Mike Russell wrote:

> It's a shame Sinar don't include a "true" curve
> in their Standard Curve set.

What is a "true" curve? One that doesn't suppress highlight and shadow
regions (like sholder and toe on film)? A lot of other stuff is going on
behind the scenes in regard to mapping tonal values to pleasing values, and
trust me, you don't want to miss out on those. A lot of work also goes into
making the tonal response similar to the human eye, and various CCD/CMOS
sensors with huge differences in sensitivity for various spectras (think UV
and IR light). The curve is simply a tool giving you a bit of control over
your images instead of it being a "done deal".

> Since the colour 
> calibration process involves shooting a known
> target (Macbeth 240 patch) you'd think it'd be
> possible to produce a file/curve where the
> readable values of the patches on screen
> correspond to their actual values on the chart.

Supposing the goal of the camera manufacturer would be to reproduce a DC
ColorChecker as true as possible (and ignore all colours more saturated than
that and a few other minor things such as skin-tone contrast and
highlight/shadow definition) it is sort of possible.

It's probably unlikely to be implemented for obvious reasons though.

The only company in the world interested in a perfect representation of the
DC ColorChecker is the company manufacturing it namely GretagMacbeth (OK
include Munsel then). Most camera (and film) manufacturers  are more
interested in pleasing colours and nice looking images. What pleasing
colours is, of course, is somewhat subjective. But so is the colour you
obtain from different brands of film and digital cameras...

Another small issue is that you'd have to read the colour patches in Lab
(the reference you want to compare it to), and to be able to compare colours
would need to compare the values in Lab. This would require an Absolute
colorimetric conversion from Sinar RGB (whatever that may be on your
particular back) to D50 Lab.

Oh, and we are discounting observer/camera metamerism, and gloss appearance,
from that equation too.

It may sound tempting and easy Mike, but it isn't.

The first time I went down that rabbit hole (in '93 on a Leaf DCB I "brick")
I spend 2-3 month on an off with a really good scanner operator, creating a
CMYK conversion in "Photone" and ColorShop that would yield a perfect
cromalin of a Macbeth 24 chart. We eventually succeeded within a percent or
two.

However when we tried to actually convert real images they fell totally
apart with weird posterization, shadows closing down, greybalance skewed
etc. results were totally unusable for anything real life - but it was spot
on for the ColorChecker...

Morale: 
We didn't (and still doesn't) make a living shooting colour checkers, so
using that as a base for defining how a conversion should be done wasn't
(and still isn't) a wise decision... There are other more important
considerations than matching a set of colours no one (Ok, so no one except
you Mike <VBG>) cares about...

Merry christmas



Best Regards

Thomas Holm / Pixl ApS

- Photographer & Colour Management Consultant
- Adobe Certified Training Provider in Photoshop�
- Apple Solutions Expert - Colour Management
- Imacon Authorized Scanner Training Facility
- Remote Profiling Service (Output ICC profiles)
- Seminars speaker and tutor on CM and Digital Imaging etc.

- Home Page: www.pixl.dk � Email: th[AT]pixl.dk
-- 


===============================================================
GO TO http://www.prodig.org for ~ GUIDELINES ~ un/SUBSCRIBING ~ ITEMS for SALE

Reply via email to