[Digikam-users] 16bit editiing and auto-whitebalance

[elle stone]

Bugzilla from arnd.baecker at web.de wrote:
> 
> On Tue, 15 Apr 2008, Gilles Caulier wrote:
> 
> Should we now switch to 16 Bit as default ?
> 
>> Done in 0.9.4. auto-gamma is implemented with 16 bits color depth
> 
> Most likely this is not digikams fault (i.e., ufraw's output
> looks similary dull to me) as no one (in open source) knows
> the secret methods of raw to jpg conversion used by Nikon
> (same with Canon of course, but there the results
> looked much better, in the cases we looked at together, Gilles).
> 
> I am not sure if much can be done, though ....
> 
> Best, Arnd
> 
> 

Hi, all,

I think it's very true that Canon and presumably Nikon have some kind of
"secret sauce" that they use to produce really good-looking images, whether
straight out of the camera as jpegs, or whether as the result of using the
proprietary Canon/Nikon raw developer software.  The "secret sauce" has two
parts - color (hue and saturation) and tonality (distribution of lights and
darks).  

Canon DPP, the "not for linux" Canon proprietary raw development software,
is based on the same "picture styles" that Canon uses for in-camera jpegs. 
Most Canon picture styles automatically add a heavy S-curve and extra color
saturation to give the picture more "pop".  Even if you (1)choose "neutral"
(the Canon picture style that gives you the most accurate "scene-referred"
colors); and (2)select "less contrast", "less saturation", "no noise
reduction", and "no sharpening" in the DPP raw development dialog, you will
find, if you know what to look for, that an S-curve and also shadow
denoising has been applied to your image.

Dcraw gives you the lights and darks that are actually recorded by the CCD. 
According to Tindeman
(http://21stcenturyshoebox.com/tools/ACRcalibrator.html; alas the site is
down at present), dcraw is one of only a very few raw developers that
actually gives you the "scene-referred" tonality.  And it IS flat-looking,
because the camera CCD records light linearly, whereas our eyes are
constantly interacting with our brain to accomodate dim and bright areas in
a scene. 

Color fidelity is a different story.  Color is where the Canon (and
presumably Nikon) secret sauce really comes into play.  With the latest
dcraw, you can apply Canon's camera-model-specific color profile(s) to the
16-bit dcraw output tiff during the raw development process, and the colors
will still NOT not be exactly the same as what Canon produces.  I've checked
extensively, using an "eyedropper", comparing the output of various raw
developers using various camera profiles - a very tedious though instructive
process.  Canon default jpeg and DPP colors are really good.  (If you want
to hear some belly-aching about Canon colors being "better", try a google
search on what Adobe Camera Raw does to Canon colors - NObody likes the
default ACR output for Canon raw files.)  It seems to me that the only way
to "do better" color-wise than what Canon produces with their "secret sauce"
is to produce your own color profile for your camera, using Argyll, for
example, and give your custom color profile to dcraw/digikam,ufraw/etc.  I
haven't taken that step yet (next on my "learning the digital darkroom"
agenda).  

So why on earth would anyone want to use dcraw to develop a raw file, only
to have an image with rather flat tonality (no S-curve), noise in the
shadows (no noise reduction unless you specifically choose the wavelet
denoising, which I never do), and colors that are not as accurate as what
Canon can produce "in-camera"?  

The answer is artistic control.  When you take a picture, presumably you
have an idea of what you want the final image to look like.  It is much
easier to achieve that final image if you don't have to "undo" stuff that
has already been done to your image.  Once Canon (or Nikon, or Bibble, or
etc) has applied their proprietary S-curves and shadow-denoising,
sharpening, etc to your image, then your shadows (highlights, edge detail,
etc) are already squashed, clipped, chopped, and otherwise altered and
mangled.  You've thrown information away and you can't get it back. 
Especially in the shadows, even with 16-bit images (actually, 12- or
14-bits, depending on the camera, but it's encoded as 16-bits for the
computer's convenience), there just isn't that much information to begin
with, as explained in Ron Bigelow's excellent article, 
http://www.ronbigelow.com/articles/raw/raw.htm.   

Anyway, I hope my long-winded discussion of why dcraw output looks "flat"
helps cast some light on a confusing subject.  "Flat is good" if you'd
rather give your images your own artistic interpretation.  The alternative
is to let the canned, proprietary algorithms produced by Canon, Nikon,
Bibble, etc interpret your images for you.  Which is not entirely a bad idea
- for most images, most of the time, those canned alrorithms are really
pretty good!   

If I've made mistakes in my explanation, please correct me!

Elle

-- 
View this message in context: http://www.nabble.com/16bit-editiing-and-auto-whitebalance-tp16701397p17135280.html
Sent from the digikam-users mailing list archive at Nabble.com.