What’s the ideal amount of in-field recording time? While some might say you can never have too much, for workflow reasons smaller amounts of footage can be much more manageable. So how much capacity is enough?
30 minute tape loads have been the standard since the BetacamSP days. 11 minutes of per roll (wether 400′ of 16mm, or 1000′ of 35mm). has always been pretty standard in the film world. 60 minutes was the run time of MiniDV.
I don’t think anyone would argue that these runtimes are optimal, as changing reels, and tapes has always been a hassle, especially if you run out in the middle of the take. They are simply limited by the considerations of physical space. There’s only so much tape that can be packed into a cassette, there’s only so much film that can fit into a camera.
With digital acquisition, the equation changes. With the meteoric rise in memory capacity, that rapidly exceeds the size of the codecs recorded, it’s certainly possible to shoot all day without reloading. In some cases, it may be possible to shoot your entire project without reloading. Is this necessarily a good thing?
I think that for most purposes, the optimal media recording time would be enough to finish the day’s shooting without reloading. This allows the offloading of your data to hard disks to be done at the relatively peaceful time at the end of the day, rather than in the heat of the moment during a shoot, reducing the chance of accidents, and wasted down time. While of course, in terms of actual minutes, a full day’s shoot greatly depends on your project, documentaries will need many more hours of recording than a scripted drama for example, but as a workflow it has reasonable tradeoffs.
There are some who will disagree, and not want to trust their whole day’s shoot to one memory card. What if it fails? you just don’t lose a tape’s worth, you lose the whole day! While I can see the logic of this, I still believe my data is more secure on one piece of solid state media, rather than on one mechanical hard drive, especially if you’re constantly copying and deleting files all day, the room for error is just compounded. Even if you’re backing up to a mirrored RAID, I would trust the solid state more. The only way that I would think hard disks would be more secure than solid state, is if you backed up to two separate drives, that were stored in different locations, however this doubles the work of the data wrangler.
Why not take the opposite approach? If you’re shooting on AVCHD for example (please don’t though), you can most likely shoot an entire feature film’s worth of material on the highest capacity SDXC cards. Why not wait to offload, until you’ve finished the shoot entirely? In the end it’s all going to be backed up to the same hard disk anyway, this way you just do it once, rather than in multiple layoffs.
Maybe for short shoots in difficult locations, this way makes sense. However, I’ve never met a director or cinematographer, who hasn’t been able to pass up the chance at looking at the footage he’s shot the previous day, either for continuity purposes, or just emotional security. This is much easier to do off camera. In addition, while I still think that the solid state card is less of a concern as a single point of failure than a hard disk, over time the camera itself becomes a vulnerability. Usually the camera gets put in all sorts of interesting and dangerous positions during a shoot, and if it falls off a jib down a cliff, or slips off the tripod into the ocean, you’re sunk. If you do have the capacity to hold all your footage in the camera, I still think you should, only back up that footage to hard disk as well, at the end of the day.
Advances in capacity and speed of solid state media are exceeding the advances in recording codecs, which allows for much different workflows than what we’re used to. What are your experiences?