12
$\begingroup$

Blender can read Apple ProRes video files.

Is there any way to write to that format?

I use a video workflow with Final Cut Pro X and used Blender for some retouching and compositing and would like to send the clips back to FCP.

$\endgroup$
7
$\begingroup$

It is likely wiser to adopt an image based pipeline and assert that all of your color transforms and bit fidelity remain intact.

Using an image as an intermediate will be more vastly more effective. The granularity required to handle the color primaries, bit depth, and transfer curve out of a codec correctly is not currently possible within Blender nor many other mainstream applications, and will often result in incorrect output. Well documented issues such as Quicktime's perceptual noise addition, incorrect transfer curve / gamma, inconsistent vendor flag reading, etc. all make for the stuff of nightmare when dealing with codecs.

If you are careful, you can easily step to an intermediate image format and use the Baptiste Coudurier FFMBC project to encode and decode ProRes with much more control, quality, and accuracy.

While it is a fork of FFMPEG, FFMBC extends the needs of the encoding and decoding of ProRes and DNxHD to meet media broadcast needs.

Further links:

$\endgroup$
  • $\begingroup$ I haven't experienced these problems with ProRes yet. It is a lossless format after all. And FCP really likes it. So I want to stay with it. $\endgroup$ – Maccesch May 30 '13 at 5:50
  • 2
    $\begingroup$ The problems are inherent to the fact that it is a codec, and the chances are that you have experienced them, but have been unaware. There are very good reasons why independent to higher end productions rally around EXR during effects and finishing work. This forum isn't well suited to discuss the nuances however. A final minor point, ProRes is not lossless per se, and every export to it will suffer. FCP only "likes it" because FCP is sold as a very Apple-centric ecosystem. $\endgroup$ – troy_s May 30 '13 at 18:03
3
$\begingroup$

I am assuming since it's Apple based software, the format has some ties to QuickTime. The closest option you have is to export a video from Blender (excluding the default AVI formats) are

(formats that can be compiled into blender, not available on all systems)
HDR TIFF EXR MPEG AVICODEC QUICKTIME CINEON DPX

Also, Quicktime support isn't available on Windows or Linux, so these can't read image/video formats supported by Quicktime.

$\endgroup$
  • 6
    $\begingroup$ FFMEPG apparently has ProRes support (both decoding and encoding), so there should be a way to access it from Blender once the developer exposes the setting in Blender. But it doesn't look like it's possible now. Source: transcoding.wordpress.com/2012/01/29/prores-ffmpeg $\endgroup$ – Mike Pan May 29 '13 at 6:48
  • 2
    $\begingroup$ @MikePan interesting it may be possible to add support through getting blender to use the codec in ffmpeg then, however we dont compile with all codecs enabled so would need investigation. $\endgroup$ – ideasman42 May 29 '13 at 19:50
2
$\begingroup$

When outputting from blender you have a multitude of options with the FFMPEG libs. It's even been hinted that you can output prores 422 on linux with a patch to the correct library. However, blender is ill suited as an encoder in this respect, most specifically due to the "grey area" in legal issues pertaining to output formats like pro-res. That said, my workflow with blender to prores has followed this pattern in the past: blender output to images (tiff\tga) Folder of images to sequence in FCP or another editor, applying the sound etc, then out to pro-res

This workflow has proven faster at encoding the video, and has even produced higher quality with compressed formats like mp4 (with the same settings). The reason is simple enough. I farm the image rendering, so the time it takes to render the frames is cut down considerably (I have 5 old computers running blender farming in linux, varying distros, and varying power, but still core for core a better option). This produces a full format video similar to a celluloid reel of frames. Placing this into a sequence shows this as fully rendered, applying a sound track lines up perfect (especially when from the original video file). When doing this with source video for multi cam edits, I usually have to allow FCP to create the proxy for use in playback, and in adobe, render a preview. But when this is done from an edl, imported and farmed after the edits are marked, it takes longer to render the images, but shorter overall to then output a finished video from them, and with parallel encodings running, I can output several finished file formats for different types of delivery (everything from youtube and blu-ray to dvd). You are better off using a dedicated apple computer for the encoding of pro-res. Occaisionally, I've even just used compressor to put the two (audio and video) together, though not recently.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.