Am 23.06.2014 10:18, schrieb Thomas de Groot:
> Further work on my Urbex scene and HDR compositing as post-process. I am
> using for this four images of the same scene in which the light
> intensity is multiplied respectively by 2, 10, 50, 250. This nicely
> simulates the under- and over-exposure settings used in traditional
> photography.

What I don't understand in your workflow is why you're apparently 
rendering the scene multiple times? Why on earth don't you just use 
OpenEXR output, use some post-processing tool to apply nonlinear 
tonemapping, and be done with it?

One tool that might meet your needs would be Ive's fantastic IC image 
converter (it can also do luminous bloom if you feel like it); for 
scripted operation ImageMagick can probably also be coaxed to do the job 
- or you might even throw together some smart POV-Ray script using 
pigment functions and function pigments.

At any rate, there's absolutely no need to re-render a scene four times 
just with different light intensity. Even if you absolutely need four 
low dynamic range images with different light intensities, you can still 
just render a single high dynamic range image, and apply linear 
tonemapping as a post-processing step to generate four different output 
images.

(You might even be able to get your desired results simply by misusing 
POV-Ray's File_Gamma setting, and writing to a file format other than 
PNG. Not strictly recommended, but might work.)

Post a reply to this message