green colour cast after callibration - any ideas?

[chris26]

after callibrating over two days and running tests, photos with excessive blue have a green colour cast. Ok ! Anyone got any ideas before I send tickets to datacolour?

[blinded]

You have either a bad colorimeter or your screen was REALLY bad before it was calibrated and now your brain thinks it's wrong.

[chris26]

Hi, screen is 9 months old, colours were good but slightly off in some wavelengths before I bought calibration. As for the spyder itself, well who knows?

[René Damkot]

Some more info on hard and software might help.

Do I understand that the blues take on a green cast?

[chris26]

Hallo Rene, how's it going?

Thanks, I'm using CS3, after slight corrections in Raw. on apple iMAC. On both photos there was a need for a colour fill adjustment layer (bluish of course) applied to the non-masked areas of swans and on another photo a few fishing boats. I checked the colour sampler on a 5x5 average and noticed that RED is lagging behind, giving rise to high CYAN values on the CMYK info palette. But on the other photo there are higher green values. Now I could actually correct these casts through point manipulation of curves adjustment layers. BUT the point is I did not have the casts before profiling. OR DID I? I notice that you can not really see any cast on the monitor. BUT If I look hard enough I see why the printer colour cast is there (I think).

Does any of this waffle help?

In all regards I have sent a ticket to Datacolour anyway. But yeh, if others have experience similar to this then that is always a good place to start.

thankyou, have a good day

Chris

[René Damkot]

Still not too sure what you are saying: Some photo's have a cast when printed?

What is your workflow? Correct icc profile for printer / paper / ink? OEM inks?

Let us know what Datacolour says.

[chris26]

Do not quite understand why you need to know workflow. Yes I am using correct colour managed methods and never had this problem EVER EVER before I had profiled with the spyder - Damn it I knew I should have bought the Gretag Macbeth!

Chris. I will post when I get datacolours first response. This will be a generic unsatisfactory answer. Then I will post on second response. Tis will be an attempt to be specific but probably will not satisfy. (I learned all this from oter user's experiences with datacolour).. BUT, am willing to apologise if i am pleasantly surprised.

Best regards

Chris.

[chris26]

Went back to dng file and ran some colour sampler tests on original, modified dng and PS file. Identified a cyan colour cast on original DNG. This due to incorrect white balance - I cheated with an underexposed image in bright sunlight reflecting off of the sea against some boats. the RGB readings showed RED lagging way behind in the neutral areas such as dark water and pale shaded white areas. This was corrected in curves in PS by setting precise point targets in the red channel and making it higher, then lowering precise point target in the blue sky. image is reasonable but not worth printing, it is much better though. HOWEVER, lesson learned - i should have been less careless in upping the exposure in RAW and more precise by first setting colour samplers in the original raw image to double check that neutral areas are indeed neutral numerically and not trusting my eyes.

STILL, I have one problem, the printer printed that colour cast quite clearly, yet the screen does not show it very much, you have to really look hard to see anything at all.

Chris

[René Damkot]

Did the *softproof* show the color cast?
Was the area out of gamut by chance?

[blinded]

Where did you buy the spyder? You could maybe copy the current profile, and trade the spyder in for a replacement, calibrate and see if it's better or worse by comparing.

[chris26]

First, blinded, I ordered it via England, sending it back is hassle at this moment but thanks for the tip i had hought about that but first need to assertain that I am not at fault.

Rene, thanks, the soft proof did not yield anything except a pleasing blue that was tinsy winsy green, but that was how I had purposed it. As for Gamut, though the file is in 16 bit, on this occaision no greyed out of gamut coverings and I am also careful to watch the out of gamut triangles on the colour picker menu when choosing an overlay or gradient colour.

I also have just done a print on gloss paper, tha funny thing here (and this does not happen very often) the colour cast was lessened drastically. My immediate comment to myself at that moment was - bloody useless canon profiling for their own paper. I really want to invest in an epson 2400 for the sake of paper profiles that are more accurate and of course epson's quality. Unfortunately finances. Anyway, The change in paper for this photo was significant in reducing that green colour cast, although it was in fact a cyan colour cast caused by lack of red in the RGB 11x11 averages on colour sampling the neutral areas.

Soft proof revealed nothing new. No out of gamut. my monitor was checked 30 minutes ago. Everything got the green tick at the bottom.

OH yes, also printed out another phot that had a green cast - my swan on a lake, same results when printed out on gloss, colour cast diminished. ALSO MOST NOTABLY his white was tainted with that cast but on the gloss the white was 200% perfect and perfectly in line with the screen.

Thanks Rene for your time

Kind regards

Chris

[chris26]

Rene, ik heb het net opgemerkt, jij bent een nederlander, ik ben hoewel engels.

[René Damkot]

Hahaha. Ja, nederlands ;)
Seems to be mainly a paper profiling issue then? Might be worth it to have a custom profile made for that one paper?

[chris26]

Probably, you're right. Any recommendations about how I go about doing this. Have no clue here.

Chris

[Damo77]

As for Gamut, though the file is in 16 bit, on this occaision no greyed out of gamut coverings and I am also careful to watch the out of gamut triangles on the colour picker menu when choosing an overlay or gradient colour.

Just as an aside, bit depth has nothing to do with gamut, as far as I'm aware.

[René Damkot]

Probably, you're right. Any recommendations about how I go about doing this. Have no clue here.
Some companies can make a profile for you. IIRC there is also someone on POTN who can.
Never used it before, since I'm fairly happy with the profiles I use.
I might have an address in the Netherlands, but I suppose that's not too much help...
Just as an aside, bit depth has nothing to do with gamut, as far as I'm aware.
Agree.

[chris26]

Not all the tonal detail of a 16 bit image is recognised by an 8 bit image and I know that you know this. So I am more likely to have out of gamut warnings when i am editing in 16 bit because the likelyhood that they will not be represented in a printer is greater. Hence my statement was an off the fly comment to indicate my luck and good fortune, that although I was editing in 16 bit, I ACTUALLY was well within the colour range of my printer NOT my Paper of course!

And! While we are on the defense, your comment a few days ago about QUOTE: "listen to Rene because he knows what he is talking about" UNQUOTE - I resented. Rene is certainly knowlegdable and has tons of information that shows how much time he has put into his work, BUT knowledgable people can have technical and scientific prejudices. Native white point is NOT recommended by the professionals in the colour management workflow of photographers. I learned this and it was also confirmed (by chance) by a tutorial video from Lynda.com, just as a side note!

Regards

Chris.

[chris26]

Thanks Rene. Is the address within the Delft _ Den Haag regio? Then I am interested since we go back to Holland regularly.

[René Damkot]

Not all the tonal detail of a 16 bit image is recognised by an 8 bit image and I know that you know this. So I am more likely to have out of gamut warnings when i am editing in 16 bit because the likelyhood that they will not be represented in a printer is greater.
There might be a bit of miscommunication going on here :)

Bit depth has nothing whatsoever to do with Gamut, only with "gradation". Let's assume you are using a large color space, like ProPhotRGB. A certain color will be just as out of gamut in 8bpc as it is in 16bpc, but the 8bpc image is far more likely to have posterization, because there are not enough "steps" to make fluent transitions.
The "outer limits" of the "color range" will be the same however.

Thanks Rene. Is the address within the Delft _ Den Haag regio? Then I am interested since we go back to Holland regularly.

IIRC these guys can make one: Click (http://www.arca-nl.nl/default.aspx). However cannot find a reference on their (new, horrible) site.

You might also try (Amsterdam) Eyes On Media (http://www.eyesonmedia.nl/print-en-finishing/) or Fimex (http://www.fimex.nl/fimex/mcsmambo.p?M5NextUrl=RAINF&M5Arg=2) / Camex (http://www.camex.nl/camex/mcsmambo.p). (first two are Pro shops, not sure if you can use them). Just send them an email.

I'd google for "custom printer profile" to find more.
Here's the first hit in Google: Click (http://www.drycreekphoto.com/custom/customprofiles.htm)

[chris26]

Thankyou Rene, I will try these first. Ok you are right and i was mistaken. The tonal ranges of a certain colour might be missed out. I have actually seen that on a few photos myself.

Regards
Chris.

[Damo77]

While we are on the defense, your comment a few days ago about QUOTE: "listen to Rene because he knows what he is talking about" UNQUOTE - I resented. Rene is certainly knowlegdable and has tons of information that shows how much time he has put into his work, BUT knowledgable people can have technical and scientific prejudices. Native white point is NOT recommended by the professionals in the colour management workflow of photographers. I learned this and it was also confirmed (by chance) by a tutorial video from Lynda.com, just as a side note!

Hmmm ... perhaps you don't take René seriously because he isn't a published author (that I know of ...).

From "Real World Color Management (http://www.colorremedies.com/realworldcolor/)" by Bruce Fraser, Chris Murphy and Fred Bunting - widely regarded as a bible of colour management:

The one thing to avoid is making major corrections in the video LUT, because doing so reduces the number of discrete levels your display can reproduce - the bigger the corrections, the fewer levels.
Adjustments you make to the monitor itself, on the other hand, are "free" - there's no loss - so a major consideration in choosing calibration settings is the means available to achieve them.
...
For LCD displays, the only thing you can typically adjust on the monitor itself is the brightness of the backlight. Changes to the color temperature are carried out in the videoLUT, so we recommend using the native white point for LCDs whenever the software allows us to do so.

[chris26]

DAMO 77 QUOTE Hmmm ... perhaps you don't take René seriously because he isn't a published author (that I know of ...).UNQUOTE

Please stop making assumptions, it's simply not cricket old chap

DAMO 77 - From "Real World Color Management" by Bruce Fraser, Chris Murphy and Fred Bunting - widely regarded as a bible of colour management:


I accept this but am interested in details, why? where? what? Bruce Fraser is recommended by Katrin Eisman and Sean Duggan also and I follow there tutorials and advice. But I have studied many colour management modules from about different people and have NEVER BEEN TOLD TO DO THIS. Quite the opposite as I have indicated earlier.
IF ONE PROFESSIONAL or if I include Rene TWO proffessionals say do this and FOUR OTHER EQUALLY QULIFIED PROFESSIONALS say do that, what is a student suppose to decide?

Regards Chris.

[René Damkot]

Guys, lets keep on topic ;)

Post 14, point 3 is somewhat related; Click (http://openphotographyforums.com/forums/showthread.php?t=3822&page=2).

From http://www.imagescience.com.au/ColourControl/colourProducts/gmUsingEyeOneDisplay.html:
Try 'Native White Point' first, as most LCDs are roughly 6200 – 6800K anyway, which is fine, and with basic LCDs you generally play a price for forcing the white point away from the native whitepoint in terms of smoothness across the gamut.
If after calibration is complete the measured native whitepoint is outside of 6000K to 7000K, come back to here and re-do the profile specifiying 6500K as the whitepoint.

The short answer is: It depends on your monitor.

If "Native" isn't too far off the whitepoint you want, I'd not mess with it.
If you have a monitor with an 8bit LUT, I'd not mess with it.

I am curious to why you were specificly told *not* to use "native" however.

[Damo77]

Guys, lets keep on topic ;)

Sorry mate. How's that book coming along? :lol:

I am curious to why you were specificly told *not* to use "native" however.

Me too.

[chris26]

Chris Murphy, lecturer and colour management tutor on Lynda.com. demonstrates with the gretag macbeth eye one, he says that those with the corrct ambient light conditions select appropriate Kelvin around 5000 IF of course the light conditions within which you work are set with special light bulbs and so forth and so forth. All others should choose 6500k except for Apple cinema users who should choose Native white point. I work in the evenings under certain light that requires a 5800k, during the day I use 6500k unless it is overcast because I work next to a window - unavoidable.

If you guys really can show me otherwise, that this lecturer is perhaps wrong in some way to suggest this then I am really open to learn ok!

Bets regards, chris.

[René Damkot]

Curious to the screens being used: Most screens don't take D50 particularly well...
I think this can be filed under "depends on monitor"...

Let's conclude: Use Native, unless you have a good reason not to, and your screen is suitable ;)

I might look into this a bit deeper in a while, when I have the time.

[chris26]

I have been busy doing that also. In the meantime I have submitted a ticket to Datacolour asking for their reasoning.