Canon Pro9000 ll colour shift noticeable in greys

[Roy Sletcher]

However if it solves the problem this it will be OK for me. Will have some time tomorrow - Wed to experiment and see if I can get things better. In all honesty I should mention I am being rather picky because despite my complaining the profiles I am getting with the Spyder are better than the generic Paper profiles supplied by Ilford and Red River papers.

Roy, I wonder whether you have had time to check it out and if so what were the results?

I can't help feeling that this is really only a 'sticking plaster' remedy and does not address a fundamental problem with the profiles being too warm in the first place. It leaves the adjustments at the extreme end of their range.

Have you tried printing a softproof, mine look alright on the screen, but the printout is still totally different.

I don't think I am doing anything incorrectly, but have run out of ideas!

The printer in isolation produces very good results as described previously.

Regards Ian

Hi Ian,

Another try after the computer ate my reply yesterday. It has happened to me before on this site, and could be anything from my ISP, or my software not necessarily Nifty-Stuff.

Unfortunately I have been somewhat remiss in not following through to further test profiles and my methodology. It is my intent to do so but somehow time gets away from me. Following are some random thoughts on my situation. Yesterday's version was more logical and coherent

You are quite right that my current solution of editing the Spyder Profile with the Spyder profiling software is more a "band aid fix" rather than a permanent solution. - still have to visit methodology test and consistency evaluation. So may variables to account for with paper, ink, computer, methods, evaluation, and more.

For the record my softproof in Photoshop looks generally OK with Neutral black ramp from 0-255, greytones ok, and open shadows and three quarter tones in the shadows. What prints is slightly warm toned (to much red or magenta, sometimes yellow I think). Also an occasional a slight yellow tint in the centre of the grey scale ramp. Close, but no cigar, and definitely should be better.

I initially started printing the 729 patch target on 4 sheets hoping this would give a more accurate profile. It was very tedious. For the past couple of years have been using the 225 patch target on 2 sheets. The results appear to be almost identical with 729 patches. The two sheets version being much quicker to process.

I am seriously considering changing to the color munki, although the difference I am seeing does not seem to warrant the about $500 cost. Could probably sell my Spyder to defray about half the cost. At least the CM is quick and easy compared to the Spyder. Still meditating on it, especially the Spyder does not give any allowance for loyalty when upgrading, and I have been using all version since version 1 in 2002.

I have test prints going back to 2009 with my Spyder, and have to say over that time the Spyder profiles have improved. This is probably due to me getting better at using it, and upgrades in the software. Earlier software version had lots of misreads of the patches which led to consistency issues. This seems to be much better in version 4.2.3

I am still committed to doing an overall review, hopefully later this week. Unfortunately my reply yesterday was more eloquent, can't seem to get the flow going today. Lest's see if the computer gobbles this one.

Roy

 

[Roy Sletcher]

Hi Rod

Thanks for the photos, I see you have at least two pairs of sandals.

I hope the load on your server does not cause it to crash after this gets around.

I also noted the 7D, which is what I am using, mainly for wildlife. Canon all the way!

Regards Ian

Rod/Ian

Well know we know the property market is a HOT item in OZ.

Rather spoils the illusion. I had Rod pegged as the handsome "Crocodile Dundee" type. Not another geek photographer. Also Crocodile Dundee had this huge knife, which I am led to believe Aussies need when eating steak off the barbie (Or something like that)

Roy

 

[rodbam]

That shot was taken a few years ago for an internet challenge, the missus was out so I set this shot up with my old G2 which took a few tries to get it right & when me missus got home she saw a black burnt patch on the paver's near the chair. She then taught me how to get burn marks off the paver's
All I can do Roy is to try & make my camera & lenses look bigger than anyone else's

 

[Roy Sletcher]

That shot was taken a few years ago for an internet challenge, the missus was out so I set this shot up with my old G2 which took a few tries to get it right & when me missus got home she saw a black burnt patch on the paver's near the chair. She then taught me how to get burn marks off the paver's
All I can do Roy is to try & make my camera & lenses look bigger than anyone else's

Well Rod, as the younguns say, "Size matters".

Personally I have always believed old and cunning trumps youth and enthusiasm.

Roy
Who has a big Canon.

 

[Emulator]

Hi Roy and Rod

I am not going to try to compete. At 77, it is the story of the old ram and the young ram.

I thought you might say you had a CO2 extinguisher tucked around the corner Rod.

I have been doing more research Roy. FYI I have found 'ICC Profile Inspector', a free software download from the ICC site, which lets you view and I think adjust ICM profile inner details. Still have not properly explored it or understand it. However, looking at a profile it shows details that are set in the profile that I did not think would be included.

A second point, the colour gamut of the Evaluation image.tiff is 'ProPhoto RGB' try coverting it to sRGB in Photoshop. It makes a difference to the prints. Printing the Spyder proof matrix image produces nice greys in my latest printout. I have been experimenting with plain paper (it's cheaper) and have produced a plain paper profile.

I think we are making progress albeit slowly.

Regards Ian

 

[Emulator]

There is little doubt that the high quality resin coated micro porous paper has a marked effect on the measurement of the SpyderPrint colour patches.

Creating a profile with plain paper illustrated the difference. The coated paper has a blue filter effect and the resulting measured colour patches (in split mode) are very noticeably different from the 'pure' colours, causing the generated profile to make unnecessarily extreme adjustments.

I have tried biasing the resulting profile measurements away from 'warm' by using a straw coloured backing sheet behind the target print being measured. This had very little effect on the measurements.

For the next attempt I placed a .002in or .05mm thick, clear, (apparently) colourless film sheet between the resin coated paper target print and the spectro head. This made the spectro head slide more easily and produced a startlingly different set of colour measurments, very much closer to the 'pure' colour patches in split mode.

Using the resulting profile produces noticeably better individual colour gradation on the Evaluation target print.

I think this approach shows much promise in overcoming the patch measurement bias of 'blue rinse' papers. The film thickness will have an affect on the optical function of the spectro, this may be an improvement or a negative effect. The 'colour effect' of the apparently colourless film, will also contribute.

More work needs to be done, but the results are very promising.

 

[ThrillaMozilla]

It sounds like the spectrometer gadget is reading the surface reflection. That's white light. When you add a film on top of the print, you get a double reflection. The spectrometer (or whatever they call it) sounds like a disaster.

 

[stratman]

Are you able to set a white point when profiling?

 

[Emulator]

Stratman

Yes, the calibration is set against the Spyder white tile Lab=91.27,-1.01,-0.81 and the subsequent white target measurements directly from the resin paper show a distinct blue tinge to the white Lab=99.87,-1.11,-97.95. With the added film the white target displays a very light grey-white Lab=96.70,0.64,-4.75.

The light from the spectro is multicoloured, i.e. multi source, not white.

Regards
Ian

 

[stratman]

I take it you cannot set a white balance from the scanned resin paper (not the white tile) to compensate for the bluish tinge?

What paper is this resin coated paper? I scanned some of the preceding posts but not all of them, so I must have missed the name.