Build Camera Profile and Match Reference

[tomczak]

Could I be briefly explained the major differences between using Build Camera Profile and Match Reference (say, using Macbeth 24 patch target). The approach is different, but the objective is the same, isn't it?

I just started experimenting with Build Camera Profile, but I have played with Match Reference in the past:

http://www.dl-c.com/board/viewtopic.php ... match#p987

(BTW: I can't see the example image illustrating the point there anymore)

I had 2 problems with Match Reference: clipping and dependency of colour correction on exposure.

The major problem I had there was that in March reference, the colour curves mapped the 6 gray patches about right, but insisted on clipping the curves tails. This made using the Match Reference somewhat difficult to automate the correction process.

The second problem was that in Match Reference there are two independent steps: RGB Colour Curves on gray patches, followed by Colour Correction in HSV. The net result is that colour correction step depends significantly on exposure (clipping aside, the exposure is corrected properly by the Colour Curves step, but that wrecks havoc to the colour patches and subsequently Colour Corrections needed), and that makes it difficult to match colours when exposures differ.

What should I watch for when trying to do pretty much the same thing, but building Camera Profile instead? Did others do that successfully? Do I have a chance to be more successful using Camera Profiles than I was with Match Reference?

Cheers!