| I haven't worked on a blown colour casted highlight reconstruction
| recently, but it strikes me that we can perhaps deal with such colour
| casts more easily in dt than any of the other raw converters. Let's
| say our reconstruction causes a typical magenta colour cast in the
| reconstructed areas around totally blown highlights - I suspect we
| could just use our parametric masks to run a magenta subtracting
| colour change just on the brightest pixels and if necessary add a
| drawn mask as well.
My view is that I'm all for sophisticated algorithms here in general
but I think that *by default* darktable should produce results that do
not require users to go in with masks and colour manipulation and so on.
Because aggressively reconstructing clipped highlights can blow up in
your face I don't think it should be on by default. Experts or people
who really need it can turn it on (and then clean up any issues).
Hence also my bias towards a simple option for 'if any channel is
clipped, make the pixel white', because I see it as the quick way to
make sure that clipped highlight reconstruction isn't blowing up in your
face. You're very likely to get a sensible result from it.
- cks
------------------------------------------------------------------------------
How ServiceNow helps IT people transform IT departments:
1. A cloud service to automate IT design, transition and operations
2. Dashboards that offer high-level views of enterprise services
3. A single system of record for all IT processes
http://p.sf.net/sfu/servicenow-d2d-j
_______________________________________________
Darktable-users mailing list
Darktable-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/darktable-users
