The Pitfalls of HD Editing For a New Documentary

Therapy Studios, in cooperation with Ptolemaic Productions and director Barry Ptolemy, is working on a documentary called Transcendent Man, which chronicles the career and visionary ideas of Ray Kurzweill and his thoughts on the future of information technology. [Among a number of inventions, Kurzweill was the principal developer of the first print-to-speech reading machine for the blind, the first CCD flatbed scanner, and the first electronic musical instrument capable of recreating the sound of a grand piano and other orchestral instruments.]
Now in its third month of production, the documentary is being shot over the course of a year with the Panasonic HVX200 P2 solid-state camera and edited at Therapy on several Avid Media composers. Finishing and color-correction will be completed using the facility’s Quantel iQ system. In fact, Therapy is executive-producing the doc in addition to doing all the post-production.
Q: You’ve put together an interesting workflow in which images are shot in HD but posted in SD resolution. Why?
A: We prefer to do our offline editing in standard definition, so we did a bunch of different tests, and realized shooting the project 24p native, not a flagged-frame 24p, was the best way to go. This helped us in a few ways. First, we can play the actual 24p native HD media (that the camera records) on-the-fly in an SD 23.98 project, which allows us to work faster with our dual processor CPUs and monitor everything in SD. Second, it gave us 20 minutes of HD storage on an 8 GB card instead of 8 minutes, which is a huge difference. When it comes to finishing the project, we can convert the project back to HD, and play the uncompressed edit right out of the Avid into our finishing box. From there we will do our clean-up, compositing, color-correction, titles and mastering. Our other option would be to do a batch-import conform in the iQ, from the MXF media, and proceed from there.

Q: What’s the main challenge for this project?
A: Because we are working off the actual HD media, one challenge is the large file sizes of HD. In the past we would digitize at DV25 resolution, so your storage needs are much less. We basically are storing the media in its native DVCPro HD 720p resolution because it creates the most streamlined workflow, but it means we’ll need around 3-5 TB of storage capacity. We have a 6 TB LAN Share system, so we’ll probably have to upgrade it soon.

However, our chief concern is how we back up this media, which now exists in one central location. How do we store it in a stable and secure format without having to stack up FireWire drives? Honestly, in our experience having data backed up on a FireWire drive is not the most stable or secure way to handle data for a project like this. And we don’t want to trust our entire investment in this film to FireWire drives. So we’re trying to figure out the best way to aggregate the data. We now back up most of our eQ online projects to DLT tape, and those will go anywhere from 160 GB to 320 GB with compression.

So we might end up with about 15 or 20 of those tapes for this film, only because in our minds that’s a far more stable format to back up our data to. So, we’ll soon start backing up the date to DLT tapes, 160 GB at a time, as we move along the process and Barry continues to acquire more and more footage.

If you lose your media, it’s gone forever. That’s one of the downsides of the tapeless workflow right now. Anyone who’s working with P2 or any tapeless format will tell you the same thing.

Q: With such a long shooting process, when does the editing begin?
A: There’s a lot of footage, so at this point we’re establishing a process of logging and organizing the scenes Barry is sending in from the road. We have a couple of editors working on this as we speak. We’re expecting about 150 to 200 hours of footage before we’re done. Logistics is a big issue.

Q: What’s the biggest misconception about posting HD?
A: There are so many formats, and a lot of people that are shooting HD are not that knowledgeable about the differences between them. Often times we see people shooting in formats or modes that they don’t know they are in. Then we get the footage and a lot of those different elements can’t be mixed together cleanly. And frankly there isn’t much you can do about it after the footage has been shot, so it’s important to plan out your projects carefully beforehand. We often get projects that have elements from different sources and we end up doing a lot of frame-by-frame work trying to get images to display in a uniform way. It can be a real pain.