DV Test?

[Daniel Low]

“Does somebody have a good test that can definitively demonstrate the artifacts of DV compression?”

Yes, try pulling of a difficult chromakey one through fine detail with lots of highlights.

“To explain- we’re selling our Avid Express DV system and switching over to Apple”

Good move.

“Along with keeping a backup of all the project and media files, we want to render out masters of the different films.

I’ve pushed for doing this with some sort of lossless codec like lagarith, or quicktime animation at 100%, or straight uncompressed- but we’ve got someone telling us that it will be no better quality than DV stream, and lossless (even with a good codec) will be a waste of space. His argument is that *everything* in Avid Express DV is internally stored as a DV stream anyway- so an export to a DV codec is 1:1 ”

If everything in the Avid is stored as 25Mb/s I frame the saving in another codec won’t improved the quality obviously, and will certainly need to be re-rendered when being brought back into a project in FCP.

Some would say to simply lay off all your media files back to DV tape or alternatively use a lossy but high quality quality codec like PhotoJPEG at above 75% quality. It’s fast and easily transportable.

“ I’m not totally convinced, as I’d think that effects would be done in the uncompressed memory buffer and then rendering out would re-compress this an additional time.”

Possibly, I’m not sure about the Avid, but I doubt it. Moving around uncompressed data is expensive for any system to perform. Low-end systems aren’t built to do this.
To test this I’d take a multi-layered project and render it out to DV and then take the same project and render it out to Animation, PhotoJPEG at 100% or uncompressed AVI and see if there’s any quality difference.

Let us know your findings

__________________________________________________________________
Please post back saying what solved your problem. It could help others, and saying ‘thanks’ is free!