Pages: 1 2 [3] 4 5   Go Down

Author Topic: Does your profiled monitor do this?  (Read 18918 times)

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Does your profiled monitor do this?
« Reply #40 on: February 23, 2018, 01:48:48 am »

That doesn't look all that strange. The steps are typical of the LUTs when truncated to 8 bits. The profile you posted earlier has 16 bit LUTs. Looks like your system is displaying 8 bits. The jaggies are due to the the 8 bit LUTs having to skip steps, or advance 2 steps depending on the slope.

Video LUTs can be either 8 bit or 16 bit depending on the card. Some cards will use more than 8 bits and dither the pixels to create in between colors.

However, the color shifts you are getting are way beyond that caused by 8 bit stepping anomalies.  These can create slightly funky gradients and is one of the reasons the best monitors use internal LUTs rather than the ones in a video card and leave the video card LUTs in a one-to-one state. But they don't produce the large color shifts you are seeing. Something else is going on. As others said, you need to break it down and find some way to test each device separately. Though, of course you would have to have access to the hardware.
Logged

degrub

  • Sr. Member
  • ****
  • Online Online
  • Posts: 1947
Re: Does your profiled monitor do this?
« Reply #41 on: February 23, 2018, 09:38:49 am »

The panel is likely 6 or 8 bit so it has to be downsampled somewhere in the chain. AS Doug pointed out, that doesn't really account for the colour shifts.
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Does your profiled monitor do this?
« Reply #42 on: February 23, 2018, 09:18:49 pm »

What about accounting for the aging of the display running it at 120cd/m2 for that many years.

BTW I loaded the Dell "Native" profile in Photoshop's CustomRGB and it reads a perfect 6500K/D65, but the OP says it looks better setting it to 5500K through the video card which I'm assuming fixes the bad highlight gradients. That sounds like a display set too bright that it loses linearity between each RGB channel.

Why not set brightness to 100 cd/m2 and profile it with a native WP and then the desired warmer 5500K using the video card just to see if it is on account of an aging display?

I can tell you through my 20 years calibrating my own displays from CRT's to Fluorescent to LED I've never had to resort to using vcgt LUT curves to achieve a specific brightness and color of white.

And I still got screen to print matches.
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Does your profiled monitor do this?
« Reply #43 on: February 23, 2018, 09:28:20 pm »

Well this doesn't make any sense. Why does the Dell 3007_native.icc profile have a vcgt tag with adjusted LUT curves. Native tells the software NOT to do anything with the video card unless adjusting brightness.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Does your profiled monitor do this?
« Reply #44 on: February 23, 2018, 09:33:04 pm »

BTW I loaded the Dell "Native" profile in Photoshop's CustomRGB and it reads a perfect 6500K/D65, but the OP says it looks better setting it to 5500K through the video card which I'm assuming fixes the bad highlight gradients.
Illustrates the idea of this loading profile in Photoshop to view CCT values (which are not the same as Standard Illuminants) as suggested, assumed to be prefect, is largely a waste of time. But that was outlined days ago.
Glad to see the OP is making progress by concentrating on finding a WP setting in the calibration product that produces a desired result. It's something most of us have to do anyway! The exception of course are those that read people make silly statements like: always calibrate your display to (fill in the blank). They think in their ignorance they are helping, they are not.  :'(
CCT 5150K on my NEC with my viewing booth, with a specific cd/m^2 produces a very, very close match for me between print and display. I'd NEVER suggest anyone use my settings unless they were using the identical display (PA272W), same colorimeter, same viewing booth, same ambient conditions as I do.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Does your profiled monitor do this?
« Reply #45 on: February 23, 2018, 09:40:33 pm »

So are you saying that the plotted xy numbers according to how they fall on the Planckian Locus don't coorelate to a hue of white. That they're just numbers and the software decides what those numbers look like?

If this is so then there's no way that you can get all calibrated displays to look the same which defeats the purpose of calibrating/profiling.

If goal post on what xy numbers look like on a display keeps being moved then your whole premise falls apart.

Andrew, you're not a color scientist and you don't seem to be offering any help to the OP.

At least I'm trying. You're just being A-HOLE as usual.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Does your profiled monitor do this?
« Reply #46 on: February 23, 2018, 09:40:57 pm »

Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Does your profiled monitor do this?
« Reply #47 on: February 23, 2018, 09:46:00 pm »

You have actual testing to do:
http://www.brucelindbloom.com/index.html?Vcgt.html

Quote from that link...

Quote
If a profile is loaded that has no vcgt tag, you can see if the video LUT is left alone or if some default curve is loaded instead.

That's all I needed to read which confirms my point. I already know about linear vcgt tags. How does the OP find out if something is being loaded in its place when there should be nothing?

That might help him if you can tell how to determine that.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Does your profiled monitor do this?
« Reply #48 on: February 23, 2018, 09:46:55 pm »

So are you saying that the plotted xy numbers according to how they fall on the Planckian Locus don't coorelate to a hue of white.
Seems you need to study what the Planckian Locus really is (or isn't).
Quote
Andrew, you're not a color scientist and you don't seem to be offering any help to the OP.
No, I'm not a color scientist, clearly you're not based on your text here. You assume (again) I'm not offering the OP help, despite his text and despite the fact you have no idea if I've helped the OP or not.
You are trying to create rabbit holes for him, meanwhile, while you were futzing around in Photoshop, coming up with flat earth ideas about color, he's moved on and improved his situation by avoiding much of the silly text here by one poster and elsewhere, from folks that are neither color scientists or experts on color, by trying differing calibration settings. So yeah, I'm fine being an A hole here attempting to aid him and teach you how this color stuff actually works.
What is your agenda in wasting his time sir?
« Last Edit: February 23, 2018, 09:53:21 pm by digitaldog »
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Does your profiled monitor do this?
« Reply #49 on: February 23, 2018, 09:48:08 pm »

That might help him if you can tell how to determine that.
Nope! He's on the right path (with my help I believe to some degree) if he hasn't found it fully already. You'd be hip to that if you actually read what he wrote (or what I write). What is your agenda in wasting his time sir?

Try reading what he wrote, one word at a time with my formatting as an aid:
The upside to the story is I experimented with calibrating the monitor to 5500K instead of the standard 6500K, and photos now look more colorful and pleasing to the eye.  So I'll set the 5500K ICC profile as default.   And move on with my life.


I think I found the problem.
I think not!
« Last Edit: February 23, 2018, 09:52:30 pm by digitaldog »
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Does your profiled monitor do this?
« Reply #50 on: February 23, 2018, 09:53:01 pm »

Oh, you already helped him in this thread?

I read through this thread and didn't find where you helped him.

And my agenda is to simply help the guy and learn if an aging display or Window's CM system could be causing issues.

You haven't helped at all. Could you just copy and paste the input that helped the OP? I couldn't find it.

I just can't believe someone as the OP is having this much trouble with a display and you still haven't told anyone why or how to fix it.
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Does your profiled monitor do this?
« Reply #51 on: February 23, 2018, 09:58:25 pm »

Oh, you already helped him in this thread?
I believe so but I'll let him speak for himself unlike you sir:
http://forum.luminous-landscape.com/index.php?topic=123300.msg1028763#msg1028763
See and attempt to understand suggestion #3 that I made. I'll even paste it for you and apply more formatting to focus your reading comprehension:
Try a vastly different profile setting for WP and if possible, set the software for a Native Gamma. Better, worse?
And now attempt again to read his reply:
The upside to the story is I experimented with calibrating the monitor to 5500K instead of the standard 6500K, and photos now look more colorful and pleasing to the eye.  So I'll set the 5500K ICC profile as default. And move on with my life.

Meanwhile, you're futzing around looking at CCT values in Photoshop and lumping them together with a standard Illumiant and worse, proposing to the poor OP, you've got some idea to help him. Nope.
Quote
I couldn't find it.
You're lost, that is why.
« Last Edit: February 23, 2018, 10:01:32 pm by digitaldog »
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Does your profiled monitor do this?
« Reply #52 on: February 23, 2018, 10:13:33 pm »

What I can't figure out is why this thread has gotten over 1000 views but the OP's YouTube video only gets 80. That means very few people reading bothered to view the video so not only is there very few people interested in seeing this as a problem but they're not even concerned about learning from this experience.

But I looked at the video again to note that the OP's video camera capture his display using AWB to show either 6500K he had it at is too cool or too warm but he states native is too yellow.

So Andrew, are you saying that it doesn't matter if it's 6500K or 5500K, they both can look identical in neutrality and not to trust what hue of white they represent on the Planckian Locus?
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Does your profiled monitor do this?
« Reply #53 on: February 23, 2018, 10:38:53 pm »

What I can't figure out...,
So Andrew, are you saying that it doesn't matter if it's 6500K or 5500K, they both can look identical in neutrality and not to trust what hue of white they represent on the Planckian Locus?
There’s a lot you can’t figure out by your own admission here more than once in this thread! It is common to see from amateurs (clearly unlike some here, your day job, what you do for a living is not professional color management consulting or photography!).
As to your question; start a new thread instead of hijacking this one. Probably study what the Planckin Locus actually is first; ask a silly questions then if you must.
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: Does your profiled monitor do this?
« Reply #54 on: February 23, 2018, 10:54:19 pm »

More clues:

My Dell3007 may be 10 years old, but it's been in storage for 8 of those years (climate controlled, dust free). So I doubt aging of the CCFL screen is to blame.

I don't believe the white balance settings on my video camera, which recorded the video I posted on YouTube, is relevant. The Youtube video was merely to show the alternating cyan/magenta seepage, not to point out the quality of the white point.

color seepage by texshooter, on Flickr

And yes, the Dell factory profile cures at least 50% of the cyan/magenta seepage problem.     
I've attached the Dell factory profile to this post. Perhaps it bears a clue to why it works so much better than the custom profiles I posted earlier.   I suspect something screwy is going on with how the LUTs are working (or not working) with my custom profiles. That doesn't mean it's X-Rite's fault.  There could be a problem with how the vsgt tags are loading, like Tim was saying. Or something could be wrong with how X-Rite's ADC is handled by the computer.  Unfortunately, I wouldn't know where to begin to test whether the problem is with the graphics card or the monitor. Did I mention this darn thing doesn't even have knobs to control contrast and color temperature? Jeez, even my first TV when I was a kid came with those.

I've borrowed a friend's monitor and will be doing some tests on it, as well. Wait, I thought I was moving on with my life?
« Last Edit: February 24, 2018, 06:29:52 am by texshooter »
Logged

Doug Gray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2197
Re: Does your profiled monitor do this?
« Reply #55 on: February 24, 2018, 12:25:31 am »

What I can't figure out is why this thread has gotten over 1000 views but the OP's YouTube video only gets 80. That means very few people reading bothered to view the video so not only is there very few people interested in seeing this as a problem but they're not even concerned about learning from this experience.

Simple really, tex has an interesting problem and people want to help or see if it's something they have encountered or learn from in case they encounter it.

The reason for the difference in view counts is that once people see the video, they pretty much understand the issue. But as more people comment and offer suggestions the overall thread view count keeps increasing. I've probably viewed the thread 30 times to see if there has been any resolution but I've only watched the video twice.
Logged

texshooter

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 575
Re: Does your profiled monitor do this?
« Reply #56 on: February 24, 2018, 12:54:33 am »

And let me add these conundrums.

1.  i1Studio recommends that I reset the monitor to its factory defaults before calibrating. But I can't because the OSD software doesn't work and because there are no physical buttons on the monitor (except for brightness and power on/off). Could the failure to reset the monitor to factory defaults cause any problems with the calibration?

2.  After I press the measure button in i1Studio, I get a one second warning message at the lower right corner of the screen, which reads
"The color scheme has been changed to Windows Basic.  A running program isn't compatible with certain visual elements of Windows."
Should I do something to fix this?

« Last Edit: February 24, 2018, 06:14:56 am by texshooter »
Logged

digitaldog

  • Sr. Member
  • ****
  • Online Online
  • Posts: 20630
  • Andrew Rodney
    • http://www.digitaldog.net/
Re: Does your profiled monitor do this?
« Reply #57 on: February 24, 2018, 09:28:54 am »

Avoid OSD (and the Spyder*) like the plague.
Did you try differing backlight display options in the software with the same settings otherwise?
I'd suggest borrowing or renting an i1Display Pro with it's software which has vastly larger options for setting the calibration of white (and other attributes). Plus i1Studio (if you're referring to the newer X-rite product) may be buggy if history is any indicator. With the i1 Display Pro, you'll have a very good colorimeter for the task and software with far more options in getting closer to your goal. IF with this new test, you still have issues, it aims towards the display, video card or something else, other than the calibration/color management part of this process being suspect.

http://iephotorentals.com/xbrite-display-pro

And please move ahead with your life by avoiding all this silly stuff about viewing your existing and undesired profiles, their numbers they supposedly show etc. Apparently tell testing shows otherwise, the profiles and/or calibration suck; that's all you need to know before moving on to the next test to resolve the issue. You got closer altering WP which shows some progress.


* http://forum.luminous-landscape.com/index.php?topic=103094.msg845726#msg845726
The higher the reported dE, the worse the unit preformed. So you'll see two Spyder's (newest models) were 9.9 and 7.2 which is pretty awful. The X-rite products were 1.4 and as low as 0.8!
« Last Edit: February 24, 2018, 09:34:29 am by digitaldog »
Logged
http://www.digitaldog.net/
Author "Color Management for Photographers".

TonyW

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 643
Re: Does your profiled monitor do this?
« Reply #58 on: February 24, 2018, 09:29:38 am »

And let me add these conundrums.

1.  i1Studio recommends that I reset the monitor to its factory defaults before calibrating. But I can't because the OSD software doesn't work and because there are no physical buttons on the monitor (except for brightness and power on/off). Could the failure to reset the monitor to factory defaults cause any problems with the calibration?
Usually I would expect some sort of hard reset option rather than relying on software alone.  It may be a combination of button presses to be held for a number of seconds to set back to factory state.  Maybe fire off an email to Dell customer support and ask how to reset without software or if not possible how to run in your version of Windows.

I'm going to take a guess that IF the monitor holds the value last input (8 years ago?) that you may have issues due to the graphics card limitations on controlling the look.  IMHO it is always best to get back to factory conditions if at all possible.

TBH, even though the monitor is new'ish to you it is 10 years old, a couple of lifetimes in technology and if you cannot get any drivers to allow access to the OSD then perhaps it may be time to consider an upgrade to a modern system and either retire the 3007 or repurpose for something else?

Quote
2.  After I press the measure button in i1Studio, I get a one second warning message at the lower right corner of the screen, which reads
"The color scheme has been changed to Windows Basic.  A running program isn't compatible with certain visual elements of Windows."
Should I do something to fix this?
No this warning generally is nothing to worry about as it just tells you that the Aero Windows theme (probably?) is not suited due to certain element (Opacity etc?) not compatible to a running program. Maybe i1Studio?  It should not affect the application running correctly.
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2436
Re: Does your profiled monitor do this?
« Reply #59 on: February 24, 2018, 05:38:39 pm »

More clues:

My Dell3007 may be 10 years old, but it's been in storage for 8 of those years (climate controlled, dust free). So I doubt aging of the CCFL screen is to blame.

I don't believe the white balance settings on my video camera, which recorded the video I posted on YouTube, is relevant. The Youtube video was merely to show the alternating cyan/magenta seepage, not to point out the quality of the white point.

color seepage by texshooter, on Flickr

And yes, the Dell factory profile cures at least 50% of the cyan/magenta seepage problem.     
I've attached the Dell factory profile to this post. Perhaps it bears a clue to why it works so much better than the custom profiles I posted earlier.   I suspect something screwy is going on with how the LUTs are working (or not working) with my custom profiles. That doesn't mean it's X-Rite's fault.  There could be a problem with how the vsgt tags are loading, like Tim was saying. Or something could be wrong with how X-Rite's ADC is handled by the computer.  Unfortunately, I wouldn't know where to begin to test whether the problem is with the graphics card or the monitor. Did I mention this darn thing doesn't even have knobs to control contrast and color temperature? Jeez, even my first TV when I was a kid came with those.

Got the same step wedge highlight anomalies in your posted images when I tweaked my CFL Dell 22in. Offset and Gain adjusts accessed through the display's factory set ROM OSD menu I had to go online to find out what button combos brought it up. I first recorded the Offset/Gain RGB numbers as is so if I screwed something up doing this I could go back.

And sure enough after applying the ROM tweaks and profiling with i1Match software/original i1Display colormeter, I got the same step wedge highlight color issues.

That's why I suggested you reduce brightness because just like with editing Raw on going too far with Expose To The Right, if there's any issues with RGB non-linearity caused by near saturation it will affect highlight color detail.

But also the native profile should not make your display look yellow and certainly shouldn't have that much of a correction in the blue channel LUT curve. I haven't seen one display at Best Buy or anywhere where displays off the shelf look yellow in their native white point.

And I just checked out that Dell factory profile you provided and it's just a simplified canned sRGB profile. It's useless because it doesn't have any  video card LUTs/vcgt curves.
« Last Edit: February 24, 2018, 05:46:59 pm by Tim Lookingbill »
Logged
Pages: 1 2 [3] 4 5   Go Up