Legacy ImageMagick Discussions Archive

Hello all,
I'm learning the basics of scripting with IM.
Is there a way to optimize the following into a single convert operation since I always start from the same source file: And two side questions:

• I saw it's possible to do general floating point maths with IM with "%[...]" but where is the complete explanation (syntax, list of functions, etc) about that ?
• is there a way to get write support for the DNG format ? (something else to install / compile ?) I'm on Ubuntu.

Try: or

Thanks, that would have been hard to figure out. Actually, after some tests I'm not convinced this is much faster than the simpler multiline command.

It is faster as it performs the exact same operations, but only reads the source image once.
It also saves of start up processing (loading commands, libraries etc) needed.

Depending on the operations you may also gain speed by not needing to save and re-read intermediate image files, which not only can take time, and is generally slow due to disk IO, but also can introduce loss of accuracy as image files format usually are 8 bit depth, when default in memory store is a 16 bit depth.

In fact, not saving intermediate files become a crucial part for very advanced fourier transform image processing techniques using HDRI (floating point).