Does a file-based workflow make it easy?

Just recently, I have been to some meetings where the discussion centered on the impact of the many changes — technical and business — that are happening in the media and entertainment industry. Much talk focused on changes in workflow, specifically the migration from videotape to files.

There are two aspects to working with files: One is internal to a facility, but the other is around the exchange of files with business partners. A bespoke solution can work in-house, but for content exchange, a common format must be agreed upon.

Commercials have long been delivered as files, but they are very short duration, so small files and a well-established workflow surrounds the aggregation from agencies and delivery to broadcast playout facilities.

Of course, programs exist as files within the NLE; it’s just that they’re rendered out to tape for delivery. In the case of HD programming, HDCAM SR is popular because the number of audio tracks allows for surround-sound delivery. (Who remembers the old 8-track tapes used to supplement Digital Betacam?)

File wrappers have no such limitations. It is perfectly feasible to deliver a program with two languages, in 7.1 and stereo, full mix, and music and effects, and still add a narrator track and audio description.

In many ways, files seem an ideal solution. They can be processed with generic IT platforms, and they provide an efficient and low-cost way to handle content.

File-based production has liberated users from the constraints of the VTR. The cost and size restricted their use to the post house and occasionally on set. Now that cameras record to cards, a simple card reader is all that is needed to download a file to a laptop to start shot logging and selection, or even editing.

But, it is never that simple. Many of the issues center on file exchange. A tape is a tape. You can switch the audio tracks around, but given the correct format deck, anyone can play a tape.

Files have options, many options, and some of the complaints I heard were around acquisition. Cameras use all flavors of MPEG-2 and MPEG-4, plus uncompressed video and raw sensor data. Several wrappers can be used for the video and audio files, and then there is metadata. The post house must transcode all the many formats that may have been used during the shoot and end up with one codec to be used for the final master. When a shoot is properly planned, and there is a digital technician on set, things can go smoothly. But not all productions are well-planned, and that is when codec chaos starts.

One consequence of file-based capture is a tendency to leave cameras running (it’s free!) and capture everything. Once the VTR is removed from the workflow, the perception is that the workflows that replace tape are free. The cries I heard from facility managers were that “storage is not free” and “ingest is not free.” Apart from the cost of storage, there is also the time to log the hours and hours of material.

Another comment was that you couldn’t insert edit into a file. If a final cut was rendered to tape, a small error in the middle of an hour-long program could be fixed with a quick edit. Now the entire file must be re-rendered, adding time to the edit. With some programs edited close to transmission, this is an unwanted feature of file delivery.

The pressure to move to file-based operations is partly to drive down costs and improve productivity. The technology can only help so much; ultimately, the producer must pay for creative services, and there is no substitute for proper planning before the shoot.

David Austerberry, editor