Editing HDV with Pixlet?

[Steve Mullen]

Some think Pixlet offers a way to edit HDV. After demuxing the MPEG-2 would be "converted" to Pixlet. Then after editing the Pixlet would be "converted" back to MPEG-2 and recorded to HDV.

FCP will import Pixlet. You can define a Sequence to be Pixlet. This means that Pixlet clips will play -- given a 1GHz G4 -- at full-speed.

Of course, FX will not be RT. Some may not care.

But, because of the way FCP exports -- using Pixlet is not optimal.

Here's why.

1) DURING CONVERSION TO PIXLET: MPEG2 > uncompressed > Pixlet

2) DURING EXPORT: Pixlet > uncompressed for FX including color correction -- then back to Pixlet

3) DURING CONVERSION TO MPEG-2: Pixlet > uncompressed > MPEG-2.

Apple has already stated that Pixlet is a lossey codec. We know it blurs fine details. If used at the END of a production, as Apple intends, this is no huge problem. Neither is the very long compression times. The goal is HD playback.

But, as you can see -- editing with Pixlet introduces 2 or 3 decompression-recompression cycles. Each recompression into Pixlet takes a very long time. And, of course, MPEG-2 decoding and encoding are not fast either.

No RT FX. Loss of quality. Slow performace.