Page 1 of 1

16-bit / channel RGB output formats with Q16

Posted: 2009-09-21T08:24:40-07:00
by Bob-O-Rama
Hi,

I'm using Magick++ and the Q16 libraries, to produce a series of files that are handed to ffmpeg and converted into video. My issues is that I need a file format which can store 16-bit / channel RGB and be read by ffmpeg. ( The underlying issue is that most video formats are 10-bit / 12-bit and hence using a format like JPG causes posterizing of delicately shaded areas. ) I currently use MIFF for lossless intermediate files for my internal processing - which works fine, albet it bulky. I just need something - even if lossy - that I can use to ferry the data to ffmpeg.

Alternatively, some means by which to create the video directly from the 16-bit depth Image objects.

Any suggestions on "keeping all my bits?"

Thanks!

-- Bob

Re: 16-bit / channel RGB output formats with Q16

Posted: 2009-09-21T09:13:02-07:00
by magick
Set the compression quality to 100 when you create the intermediate JPEG images. Its still 8-bit but the quantization artifacts should be diminished.

In the mean-time, you can try PNM. Its a 16-bit format supported by ffmpeg.