Home Products Download Order Contacts


Subject: Re: raw 3.1 professional custom profiles for highend digital cameras

Is it preferable to retain the initially lighter or the more normal looking
image in terms of colour and tone in the higher and middle levels.

I think this question deserves the proverbial cigar.

I believe the answer should be as follows: it depends what the image will be used for.

If it is used as is, involving normal adjustments, almost directly to output, you can best use the normally exposed version because it is the more efficient workflow.

If it will be used as a basis for further manipulation, possibly involving extreme adjustments, then and only then will it be advantageous to use the lighter version.
(It can be argued that extreme resizing is part of this).

Why I think this is the $64K question: because if we are discussing the possible merits of a different exposure paradigm, we should also consider how the RAW converter implements the tone-curve. It seems counter productive to expose for the (precise) highlights, and then have the tone curve be implemented around a middle gray paradigm/pivot-point.

It basically means that any exposure requires a custom curve adjustment because the brightness/contrast combination can not be shifted far enough. And even if it could, in either case you end up with completely unpredictable results. You obviously get what you see, so you can make the adjustments, but it is not predictable as in; process a ton of images and only adjust the exceptions...

I don't know what would possibly be a better toning control, but I can imagine that the contrast curve might be fixed to the highlights and the pivot point moves, such that increasing contrast will always ensure the entire curve is below the previous...

Whether a Photographer will feel comfortable with such a control remains to be seen. It could of course become a user preference?

ps. Regarding the 15bit+1 issue. If they are talking about signed integers they mean 32bit integers. The 15bit+1 is unsigned and meant for really fast multiplies (read blending). And goes like so: an image pixel ranges from [0 ... 32768]. Multiply image pixels (as in a blending mode) and you get a maximum value of 32768x32768. This still fits in a 32bit signed integer. A really fast bit-shift operation can be used instead of a very slow division to bring this value back to the image pixel range.


View All Messages in adobe.photoshop.camera.raw

raw 3.1 professional custom profiles for highend digital cameras =>


Copyright 2006 WatermarkFactory.com. All Rights Reserved.