A lot of the time I see people knocking HDV (and obviously I am aware of its limitations) but occasionally it feels that some folks bash it because HDV has a consumerish stink about it these days, and forget what a real marvel it is (was).
I personally have never had any 'nightmares' at acquisition, ingest or post but like any media, I have also learned what it will tolerate and what it will not (heck, even people using 35 have to do the same right?)
So, I was really curious at this stage of my project whether converting (thousands) of HDV clips to ProRes 422 *before* I grade was really going to be a worthwhile exercise.
On the attached web-page you will see a clip that I really pushed in Color to bring out the artifacting. I graded the top clip as a native HDV file but converted the second one to ProRes 422 HQ (interlaced, top field selected, 4:4:4 sampling unchecked) put it in FCP then did the grade.
http://web.mac.com/pramsbottom/Site/Color_tests.html
Note, right-click to download the full 1080 file
ProRes didn't solve everything but it is definitely an improvement, and I'll certainly do an batch convert of my HDV stuff based on these results.
The only odd thing I noticed was that the exported ProRes file was ever so slightly smaller that the 1080 canvas in FCP, and needed a tiny stretch to get it back to full frame.
File size-wise, the HDV original was about 9MB, the ProRes conversion was 50MB+, so increasing my storage needs 5-fold is no small consideration but I think it is worth it.
If I had everything on HDV tape I might have considered recapturing to ProRes to see whether the conversion directly off tape was better than what could be achieved running these files through compressor but as a lot of these files are now only available to me as QTs and because I have so much material + the fact the transcode from tape process doesn't sound 100% baked right now, I will probably not bother.
I hope this is helpful to folks.
