Thanks for responding. I wasn't aware of DCamProf and will dig into that. I
do work in art reproduction and creating profiles for specific camera/light
setups is a routine part of the work. My main interest was if spotread
could continuously write to a file but it seems that it was not designed to
do so. I will continue building the reference file i need manually between
patch measurements.
My favorite raw editor is Rawtherapee and I had forgotten about RawDigger.
thanks for mentioning that.
Now getting to the spectral workflow, I am aware of this kind of approach
and is something I want to pursue at a later date. until then i will need
to become more familiar with Argyll's .ti3 format and capabilities.
Thanks everyone.
On Mon, Apr 9, 2018 at 6:10 AM, edmund ronald <edmundronald@xxxxxxxxx>
wrote:
General purpose profiles require skill, but a quick local profile for a
given shooting condition is useful and easy - even a normal colorchecker is
sufficient.
Edmund
On Apr 9, 2018 04:34, "ben" <ben@xxxxxxxxxxxxxxxx> wrote:
On Apr 7, 2018, at 9:22 AM, Joseph Aaron Campbell <joacampb@xxxxxxxxx>
wrote:
other methods of profiling camera are welcomed.
Camera profiling is not for the faint of heart.
For starters, a given profile is only valid for exactly one illuminant --
and, unless you have a way of including a characterization of your scene's
illuminant in the process of building your profile, it's logically
impossible to create a valid profile.
Next, charts are entirely inadequate for characterizing a camera's sensor
unless the spectral characteristics of the chart are a good match for the
subject.
What this means in practice is that chart-based profiling has a _very_
narrow applicability: if you're doing fine art reproduction in a studio and
you've handmade your chart with as exhaustive as possible a set of the
pigments your artists are using.
For a more general-purpose profile-based workflow, you need a spectral
model of your camera sensor's response. Ideally, you'd also get an ambient
spectral measurement of the scene's illuminant, but you can often do "close
enough" with knowledge of the type of illuminant (daylight, incandescent,
standard fluorescent, etc.) plus a ColorChecker. You also need to know the
spectral transmission efficiency of your lens as well as any filters (etc.)
in front of the lens. If you have all that, the math to create a synthetic
.ti3 to hand to Argyll is pretty straightforward.
If you don't understand what I typed...you'll either first need to bone
up on your color science chops to get to the point you do, after which
you'll know how to take it from there...or you're very, very, very much
better off using the profiles supplied with your RAW converter. (And, if
you need a good RAW converter, RawDigger is the one to beat.)
Those profiles that come with your RAW converter aren't good enough for
high-end reprographic work, but anything you're likely to create with an
off-the-shelf chart is going to be worse.
That writ, off-the-shelf charts can have a significant place in a
colorimetric workflow -- namely, normalizing channel ("white") balance and
exposure, as well as fine-tuning model estimates of an unmeasured
illuminant's color temperature. The ColorChecker Passport is especially
good for this, including its durability in the field. Just don't expect you
can build a decent profile from one.
Cheers,
b&
