Reading mode: Light Dark

Photoshop CS Calibration. Who's done it?

Go to last post
If this is the case, it is, for all practical purposes, worthless.

With Magne's profiles, you get one standard profile. You might have another for a different level of saturation, but that's it.

I did calibrate ACR using an AWB shot of the Color Chart. noon lighting and it seems to work okay except, of course for tungsten and flash lighting.
Perhaps Tom Fors could weigh in here and clarify this issue.

As I mentioned, he says nothing about multiple exposures, but OTOH, other methodologies do indicate the need for separate files for different lighting.
Tom
Stephen Cameron said:
("Like you, I am quite happy with calibrated ACR. As long as you
have a well exposed Gretag-MacBeth chart, ACR takes into
consideration the varying color temps, referencing it to the shot
you use for the script--as far as I can tell. (If it didn't, you'd
be faced with making a set of conversion numbers for every
conceivable color temp!!) If I'm wrong about that, I'll stand
corrected, but it seems that way to me")

Tom,
More than one Macbeth Gretag Color Chart shot is needed as the
reason for the profile is to balance ACR for different light
sources as well as the device, (camera, lens combo). Direct sun,
Open shade, Cloudy, and Tungsten are the four main ones and
Fluorescent if you want to mess with that bag of worms.

So far for me I have seen large variations in the same light
source, Direct Sun, with time of day and angle of exposure relative
to the sun. And it seems the exposure histogram should match the
future shot you intend to correct. And of course each time you run
it on the same chart it comes out a little different. One thing
I've noticed though is that closing the history palette while
running it seems to really speed things up in addition to
specifying the lowest resolution import from ACR to Photoshop
during the run.
40mins as compared to 120mins.
Click to expand...
 
Tom Melanson said:
scottjj said:
Needing to run a 90 minute long script for every event
and lighting change that I shoot in makes me suspicious.
Click to expand...
I believe this is a mis-characterization as to what the script does.
The Thomas Fors script says nothing about running a script for
"every event and lighting change". One, "evenly lit, well exposed"
file of the Gratag-MacBeth Color Checker is all that is necessary.
Click to expand...
At basketball Saturaday night I shot the GM Color Checker in two locations. The first was laying on the court just outside of the paint. That is as about a bright a location as can be found on the court since the chart was staring right into the lights. I shot no-flash, bare flash and flash with Stofen. Then I shot the Color Checker clipped on a back wall about 10 feet from the end of court. There the chart was lit by the same lights but with a bit of the light reflected off the wood floor. I shot the same three versions.

In theory all six images of the same chart shot by the same camera under the same lights should be the same except for light intensities, WB and maybe a touch of the gym lighting color shift. The gym lighting is warmer than the flash and the Stofen diminishes the flash output and gives a slightly warm look to images.

Three scripts were run, two of them from the court images. The three charts are all different from each other but images processed using their numbers are fairly close. The differences between them tend to be about the same intensity as each one is different from Adobe's default setting for the 1Dm2. The good news is that one of them does has a more realistic color rendition than the ACR default. So, since we're using the script method to achieve color accuracy, which one of the four profiles is "correct"?

I'm not slamming either the theory of profiling or the hard work put into writing the script. I'm just trying judge whether all the fuss is really resulting in a truly accurate tool to use. For example on those shots where the flash failed to fire should the non-flash script data be used? Isn't flash color compared to no flash color really a difference of WB? Which one of my accurate scripts is the most accurate script?
 
I don't think anything is perfect when it comes to "do-it-yourself" profiling. The most accurate profiles come from Magne Nilsen. It's fair to say he is one of the world's experts in the field.

Given the relative complexity of what one is trying to accomplish and the tools one is given, it doesn't surprise me that some settings are different with Tom Fors script. However, in my experience--as in yours--these are distinctions without differences.

One thing I have come to a conclusion about after a lot of checking is that ACR, regardless of other issues, is more accurate in reproducing blues than is C-1. And this is not a WB issue. Dark blues always go toward a slight, but noticeable, deep purplish hue with C-1 and are dead on with ACR, calibrated or not.

With a critical shot, I'll aways convert the RAW file with both C-1 and ACR and use whichever is closest to "real life".

Like many others, I eagerly await a Magne Nilsen profile for C-1 and the 1DsMkII.
Tom
scottjj said:
Tom Melanson said:
scottjj said:
Needing to run a 90 minute long script for every event
and lighting change that I shoot in makes me suspicious.
Click to expand...
I believe this is a mis-characterization as to what the script does.
The Thomas Fors script says nothing about running a script for
"every event and lighting change". One, "evenly lit, well exposed"
file of the Gratag-MacBeth Color Checker is all that is necessary.
Click to expand...
At basketball Saturaday night I shot the GM Color Checker in two
locations. The first was laying on the court just outside of the
paint. That is as about a bright a location as can be found on the
court since the chart was staring right into the lights. I shot
no-flash, bare flash and flash with Stofen. Then I shot the Color
Checker clipped on a back wall about 10 feet from the end of court.
There the chart was lit by the same lights but with a bit of the
light reflected off the wood floor. I shot the same three versions.

In theory all six images of the same chart shot by the same camera
under the same lights should be the same except for light
intensities, WB and maybe a touch of the gym lighting color shift.
The gym lighting is warmer than the flash and the Stofen diminishes
the flash output and gives a slightly warm look to images.

Three scripts were run, two of them from the court images. The
three charts are all different from each other but images processed
using their numbers are fairly close. The differences between them
tend to be about the same intensity as each one is different from
Adobe's default setting for the 1Dm2. The good news is that one of
them does has a more realistic color rendition than the ACR
default. So, since we're using the script method to achieve color
accuracy, which one of the four profiles is "correct"?

I'm not slamming either the theory of profiling or the hard work
put into writing the script. I'm just trying judge whether all the
fuss is really resulting in a truly accurate tool to use. For
example on those shots where the flash failed to fire should the
non-flash script data be used? Isn't flash color compared to no
flash color really a difference of WB? Which one of my accurate
scripts is the most accurate script?
Click to expand...
 
The GretagMacbeth color chart comes in a beautifully printed box with 2 images of the chart. In both images there are 2 squares that are meant to be blue that are printed purple.

If the GretagMacbeth chart and profiling is a good, professional approach to accurate color profiles, how is it that GretagMacbeth can't even get it right?

--len
Tom Melanson said:
If you've gone to
http://fors.net/scripts/ACR-Calibrator/
and calibrated your ACR RAW converter to your camera, tell us what
you've
found.
I've just calibrated ACR with outputs from the 1DsMkII and the 20D
and compared the results to the output from the C-1 RAW conversion.
Thomas Fors has put forth a lot of work on this and the results
are quite interesting.
My twin Xeon 2.8's with HT engaged ran 78 minutes of solid
computing to calibrate ACR to the reduced 1DsMkII file of the
Gretag-MacBeth color chart, but the results seem to be worth it.
Naturally, the 20D calibration took half the time.
Looks like ACR offers some pretty sophisticated tools for those
interested in getting the color really right--at least til Magne
delivers his profiles for the latest Canon cameras :-).
Feedback?
Tom
Click to expand...
--
--len
 
UnderWater said:
The GretagMacbeth color chart comes in a beautifully printed box
with 2 images of the chart. In both images there are 2 squares that
are meant to be blue that are printed purple.
If the GretagMacbeth chart and profiling is a good, professional
approach to accurate color profiles, how is it that GretagMacbeth
can't even get it right?
Click to expand...
Hey! I use the Mini Color Checker and it only came in a plastic zip-lock bag. I want my box!
 
The two patches to which you refer are: "purplish blue" and "blue flower". Both are shades of blue that have a purplish hue as you indicate, and how they are supposed to be. Purplish Blue would be called "purple" by most people.
Surely you are not questioning the Gretag-MacBeth color checker.
It is indeed a "good, professional" chart.
You can see the chart and the patches here:

http://www.gretagmacbeth.com/index/products/products_color-standards.htm
Tom
UnderWater said:
--len
Tom Melanson said:
If you've gone to
http://fors.net/scripts/ACR-Calibrator/
and calibrated your ACR RAW converter to your camera, tell us what
you've
found.
I've just calibrated ACR with outputs from the 1DsMkII and the 20D
and compared the results to the output from the C-1 RAW conversion.
Thomas Fors has put forth a lot of work on this and the results
are quite interesting.
My twin Xeon 2.8's with HT engaged ran 78 minutes of solid
computing to calibrate ACR to the reduced 1DsMkII file of the
Gretag-MacBeth color chart, but the results seem to be worth it.
Naturally, the 20D calibration took half the time.
Looks like ACR offers some pretty sophisticated tools for those
interested in getting the color really right--at least til Magne
delivers his profiles for the latest Canon cameras :-).
Feedback?
Tom
Click to expand...
--
--len
Click to expand...
 
Hello,

Can anyone tell me if there are any things I have to enable in order for PSCS to find and run the script? I see it in the file menu but the dialog keeps on saying:

"Could not complete the ACR-Calibrator command because Photoshop was unable to find the Javascript plug-in."

What am I doing wrong?

Thanks,
Jorge
 
On their box for the full size card (not the mini).

Had the same issue myself once with blue pansies - came out purple whatever I did with my calibrated Nikon/Epson gear, but quite accurate blue on my Canon cameras/printers.

You can get very accurate on product printing, so this mistake on their part is very amusing to me.

--len
Jason Hutchinson said:
Tom Melanson said:
Surely you are not questioning the Gretag-MacBeth color checker.
Click to expand...
It sounded like he was questioning the box. Lots of variables
there, perhaps some out of their control.

Jason
Click to expand...
--
--len
 
They just got the color grossly wrong. I mean you wouldn't call blue ink just impure if it came out purple would you? Or yellow ink if it came out orange? I'm not talking about perfection here.

My "conspiracy theory" is that they didn't want people to cut up the box and use it as a second card. :)

--len
Brad Sanders said:
Now you're talking about CMYK color gamut, ink purity, press
operator talent...
--
Brad
Click to expand...
--
--len
 
UnderWater said:
The GretagMacbeth color chart comes in a beautifully printed box
with 2 images of the chart. In both images there are 2 squares that
are meant to be blue that are printed purple.
If the GretagMacbeth chart and profiling is a good, professional
approach to accurate color profiles, how is it that GretagMacbeth
can't even get it right?
Click to expand...
It is because they printed the box with only the four CMYK colors. If they printed the box with more colors it would have to be just as expensive as the chart!
Download the Lab chart from
http://www.babelcolor.com/main_level/ColorChecker.htm

and do a CMYK soft proof in Photoshop. When you activate the "Gamut Warning" function you will see that those patches are outside of the CMYK color space.

Barnett
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top