[argyllcms] Re: help with camera profile

  • From: Iliah Borg <ib@xxxxxxxxxxx>
  • To: argyllcms@xxxxxxxxxxxxx
  • Date: Thu, 2 Aug 2012 10:25:42 -0400

On Aug 2, 2012, at 10:07 AM, Stephen T wrote:

> "Cannot the problems with shadows/highlights be related with the dynamic 
> range of the camera?"
> 
> No single exposure of a test chart can fully exercise the dynamic range

But 2 exposures separated by 4 or 5 EV (depending on the camera) wiil.

> and colour gamut of a modern digital camera

Camera records all the colours presented to the sensor. RIT Labs have a good 
explanation of this in their FAQ section
http://www.cis.rit.edu/mcsl/faq3#255

> . What we need is a profile model (and workflow) that can extrapolate beyond 
> the bounds of the test chart and consistently deliver pleasing results.

Pleasing results are not a matter of math or physics, it is a matter of 
experiment and taste.

> Linear transforms are well-behaved and predictable.

And they are usually adequate; however - I work with non-linear sensors among 
others, and any ordinary sensor can be non-linear close to shadows and 
highlights. On top of that, CFAs are not always following Luther-Yves condition 
close enough and data coming from those can't be converted back to scene using 
a matrix transform. Next, Foveon-type sensors and multi-spectral imaging are a 
separate, though important case.

> For general photography, matrix profiles seem to perform well

Indeed so.

--
Iliah Borg
ib@xxxxxxxxxxx




Other related posts: