[Frameworks] Super 8 Frame Size question

David Tetzlaff djtet53 at gmail.com
Fri Oct 7 08:30:24 CDT 2011


> Could you explain a few term to me that you used in your message, this would be extremely helpful for future reference. Although I got the gist of there meanings some clarification would be awesome:
> 
> Vertical Masking 

Dunno if that's a standard term, just my description for something that blocks off the top and/or bottom of the full recorded frame. In actual film projection, you never see the full frame, as the gate masks the edges of the image. On theatrical projectors, the masks are adjustable to fit different wide-screen aspect ratios. All CRT monitors overscanned, and had masks around the whole image. These varied from set to set, resulting in the concept of 'image safe area'.


> Pull Down

The process by which 24fps film gets converted to 29.97fps NTSC video. It's pretty complicated and confusing, so for a full explanation go to Google or Wikpedia. Basically, thinking 60 fields instead of 30 frames, each film frame gets copied to x number of video fields, either 2 or 3, in a regular pattern that makes the math come out right. None of the fields are blended but, some of the video frames have an upper field from one film frame, and a lower field from the next. 

> Do you have any advice on what presets I should change for best quality in regards to using super 8/ square pixel footage?

I'm not THAT much of an expert on compression. I just know more than most people on Frameworks.

A few things I've concluded:

Regardless of what format your original is in, it's usually a good idea to bump it to ProRes for editing, especially if you're doing things that involve rendering (like speed changes, supers, etc.). This will preserve the best quality throughout the edit process and yield the best overall master for duping off copies in a variety of formats.

Modern computers are powerful enough that using the best conversion settings doesn't add that much time to the process, which still takes a long enough time that you're going to go away and do something else anyway. So you might as well use the best settings, which may or may not help, but almost certainly won't hurt.

> For example, if I'm using a DV-Pal kind of set up should I change field dominance (I kind of remember this using streamclip).

Field dominance is determined by the codec. 

The 'trick' in doing the conversion is changing the settings in Compressor to the highest quality, which isn't all that straightforward...



More information about the FrameWorks mailing list