Highlighted
Moderator
Moderator

Version 3 of the U2410.ICM file is out

Go here

If using Firefox, remember to load it in Firefox- Add-ons- Color Management- Preference- All Images. Once loaded, open Photoshop and direct it to use that ICM.

Windows 7/Vista
* Open Color Management by clicking the Start button, and then clicking Control Panel
* In the search box, type color management, and then click Color Management
* Click the All Profiles tab, and then click Add
* Locate and select the new color profile, and then click Add
* Click Close

* Open Control Panel- Hardware and Sound- Display- Change Display Settings (left hand pane)
* Select Advanced settings- Color Management tab- Color Management
* Check the "Use my settings for this device" box
* Click on the Add button
* Scroll down until you find DELL U2410 Color Profile
* Double click it
* Click on Set as Default Profile and close the Color Management screen

Windows XP
* Right click the Desktop
* Click Properties
* Click Settings
* Click Advanced
* Click Color Management
* Click Add
* Locate and select the new color profile, and then click Add
* Click Apply- OK


0 Kudos
23 Replies
Gecks
2 Bronze

Re: Version 2 of the U2410.ICM file is out

It's looking good Chris!  The Windows Picture Viewer issue appears to be resolved.  No issues in Photoshop either.

Will do some more testing, but it's looking good at this point.

0 Kudos
P.A.K
2 Iron

Re: Version 2 of the U2410.ICM file is out

Haven't had time to check much, but this profile is very much improved over the first beta ICM.

However I did notice that it creates some visible banding, which can be seen on images like the Lagom gradient test ( http://www.lagom.nl/lcd-test/img/gradient-h.png ) with applications which use the profile, and which is NOT present when the ICM file isn't used. I'm wondering if this new ICM could be fine tuned to remove/minimize the issue more?

Here is an example of what is seen. Please note that you must have colour management disabled to view this since otherwise the ICM profile may introduce banding on both the top and bottom of the image. Also, care should be taken over scaling the image (magnifying etc) since some applications can introduce banding on this type of image.. For example IE8 might make the bottom of this image band, whereas (viewed properly) banding should only be apparent on the top of the image. I suggest using something like Windows Photoviewer or Photoshop to view it. The windows paint program will also suffice, and you can use it to magnify things and make the issue more clear.

The full image is available here http://img138.imageshack.us/img138/5283/u2410betaicmv2.png

Smaller preview

 

This banding isn't only visible in applications like Windows Photoviewer and Firefox, it also appears to be visible in Photoshop too. As said, if possible, it would be nice if this improved ICM could be fine tuned to reduce or even remove this visible banding. But it's almost there, and seems to be a big improvement over the last version so far.

 

UPDATE - Please see my updated notes below. I get the feeling that this is related to the trade off you do when converting images to an sRGB colour space when using the ICM profile. Hopefully there are still some very minor tweaks which could be done even if this is the case 🙂

0 Kudos
jsicard
1 Copper

Re: Version 2 of the U2410.ICM file is out

Hello - I've been tracking this thread since the beginning - trying to avoid returning this monitor. I'm considering purchasing the Spyder 3 calibration software, and creating a spepcific icm with it. Would that fix the issues we're facing here?

Kinda new at this, so appologize if it's a futile question...

John

0 Kudos
David Cheney
1 Copper

Re: Version 2 of the U2410.ICM file is out

Yep, looking pretty good from what I'm seeing... but at this point I'm following the lead of P.A.K - and perhaps Dell is too 🙂

Dave

0 Kudos
dfwald
1 Copper

Re: Version 2 of the U2410.ICM file is out

A nice improvement finally making my Photoshop Elements 7 colors to look the same as Windows Photo Viewer but I find that now the colors are a bit undersaturated.  The default monitor looked oversaturated and I feel that now you've gone a bit too far in the opposite direction.  I have colors set to vivid in my DSLR camera settings yet all my photos now look a tiny bit less saturated in color than what appears to my eyes.  With my vivid setting, it should look a little bit more saturated than the colors I see with my eyes.   My pictures look like they lack contrast now and don't pop at all anymore.  In your next beta version, try to fix this undersaturation problem.   Are other people here seeing the same thing?

0 Kudos
P.A.K
2 Iron

Re: Version 2 of the U2410.ICM file is out



Yep, looking pretty good from what I'm seeing... but at this point I'm following the lead of P.A.K - and perhaps Dell is too 🙂



Hmm, let's hope not emoticon.Smile.title

Anyway, whilst listening to the radio, I had a look at the actual Red/Green/Blue values on the Lagom test image last night. In order to limit the colour gamut of the display to sRGB some values must be changed when in wide gamut mode. So what I documented above, in my first post, might well be related to those compromises. But hopefully there still might be a way to improve things so that everyone can use the ICM file without the need to add "as long as you don't mind a small amount of banding it introduces".

The first thing I noticed when using a gradient test image like the Lagom image, which is designed to show the full range of values from black (0) to white (255). is that the ICM causes the darkest RGB values to go from 0 to 3,7,10, then 12 under Windows Photoviewer. Without the ICM it's the correct 0,1,2,3,4 etc. This means the ICM is already 8 steps brighter than without the ICM at the RGB value 4,4,4 (Using the ICM it'd be shown as 12,12,12 on your screen). It also means 8 black steps are completely ignored by the ICM between 0 and 12.

However, before blaming too much on the ICM, when you look at Firefox it actually renders the darkest end differently - as 0,6,8,10,12. It's doing the exact same limitation, but some of the numbers are different. So although some of this is down to the ICM, it's also telling you that there is also some difference between the way different applications are using the information to render things, because Firefox renders the colours slightly differently than Windows Photoviewer does.

If you're looking to limit the range of colours it's also usually better to do most of it in the lowest dark values. Although that will reduce the black level detail, and also the perceived contrast to some degree, it's still usually a much better compromise than the alternative - which is to make those changes higher up in the colour values where you'll notice it much more. If that's the case it would explain why those black values jump at the start when using the ICM. There's a further suggestion that things are related to limiting the range because some of the values are "doubled up" on the Lagom gradient test image. EG Under Windows Photo Viewer the RGB values 22 - 28 - 34 - 42 - 50 - 61 - 73 - 88 and 159 are doubled. That is to say, instead of going 21,22,23 it goes 21,22,22,23. So it skips some values. This is one of the things that shows as banding in the image. Also keep in mind those numbers will probably change if a different application is used.

 

How much of this is actually a necessary trade off etc, and how much can be avoided, is what I really don't know. I also don't know how all this relates to the internal hardware LUT, since I'm not sure it's even being used.. The internal hardware LUT on the U2410 can supposedly process things internally as 12-bit colour values. That is to say Red, Green and Blue can have 4096 values, and are internally processed with this greater accuracy before being sent to the panel itself which is a native 8 bit panel - meaning red, green and blue can have 256 values each. (And yes, technically the panel can apparently do 10 bit with dithering applied, but things are already really overly-confusing as it is, huh? 😉 ). But the bottom line is it'd be nice if there's still a way to slightly improve or even remove this banding behaviour..


Other than the jump in black values at the start, and the double values, the other differences under Windows Photoviewer (just to document them) were
At 111 it goes 111,111,111 - 111,110,111 - 112,112,112.  (In the middle, between these 2 values, you'd definitely want the green to be 112 or even 111.. how much is down to the ICM I'm not sure)
Continues correctly from 112-172.
At value 172 it goes 172, 172, 172 - 173,173,174 -  174,174,175.
Continues correctly from 175-181.
At value 181 it goes 181,181,181 - 183,182,183 - 184,183,184 - 185,184,185
Continues correctly from 186-237.
At value 237 it goes 237,237,237 - 239,238,239 - 239,239,239 - 241,240,241 - The 239,239,239 value would be better set to 240,239,240. Again, how much is related to the ICM, and how much is the application I'm really not sure.
It then continues correctly from 242 to 255.

Even if the above information can't change much at least it serves as a rough guide for where you're likely to be compromising on image quality when it comes to using the ICM file..

0 Kudos
Gecks
2 Bronze

Re: Version 2 of the U2410.ICM file is out

P.A.K. - I've followed most of your posts and while I agree with your observations, I am wondering if a *perfect* ICM is even possible, especially with the two RGB modes.   There are likely too many variables.  Color management is in its infancy at this point, with Windows 7 (scRGB) not really doing it properly, IE does not, Firefox has its version and OS X is a different scenario as well (its recent gamma point adjustment notwithstanding).

So given that this monitor is capable of a very wide gamut, how does it accurately and without visible distortion, limit that gamut to a smaller color space like RGB?  Throw in the way that other third party software interprets color management and you would have a headache indeed, IMO.

I too am hoping that Dell can produce a profile that works perfectly, although I am wondering if it is even possible.

It will be nice when software and OS's catch up on this front.  Wide gamut monitors are still far too ahead of their time, IMO. 

For a wide gamut, the U2410's two RGB modes are quite impressive.  This would be an absolutely perfect monitor for me if the anti-glare was turned down just a tad, but that's a whole other story.

 

0 Kudos
Moderator
Moderator

Re: Version 2 of the U2410.ICM file is out

I too am hoping that Dell can produce a profile that works perfectly, although I am wondering if it is even possible.
* At some point, engineering will say they have done enough work on the ICM.


0 Kudos
P.A.K
2 Iron

Re: Version 2 of the U2410.ICM file is out

I am wondering if a *perfect* ICM is even possible.

Yep, true, which is why I was sort of pondering that in the comments of the longer posts above. The least we can do is point out the current limitations and see how well they might be minimized (if at all) emoticon.Smile.title

So given that this monitor is capable of a very wide gamut, how does it accurately and without visible distortion, limit that gamut to a smaller color space like RGB?



You do what the U2410 does and include a 12-bit hardware Look Up Table inside the screen which can calculate the compromises to an accuracy of 4096 points (more than 68 billion possible colours), then render it to the screen. Although the screen still renders with lower accuracy it probably makes the U2410 superior to most equivalent wide gamut monitors (like HP's LP2475W) around the same price range in this regard. If you switch to sRGB mode you'll see that it's operating in a very decent emulated sRGB colour space, yet it does this without noticeable dithering or banding - or at least it does if you use firmware revision A01 emoticon.Wink.title

Of course the reason we want the ICM profile is, in theory, it can be left to operate at the full gamut and colour managed applications will automatically render to sRGB or Adobe RGB etc. No need to switch modes or even think about it any more. However I honestly don't know if an ICM profile can be made to produce sRGB results as good as the U2410's sRGB mode whilst staying within the U2410's wide gamut settings. I suspect there might be a limitation there, but that's (hopefully) what we're about to find out.. 🙂

If it turns out it's not possible then people can still switch modes manually. Not as good as having an ICM file which auto-magically solves things (if the app is colour managed!), but at least it's possible on the U2410.. emoticon.Smile.title


 




0 Kudos