- This topic has 12 replies, 4 voices, and was last updated 8 years ago by
.
-
Posts
-
I’m using an Asus vs278q-p, a basic backlit LED LCD, with a Spyder5. I’m trying to do color correction for a short film I did at an orphan’s home in Mexico last summer. So far I’ve probably spent 200hrs trying to calibrate the color and redoing the correction. I’ve done the calibration and correction multiple times on 3 separate displays and when I’ve had the color looking good on one it just hasn’t looked right on the others or additional screens I’ve tried. The colors often look over saturated and the lighting dark on the other screens.
I’m trying to figure out if I’m doing everything as well as possible for a low-cost display like this and whether the results show it’s working. I’ll show you what I have so far and any advice and assistance any of you can offer is greatly appreciated.
On the Display/Instrument tab I have the Mode set as LCD (white LED), and Correction set as Auto (none). On Calibration everything is set as (as Measured). On Profiling the settings are Lab LUT for Profile Type, Profile Quality as High and Testchart as Auto-optimized, with 1553 patches. I have the other Colorimeter corrections imported and calibrated the RGB and contrast before starting the profiling, which was set for
Here are a few short clips I uploaded and the results from my verification log:
Verify calibration
15:30:57,155 ————————————————————————————————————————
15:30:57,155 Command line:
15:30:57,155 dispcal
15:30:57,155 -v2
15:30:57,156 -d1
15:30:57,156 -c1
15:30:57,156 -ye
15:30:57,156 -P0.977892409727,0.350096711799,1.74303405573
15:30:57,156 -t
15:30:57,157 -b
15:30:57,157 -z
15:30:57,157
15:30:57,169 DisplayCAL: Starting interaction with subprocess
15:30:57,255 Argyll ‘V1.8.3’ Build ‘OS X 64 bit’ System ‘Darwin Darwin Kernel Version 14.5.0:
15:30:57,255 Mon Jan 11 18:48:35 PST 2016; root:xnu-2782.50.2~1/RELEASE_X86_64 14.5.0 x86_64’
15:30:57,256 Setting up the instrument
15:30:58,768 Instrument Type: Datacolor Spyder5
15:30:58,768 Serial Number: 50072755
15:30:58,769 Hardware version: 0x0a0f
15:30:59,028 Place instrument on test window.
15:30:59,030 DisplayCAL: Waiting for send buffer
15:31:00,566 DisplayCAL: Sending buffer: ‘ ‘
15:31:00,622 Hit Esc or Q to give up, any other key to continue:
15:31:00,623 Display type is ‘e’
15:31:00,623 Target white = native white point
15:31:00,623 Target white brightness = native brightness
15:31:00,624 Target black brightness = native brightness
15:31:00,624 Black point device hack is enabled
15:31:00,624 Target advertised gamma = 2.400000
15:31:00,625 Commencing display verification
15:31:00,630 Switching to native response for base measurements
15:32:34,140 Patch 9 of 9
15:32:34,140 Switching back to calibration being verified
15:32:34,142 Black = XYZ 0.2275 0.2071 0.4607
15:32:34,143 Red = XYZ 51.472 27.486 2.225
15:32:34,143 Green = XYZ 40.800 87.761 7.750
15:32:34,143 Blue = XYZ 17.238 5.561 88.305
15:32:34,144 White = XYZ 124.091 128.903 109.157
15:32:34,144
15:32:34,145 Initial native brightness target = 128.902526 cd/m^2
15:32:34,145 Target white value is XYZ 124.091345 128.902526 109.157490 [xy 0.342650
15:32:34,145 0.355936]
15:32:34,146 Adjusted target black XYZ 0.2275 0.2071 0.4607, Lab 1.451 0.881 -4.071
15:32:34,146 Target black after min adjust: XYZ 0.2275 0.2071 0.4607, Lab 1.451 0.881 -4.071
15:32:34,147 Gamma curve input offset = 0.000000, output offset = 0.001607, power = 2.409968
15:32:34,147
15:32:34,148 Doing verify pass 1/1 with 100 sample points
15:44:41,148 Patch 100 of 100
15:44:41,148 Verification results:
15:44:41,149 Brightness error = -0.065694 cd/m^2 (is 128.836832, should be 128.902526)
15:44:41,149 White point error = 0.070410 deltaE
15:44:41,149 Maximum neutral error (@ 0.194481) = 8.120399 deltaE
15:44:41,149 Average neutral error = 4.598861 deltaE
15:44:41,150 Number of measurements taken = 100
15:44:41,150 The instrument can be removed from the screen.
15:44:41,155 DisplayCAL: Reached EOF (OK)Is this going to end beinng played on TVs? Not on computers? My experience on doing color correction for TV is that I have to trust my instruments and not that much my eyes or what I see, I have a blackmagic vectorscope and I got used to look at it more than the image itself, at least to color correct video, of course have a computer monitor looking good is important, but you should work looking at something like what yiur audience is going to be looking at.
Hello Victor, the film will be shown at a few festivals, in San Diego, Vancouver, Seattle, etc, if accepted. It will also eventually be available on Facebook and the organization’s website.
We’d need more information about your workflow. I see you’re on Mac OS X. What software are you using for viewing and editing of video and still material? Are you using 3D LUTs for color management or are you relying on the display profile? What settings did you use in DisplayCAL to calibrate & profile (the easiest way to relay all your settings is to attach the profile you created by clicking “Create compressed archive” next to “Settings”)?
Hello Florian, I’m using FCP X for editing video (I don’t do much still work), with the video encoded to ProRes HQ and output to MP4. Playback on this machine with QuickTime looks good, but not so good on my other PC machines with different players.
I’m not sure I know the difference between using 3D LUT and the Display Profile. I have been installing the profile system-wide after the color patch process completes. I can see the DisplayCal profile selected in the Dislpay/Color section of system preferences.
I’ve attached the ICC and Log files.
A word of caution with QT player, it is not color management aware, it is the worst player ever invented in my opinion, things will look washed out outside of a MAC environment. Are you sure FCP X is aware of ICC or LUT profiles installed? Lastly, you will have to deal with the fact that you can not control what others do with their displays, so that is why I told you if is video, use a video card and check you levels and colors with it. I recommend my clients on PC to use VLC player which is much better than quicktime and uses your hardware, and does not destroy the color you worked so hard to get.
Of course this is totally unrelated to the fact if your monitor is correctly calibrated or not, it is just the fact that what you see on a QT player on a MAC will not look the same on a PC.
A word of caution with QT player, it is not color management aware
On Windows that may be true (haven’t checked), but on Mac OS X it definitely is fully color managed. FCPX as well. The problem with it though is that Apple is using a decoding gamma of 1.96 instead of the more common “legacy” 2.2 or even BT.1886, meaning footage played back in QT under OS X won’t likely match anything else out there.
Playback on this machine with QuickTime looks good, but not so good on my other PC machines with different players.
Most video players in the Windows world are not color managed. A notable exception is MPC-HC where you can enable it.
I’m not sure I know the difference between using 3D LUT and the Display Profile. I have been installing the profile system-wide after the color patch process completes. I can see the DisplayCal profile selected in the Dislpay/Color section of system preferences.
As far as QuickTime and OS X are concerned, the display profile is all that is needed. A video 3D LUT would only come into play on dedicated systems (i.e. a hardware LUT box or software solution).
I’ve attached the ICC and Log files.
Thanks, looking good as far as I can tell.
The problem with it though is that Apple is using a decoding gamma of 1.96 instead of the more common “legacy” 2.2 or even BT.1886, meaning footage played back in QT under OS X won’t likely match anything else out there.
I’ve attached the ICC and Log files.
Thanks, looking good as far as I can tell.
Hello Florian, does this mean I’ve done everything possible with regard to getting the best results for this screen and machine for playback on film festival screens? Do I need to make any adjustments for the Gamma issue you mentioned?
Hello Florian, does this mean I’ve done everything possible with regard to getting the best results for this screen and machine for playback on film festival screens?
Without knowing anything about their playback setup, I’m unable to tell.
Do I need to make any adjustments for the Gamma issue you mentioned?
I don’t think you can, but I don’t have experience with FCPX so it’s hard to tell.
In my experience, working for computer playback is not the same as working for “video” and is hard to tell without having a proper video monitor. If this goes to a festival it is possible that will be projected with video projectors, most of these devices are usually calibrated around D65 and close to sRGB and do not have the same level of contrast you can achieve on yiur computer monitor, neither they can have or use a full black or a full white, that’s why even though I deliver my TV ads which I never know where they are aging to play, I have certain specs to check, and get used to look at a vectorscope to see you are not over saturating and the hue of yiur colors, to balance an image, remove a color cast, to look at a Parade scope can also help for that… I’m blind without my scopes when I’m color correcting. After hours of work, I trust my scopes more than my own eyes.
I would suggest take a look if FCP have these instruments, and check you are not using values under black (super black) because these can look good on yiur Mac monitor but will look overly contrasted on a video projector, check that you are not over white, in 8 bits this is usually 16 and 235, but of course, all this can be irrelevant depending on their projection system
I recently finished a super wide (6700×1080) video for United, was projected with a system that used a computer to feed 4 projectors that overlap, I delivered in ProRes HQ 444 which is RGB, so black levels are not that much of a concern that with a YUV codec like DNX-HD, but most systems automatically adjust for those difference when exporting expanding if you export to an RGB coded or leave it as is if you export to a YUV codec, same as for most devices that playback, have that in consideration too that’s why I trust what I see on my vecto and waveform scopes. If you do that you should be OK
when it comes to work on a budget, there is a pretty good option to get a cheap Blackmagic studio card and a Samsung HD tv for monitoring, that would give you a good idea of what people using projectors or TVs will see . I say Samsung because they are usually pretty good and close to the right temperature and have good adjustment controls .
Hope this helps
I do notice that the color seems to shift with MP4 encoding and the picture is usually darker on Windows. It doesn’t look like FCP has any gamma adjustments available in recent versions. There’s probably nothing going on with the gamma, but if there is, I don’t know that I can pin it down.
Yes, I have been using the scopes recently and it has been a big help in getting me closer. I’ve used the Luma chart to keep the IRE between 0-100, the vectorscope to adjust the midtone colors for skin tone and the RGB Parade to adjust the Shadow and Highlight colors on clips that happen to have a true black or white area in the image. I haven’t been able to find many more tips about the scopes, but I know there’s more to learn. I may sign up to view the Larry Jordan tutorials for a month.
After using the scopes on this screen I’ll probably take my workstation and hook it back up to my parent’s 42″ Samsung LCD, which clearly has higher capability, for a final once over.
Thanks for the help guys, please let me know if you have any further suggestions.
