Is a MacBeth Color Checker very useful?

[Jesper]

I'm shooting in RAW mode most of the time with my 10D. Often I don't like the colors that I get right out of the camera in RAW mode...

Now here's an interesting article about calibrating Camera Raw in Photoshop CS (http://www.creativepro.com/story/feature/21351-1.html) using a MacBeth color checker. It sounds like a good way to get the colors as close to how they are in reality as possible.

The problem is, the MacBeth color checker is not a cheap thing that I'd buy without knowing if it's useful or not - it costs more than € 100 (US$ 120)! (at Colour Confidence (http://shop.colourconfidence.com/product.php?xSearch=&xProd=78&xSec=1&xCmd=cc&xCur= 2)).

So, has anyone used a MacBeth color checker and how useful is it? Is it worth the amount of money it costs?

[John_T]

I was just reading the same article this morning. Apparently it all goes back to a problem that occurs with PS Camera RAW 2.2 because Canon and other manufacturers are loathe to give Adobe the full works on their proprietary RAW conversion formulas. On the Adobe and DPReview forums there has apparently been an uproar about this since April.

To my knowledge, you simply download the MacBeth color checker and shoot it on a calibrated monitor. It doesn't cost anything. I downloaded it, but haven't gone further yet.

Another method to calibrate your 10D in Camera RAW is to shoot identical images, one RAW and one Auto jpeg, then with the jpeg open in PS and the RAW open in Camera RAW, play with the Calibration tab sliders until you get the RAW image to match the jpeg colors. Don't know accurate this would be, but it's worth a try.

I haven't done either yet, but trying to use Camera RAW without having calibrated it to the camera is a mess. :roll:

[Scottes]

Interesting stuff. I like the JPG/RAW test - although I use CaptureOne so I don't know what that would be like.

As a note I just saw a mini version on MacBeth's site (I believe) for $55 or $57 US. Nope, it wasn't MacBeth's site, but I know it was cheap. Adorama lists the mini for $80 US.

[John_T]

I wonder if we are talking about the same thing. I just downloaded the MacBeth ColorChecker here:

http://www.colorremedies.com/realworldcolor/downloads.html

Didn't cost anything.

[4walls]

John_T, Would you not have to shoot the MacBeth chart with your own camera in
order to accurately adjust your monitor and camera raw settings?

I assume that you (Scottes) have calibrated your monitor already, right?

[John_T]

Doh, I get it. You need to buy the physical target that is printed color accurate and shoot that. That is also a bit tricky because you have to set up the target in the correct light to get the correct colors. If you read this thread, you will see how tricky the whole methodology is even among the experts.

http://www.adobeforums.com/cgi-bin/webx?13@146.VWThcw6xNPJ.3141056@.3bb45ec0/0

I just printed the target out absolutely untouched with my i9950 on Canon MP-101 and I'll bet the print doesn't vary enough from the original to matter much. Maybe my logic is off, but if I calibrate ACR to a shot of this print, I'll be calibrating simultaneously to output on this printer.

If you do decide to print it yourself, I would recommend in the PSCS Print with Preview dialog:

Source Space: Lab Color,

Print Space: Profile: Printer Color Management,

Intent: Relative Colormetric.

In the Printer dialog choose a best quality matte paper made by your printer manufacturer, quality the highest, color Auto, no ICM.

This apears to give the purest results. To me. :wink:

Color management is Pandora's box...

[Scottes]

My understanding of the Macbeth (one use of it at least) is to shoot a frame in the light used for the shoot. That is, have a model hold it and take a picture, then start the shoot. When done the first shot allows you to "calibrate" the color settings and white balance, allowing you to then batch process the rest of the shoot.

Another might be to shoot it once in perfect light and calibrate CS's RAW. Another would be to scan it and use the scan to calibrate your scanner. And finally once everything is calibrated you could print it and use it to calibrate your printer, but I don't see much accuracy in this last method without a spectrophotometer to measure the printed color samples.

The Macbeth is an extremely accurate color sample - judging it by eye would allow for a "subjective calibration" - pretty close I'd say, but I couldn't use the word "accurate" without the use of some type of measuring device to measure the samples.

John's method of downloading and printing would fall into the same "subjective" area - he's probably more accurate than I am - I'm uncalibrated all the way - but still not yet completely accurate.


Unless I were doing model shoots in controlled lighting situation then I would not bother buying a Macbeth. Then again I'm knowledgable on the subject but far from an expert. And I might be missing something.

[maderito]

Doh, I get it. You need to buy the physical target that is printed color accurate and shoot that. That is also a bit tricky because you have to set up the target in the correct light to get the correct colors. If you read this thread, you will see how tricky the whole methodology is even among the experts.

http://www.adobeforums.com/cgi-bin/webx?13@146.VWThcw6xNPJ.3141056@.3bb45ec0/0

I just printed the target out absolutely untouched with my i9950 on Canon MP-101 and I'll bet the print doesn't vary enough from the original to matter much. Maybe my logic is off, but if I calibrate ACR to a shot of this print, I'll be calibrating simultaneously to output on this printer.

John_T,

I had to think about this a bit, but I concluded that you are guilty of (to use your favorite term) "the dog chasing its tail."

If you print the Macbeth chart on your trusty Canon, and a blue square prints purple (perhaps to everyone's eye but yours), then after calibrating ACR to a shot of this target, all the same blues will print purple in your images. It's like trying to perform a custom WB on a target that's not neutral gray or white.

You need to start with (and pay the bucks for) a target that is accurate by some independent, universally accepted criteria.

OTOH - if your Canon (and mine) prints perfect colors, then you would be chasing perfection - still a dubious goal. :roll:

[John_T]

Scott and Woody, of course you are both right, but of course I've gotta see how far I can get with this Jerry-rigged trial before getting serious about it. I don't expect it to be spot on accurate, but I would like to establish a constant to work to, and see if I can refine from there and get a usable profile in ACR.

I really would like to use ACR because of it's features, but is short on the complete Canon camera profiles, which understandably, Canon doesn't want to fork over. Open source is a nice idea, but not very good for business. Adobe are trying to dig it out, but according to Thomas Knoll, author of ACR, it may take a while and come out in Camera RAW 2.3. I'm just trying to set myself up in the meantime. Har-dee-har...

I am calibrated monitor through to print, but even so I'm aware that it is all relative and not absolute. I am very surprised how pure and clear the Color Checker printed out, so what the heck, I'm going to have a go at it. Since everything is ultimately relative and subjective anyway, nothing to lose. If it turns out to be junk, I can just dump it and wait for 2.3. :lol:

[robh]

Have a look at this thread from Dpreview. It provides some interesting reading, and settings for PSCS ACR (especially for us Pro1 users).

The settings (with some slight adjustments) have amazed me, at least for shots under average daylight..

http://forums.dpreview.com/forums/read.asp?forum=1010&message=9511092

cheers ... Rob

[John_T]

Hi Rob. Yeah, I went through the whole MCC thing with difficulty, read the DPReview thread and traced it out to a few other places. It's a subject on the Adobe forums too. I've only had limited success, and certainly nothing I would save as a default start to RAW conversion. I think the Canon color handling formulas go much deeper and making a few tweaks in ACR isn't going to cut it. Perhaps with Camera RAW 2.3 Adobe will have made progress, if not all the way.

There are several alternatives such as C1 that Scott uses, but it doesn't support so many cameras.

I'm going back to the Canon software for the time being, and though it doesn't have all the tweaks one might desire, it does have all the formulas, and if it doesn't crash, it's reliable. :roll:

[robh]

John_T, thanks for the the additional information.

I'll keep my fingers crossed for what ACR 2.3 might bring!!!

cheers... Rob