>> If a person is shooting in-camera JPEG, I honestly cannot fathom a
>> reason why you would ever select aRGB in-camera. And, for me, I am
>> thinking of the output. I tend to work backwards (does this surprise
>> anybody?). If the output is sRGB, why subject myself to conversion
>> steps if I don't need to?
>
> Simple, 'cause when you bend the tones, you get gaps and piles in the
> histogram. Sometimes not noticeable, sometimes
> ugly. Worse, sometimes just off, and hard to see how/why. And once it's
> happens, and is noticed, it's back at least
> several steps.
>
> But I may be unjustly maligning your idea. Do you convert to 16 bit at the
> beginning?
The premise is "in camera JPEG". An in camera is always going to be an
8 bit per channel file. If you are shooting 8 bit with aRGB, you are
much more likely to end up with TONAL compromises at the expense of
expanded color pallet. You've only got 256 possible values per color
channel. You've got to steal them from somewhere. Personally, I would
much rather steal bits from the top two brightness zones, at the
expense of the highlights in order to get more bits in the shadows.
But the idea behind "color spaces" is essentially that you remap bits
to accomplish a desired task.
Now, where this all makes perfect sense is if a photographer is using
one of those new-fangled Fujifilm X-Thingies that do everything but
the laundry, and you shoot stock. As good as the X-Thingies are, I'd
be terribly tempted to just shoot in-camera JPEG+RAW, but never touch
the RAW file unless something needs addressing. Since I'm such as
great photographer, I can shoot the pictures without need of any (or
minimal spot) editing. In this case, I can just take the JPEG files in
aRGB and send them straight to my stock agency, as they still prefer
aRGB. With the right in camera settings appropriately assigned for the
image at hand, the color and contrast of the images are pretty much
nuts on.
Yet, as somebody who has the vast majority of his images destined for
on-line display and printing through a lab that prefers sRGB, I see
absolutely no practical purpose to shoot in camera JPEG at aRGB.
That's just stupid busy work.
> I agree, and when I work on sRGB images from the web, I go 16 bit, but stay
> in sRGB.
Any editing immediately involves the conversion to 16 bit. However,
not all 8/16 converters work equally well. I have found that my very
best and consistent results are with Picture Window Pro. PWP doesn't
do much, but what it does do is nerdy good. Some converters work by
hacking the bottom off or just direct map 8 into the top half of 16.
Others will remap the bits causing a "comb" look to the histogram.
There are even some which will interpolate the bits to smooth out the
comb. Things can get really, really ugly, if you take an 8 bit file,
convert to 16 bit, make a minor tweak and then convert back to 8 bit.
If the the 8-16 conversion resulted in combing and the 16-8 conversion
uses every other bit tossing, you can end up with an adventurous file.
It's been a long time since I've ever dealt with that, though, because
I've been using PWP for a long time.
--
Ken Norton
ken@xxxxxxxxxxx
http://www.zone-10.com
--
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/
|