018-02-27 18:21:23,878 Setting up the instrument
2018-02-27 18:21:24,828 Product Name: Colormunki Display
2018-02-27 18:21:24,828 Serial Number: CM-14.A-02.148311.03
2018-02-27 18:21:24,828 Firmware Version: v1.03
2018-02-27 18:21:24,828 Firmware Date: 05Jun12
2018-02-27 18:21:24,904 Created MadVR window
2018-02-27 18:21:25,210 Place instrument on test window.
2018-02-27 18:21:25,211 DisplayCAL: Waiting for send buffer
2018-02-27 18:21:26,711 DisplayCAL: Skipping place instrument on screen message…
2018-02-27 18:21:26,713 DisplayCAL: Sending buffer: ‘ ‘
2018-02-27 18:21:26,713 Hit Esc or Q to give up, any other key to continue:
2018-02-27 19:37:36,252 Patch 1519 of 1519
2018-02-27 19:37:36,253 The instrument can be removed from the screen.
2018-02-27 19:37:36,461 DisplayCAL: Reached EOF (OK)
2018-02-27 19:37:36,463 …aborted.
2018-02-27 19:37:36,483 Profiling has not been finished.
2018-02-27 20:24:38,009 Executing BaseApp.OnExit()
2018-02-27 20:24:38,009 Running application exit handlers
2018-02-27 20:24:38,009 Waiting for thread ScriptingHost.ConnectionHandler to exit
2018-02-27 20:24:38,085 ScriptingHost.ConnectionHandler exited
2018-02-27 20:24:38,085 Exiting DisplayCAL
2018-02-27 20:24:38,085 Ran application exit handlersWhen I tried again it gave me these errors:
21:05:59,740 Created MadVR window
21:06:00,069 Place instrument on test window.
21:06:00,071 DisplayCAL: Waiting for send buffer
21:06:01,575 DisplayCAL: Skipping place instrument on screen message…
21:06:01,621 DisplayCAL: Sending buffer: ‘ ‘
21:06:01,624 Hit Esc or Q to give up, any other key to continue:
21:07:42,448 Patch 34 of 34
21:07:42,450 The instrument can be removed from the screen.
21:07:42,601 Written ‘madVR 2018-02-27 21-05 S XYZLUT+MTX.ti3’
21:07:42,704 DisplayCAL: Reached EOF (OK)
21:07:42,707 …aborted.
21:07:42,723 Profiling has not been finished.Try ArgyllCMS 32-bit in case you’re using the 64-bit version.
Regardless of that, you should change your madVR output levels config to the correct value.
Try ArgyllCMS 32-bit in case you’re using the 64-bit version.
Regardless of that, you should change your madVR output levels config to the correct value.
I have no idea where this error came from, I didn’t have any problems before the new argyll and displaycal. Where to change the madvr levels?They are on PC Levels (0-255) always. http://imgbox.com/tzlMBi3J I watch movies on my pc. I use a PC monitor.
Where to change the madvr levels?They are on PC Levels (0-255) always
Then you already know where to change them? 🙂 In madVR settings under devices -> your device.
There’s several ways to deal with output levels with regards to madVR:
- Best solution, but TV needs to support 0-255 input correctly(!): Set both madVR and graphics card to 0-255. Both desktop and madVR will look correct.
- TV only supports 16-235 input: Set madVR to 16-235 and graphics card to 0-255. madVR will look correct, but the desktop will have crushed blacks and clipped whites.
- TV only supports 16-235 input, preferred solution when using madVR with 3D LUT and you want the desktop to look correct as well: Set madVR to 0-255 and graphics card to 16-235. Both desktop and madVR will look correct.
Where to change the madvr levels?They are on PC Levels (0-255) always
Then you already know where to change them?
In madVR settings under devices -> your device.There’s several ways to deal with output levels with regards to madVR:
- Best solution, but TV needs to support 0-255 input correctly(!): Set both madVR and graphics card to 0-255. Both desktop and madVR will look correct.
- TV only supports 16-235 input: Set madVR to 16-235 and graphics card to 0-255. madVR will look correct, but the desktop will have crushed blacks and clipped whites.
- TV only supports 16-235 input, preferred solution when using madVR with 3D LUT and you want the desktop to look correct as well: Set madVR to 0-255 and graphics card to 16-235. Both desktop and madVR will look correct.
I’m not using madvr with a TV, I’m using it with my monitor which requires PC Levels 0-255, like I said before, I never had any problems with my calibrations , today I changed argyll cms to 32 bit version,and the calibration was OK. This time I didn’t get the levels error. I hope everything will be good from now on.
Hi Florian,
I’ve got the same issue on Mac OS X High Sierra with i1Pro in slow calibration mode. At the finish of calibration process dispcal got stuck with the message “Black drift was nan DE”. Is there any solution for this excluding faster calibration modes?
The versions of software were 2.1.1 for Argyll CMS and 3.8.7 for DisplayCAL
The i1 Pro is not very suitable for longer calibration runs because its black calibration drift may become too high. You can counteract that by enabling black drift compensation. Ideally, you should only use the i1 Pro to create a colorimeter correction, and then use a colorimeter for the actual calibration/profiling.
Thank you! Black drift compensation is always enabled and slow calibration does finish successful on my Macbook pro but hangs on my friend’s one with the same device and same options enabled. Medium speed mode runs OK both ways but curves match worse. Only difference is that I calibrate my Macbook in the full darkness and with ‘set black background’ checkbox enabled. It has been working through several versions of macOS, ArgyllCMS and DisplayCAL and even DispcalGUI and with both i1Pro and ColorMunki spectrometers. Maybe some macOS bug nevertheless?
Some versions of Dcal or ArgyllCMS halt on with the device recalibration request. Does the last version so? I have not encounter that yet.
BTW, what does it mean, ‘nan DE’? Is it Australian version of ‘none Delta E’?
