DVD-Cams: Convenience At A Price

Submitted by Robert Alstead on Tue, 05/18/2004 - 16:00

On the face of it, buying a camcorder that records to a DVD rather than miniDV has several advantages.

The DVD-camcorder electronics may not be as highly specced as miniDV cams in the same price range, but here is a camera that is apparently so easy and convenient to use.

You can simply put a DVD disc into the camera, shoot some footage and then whip it out and play it back on your DVD player.

It's digital so it's easy to archive and duplicate footage, and you don't have to worry about the wear and tear on the camcorder heads that you have when using a mini-DV camera as a VCR.

But, and it's quite a big but, your footage is now encoded in MPEG-2.

MPEG-2 is a compression format that was not designed for editing. It uses a much more ruthless compression scheme than the miniDV format, which means that you will have a lower quality picture and some headaches if you plan on editing footage that has been captured with MPEG-2.

The picture itself may not look too bad on the television, but that's largely through the magic of the Motion Picture Experts Group, the body of engineers who formulated and gave their name to the MPEG standard.

Like the JPEG image format, MPEG files are "lossy", which means certain information from the video signal is thrown away to squeeze more footage onto a DVD.

MPEG-2 video uses sequences of GOPs (Groups of pictures). Each GOP, typically 12 frames long in PAL (or 15 frames if you are using NTSC), comprises of a combination of I-frame (or Intra-frames), P-frames (predictive frames) and B-frames (bi-directional frames).

The I-Frames are the "alpha" or "key" frames. They are the biggest frames in terms of byte size because they contain all the information acquired from the video signal.

The P-frame only contains changes between the previous I-Frame or P-frame. The B-Frame, contains changes between the previous and/or the upcoming I-Frame or P-Frame.

The GOP size and P-frame frequency can change, but a second, or 25 frames, of PAL MPEG-2 might look like:


If you are not too worried about frame accurate editing then a "GOP-accurate" MPEG editor will allow you to cut where there is an I-Frame (a GOP boundary). But if you want to make a more polished edit then you will find yourself having to cut on one of the in-between B or P frames and re-rendering each GOP, leading to degradation of your image.

If you want to edit your footage then the initial benefits of the DVD-camcorder in terms of convenience and ease of use begin to look less clear. From an editing point of view it is less convenient.

DVD is still an obvious way to distribute your edited projects, but it can't compete with miniDV (or higher formats, if you can afford them) for recording footage.