de-anamorphizing in post production

[Mr. Paul White]

Another poster has opened a new thread:

http://www.reduser.net/forum/showthread.php?t=5226

Seems the Red team is working on a new feature... Awhile, I'm asking how to handle with it in post production? I mean de-anamorphizing not only for monitoring but also the workflow itself.

[Kevin Halverson]

I post, one simple method is to set the PAR (pixel aspect ratio) to the reciprocal of the lens character. If your using a cinema style anamorphic lens (2:1) then the PAR would need to match to keep both the displayed and final images correct. Other anamorphic approaches (like the Panasonic adapter) are much less aggressive (only 1.33:1) and require a less exaggerated PAR shift.

[Zack Birlew]

One guy in a 2.35:1 thread somewhere mentioned that footage using the Panasonic anamorphic adapter plus a native 16x9 camera could be set correctly in FCP by setting the pixel aspect ratio to -36. I've done tests with my HV20 and the adapter and it really does work nicely, especially for DVD output.

For special effects, I'm not sure what to do first. Squeeze the effects? Unsqueeze the footage first? One problem I ran into was that if you wanted to layer footage on top of each other and you had to alter the center point of one of the layers to be lower or whatever, the 2.35:1 unsqueezed bar would follow no matter what and it would bleed over into the other layer's black bar area too. So, in that sense, you've got to be careful or you've got to be prepared to work with the footage when it's squeezed, render out a single video file, and then unsqueeze everything together and hope it blended nicely together. It's a learning process, but it's easy to figure out. I wish that some manufacturer would have made a 2.35:1 monitoring solution by now though.

[Kevin Halverson]

Using a Panasonic anamorphic adapter with a native 1.778:1 imager will yield very close to a 2.35:1 aspect ratio (1.778 x 1.33 = 2.355:1). Not being a FCP person, I can't comment directly on the behavior of this NLE, but if the setup options allow for user selection of PAR (pixel aspect ratio) then simply set it based upon your intended final output format.

For SD NTSC projects (normal .88 PAR) then charge to a total PAR of 1.496.

For a native 16:9 project (assuming square pixels) then a PAR of 1.33 is correct.

SD PAL projects have differ as their native PAR is closer to square.

You shouldn't have to "render" the footage if the NLE is PAR aware. Just change the attribute when the footage is imported and it should behave properly from that point forward. Monitoring from within the NLE should be correct as long as it is truly PAR aware (Adobe products handle this properly without and issues).