Re: FW: Digital Camera Profiling
Re: FW: Digital Camera Profiling
- Subject: Re: FW: Digital Camera Profiling
- From: bruce fraser <email@hidden>
- Date: Fri, 18 Apr 2003 13:36:54 -0700
At 8:36 PM -0400 4/17/03, Derrick L. Brown wrote:
Well on your request I'll provide more detail.
Good camera profiles are based on consistent grey and/or white balance
procedures being followed throughout the course of any given shoot..
As it has been noted, digital cameras can easily see 20 to 50 degrees of
kelvin shift very quickly.
One of the more important tools a modern digital photographer can own is a
color meter.
As a fan of making camera profiles we evangelize that in order to maintain
consistent results, in camera grey/white balancing is a must. If you do not
do this be prepared for unpredictable color shifts (casts).
OK. I totally agree. But that's a very different statement from
At 5:03 PM -0400 4/17/03, Derrick L. Brown wrote:
carefully
built profiles allow you better color, tonality and faster workflows, No
exceptions.
Now believe it or not, I'm actually going to say, ACR is certainly better
than having nothing. If you are confronted with digital captures and you
know little about how they were performed then it offers a "starting point".
This is where I did compare ACR to Knoll Gamma (of the past).
Please hear this out.
In ACR you open a raw capture, a "White balance" is read/diplayed. Then
pick a targeted workspace. A multiple choice question. Change the
selection with no apparent changes to the image.
The user now continues down the path with "sliding bars" to create pleasing
color.
To this end you have completed your first corrected image.
Now open the next raw capture. Instead of the previous image which may have
been shot in the open field at noon, this image is one hour warmer (later)
and under the shade tree. Firstly white balance will obviously be way off.
So now you observe the read/displayed balance value reading cooler than it
should be. So you in turn shift the "Temperature" slider more warm. But
alas, now we are getting a major color shift now so please adjust the "Tint"
slider. Again, I cannot help but thing of how time intensive this is
process becomes.
That's certainly one way of working with CameraRAW. It's probably the
method that will let picky photographers best realize their vision,
but it is indeed labor-intensive. However, it's not by any stretch of
the imagination the only way of working with CameraRAW.
You don't need to choose a working space every time you launch
CameraRAW. Most people choose one, and stick with it. IAC, I've never
seen a custom profile that provided a good EDITING environment, so
whether you use CR or custom profiles, you generally have to pick a
working space. That's a red herring.
If you're shooting in a situation where you have the time to set a
custom white balance on the camera, you do so. When you open the
image in CR, the color temperature defaults to "As Shot", reflecting
your choice of WB. If the color is right, you're done aside from
making subjective edits for tonality. If the color is wrong, you
correct it, and save it as a setting for that WB setting. Using a
shot with some kind of target makes the process less subjective.
Repeat as necessary for different WB settings. My experience has been
that dialing in CR for a particular camera is generally a somewhat
quicker process than building a custom profile.
Some cameras have trustworthy auto-WB. Some do not. For those that
do, you can simply leave the CR default as "As Shot." For those that
don't, again you'll have to create and save some settings. My
admittedly limited experience has been that, with cameras whose
auto-WB leaves something to be desired, in situations where you can't
WB every shot, the control offered by CameraRAW interpolating between
two generic profiles for the camera gives me better results than any
single custom profile. This is particularly true of cameras like the
DCS 460, which has vastly different responses under tungsten and
under daylight. But that's a worst-case scenario.
Again, all I'm really saying here is if photographers learn why (and how )
to deliberately grey/white balance a custom profile can bring them 95% of
the way to their goal all the time.
I don't disagree with that at all. But it's a far cry from your
original statement, and it's an impossible criterion for many kinds
of imagery, Including a good many types of imagery for which people
get paid.
If they ignore it or worse yet perform
it improperly, you in fact will create headaches for either approach.
Almost certainly true, but true to very different degrees on different cameras.
I am also saying that though you obviously will spend additional monies to
build your own profiles, you will in fact be camped out for far less time in
front of the computer.
Subject to the above fairly serious caveat about grey/white
balancing, with cameras whose white balance actually works across the
range of lighting you use, that is definitely true. But in all those
cases where either the kit or the situation doesn't allow white
balancing every shot, CR may offer a more flexible and equally speedy
solution.
When Andrew and I independently raised the proposition that a custom
digital camera profile may not be the best solution for all cameras
and all shooting situations, a distressingly large number of people
who think they believe in color management more than Andrew and I do
reacted as if the barbarians are at the gate and interpreted all our
statements as saying that custom profiles can't work. Neither of us
ever said that. They're just tools. They work in some situations, and
fail in others. CR offers an alternative approach that we've both
found interesting and, more importantly, useful.
I suspect that the best of all possible worlds would be CR used with
two custom profiles for the camera, one created for tungsten, the
other for D65, like the generic ones in CR. However, that would make
it a massively more complex product, and unless the custom profiles
were made to at least the same standard as Tom Knoll's generics, it
might produce worse rather than better results.
The bottom line, though, is that neither Andrew nor myself has
dismissed custom camera profiling. We've simply reported real
experiences. A lot of the naysaying about CameraRAW seems to be
coming from a significantly less informed perspective, and moreover
one that glosses over the very significant difference between
rendered and unrendered imagery. The basic science on which ICC color
management is based was designed to predict the degree to which two
solid color patches, of a specific size, at a specific distance,
under a specific illuminant, in a specific context, would appear to
match or not match. It's a huge stretch to apply that model to
individual pixels in a complex color event like a photographic image.
It works remarkably well, but fairly often the colorimetrically
correct interpretation of a pixel doesn't produce a photographically
useful result. That doesn't mean we have to throw out the baby with
the bathwater, but it does mean that we need to question extreme
claims like the one quoted at the head of this message. I suspect
that our areas of agreement are much bigger than our areas of
disagreement, and I apologize if I seem to be giving you a hard time.
I'm just trying to keep us all honest, and I hope you'll all continue
to do the same for me.
Best,
Bruce
--
email@hidden
_______________________________________________
colorsync-users mailing list | email@hidden
Help/Unsubscribe/Archives:
http://www.lists.apple.com/mailman/listinfo/colorsync-users
Do not post admin requests to the list. They will be ignored.