I am working on capturing a batch of miniDV tapes for documentary purposes, creating Quicktime movie files containing DV-NTSC video. The resulting DV-NTSC video will of course be interlaced, with anamorphic pixels per the Rec. 601 standard. I will eventually be passing a set of Quicktime movie files of this material to an unknown future editor, possibly one who has no experience with legacy, interlaced, anamorphic, standard definition video.
I see unprocessed interlacing in standard definition source material on videos all the time, on YouTube, in documentaries, etc. And I hear people saying, “it’s retro, it’s cool”, which I dispise. I don't want to let that happen to the DV video that comes from me.
Therefore, I intend to create a set of Quicktime movie files with the DV deinterlaced with doubled frame rate, using Telestream's Version 22.214.171.12458 (which is not supported anymore, but it still works for me.) I would also like to do the appropriate cropping (8 pixels off left and right sides), and scaling from 720x480 to 640x480 at the same time. Those will be the Quicktime movie files that I pass to a future editor. This way, a naive future editor can't screw up the presentation of the DV video.
But can it still be DV? Is there such a thing as DV as progressive video, with square pixels, 640x480? Or do I have to encode the video to something else?
+ Reply to Thread
Results 1 to 9 of 9
Well you can call it what you like but AFAIK DV is a very strict format.
It may have some of the qualities of DV - bitrate for video and audio - but it will not be DV. It will just be Digital Video.
DV requires 720x480. You won't be allowed to encode it. The deinterlacing, cropping and resizing is a stupid idea anyway, if the intent is to set it aside for editing later. And reencoding to DV AVI incurs more quality degradation. Oddly enough, while testing, I discovered 59.94fps was allowed. This is nothing you couldn't have figured out on your own.
Why would you assume that a future editor would have less knowledge, or not have a better way of handling the material than you do?
videoeditors cannot do anything worthy if you decide to deinterace it within NLE, resize it, upscale it etc. They will screw these procedures up.
As soon as you start mix old DV footage and HD footage, things would fail or not be optimal.
What NLE can do is to resize or upscale of progressive footage, no problems there, so you are right, you can settle down and bob DV footage to double frame rate, OUTSIDE of NLE using something better, like QTGMC (or whatever you prefer and it is good). You do not even have to resize. Just get bob 720x480 59.94fps, every pixel is handy. Then upon loading into NLE for editing or mixing with HD 59.94 footage you just interpret it as 4:3 (or 16:9 if it is) with NTSC DV aspect ratio.
Depending on how much of that footage we are talking about, you can even bob it into a lossless, because we are talking just SD, not HD. But anyway, if you encode some low latency ( short GOP, easy settings, like fast decode) high quality H264, that would work also.
If you intend to preprocess with the purpose of futureproofing compatible formats for other editors, use AVISynth to process, deint using something like the QTGMC process to 59.94p, while converting color sampling from 4:1:1 yv12 to 4:2:2 yuy2, probably NOT resizing (though 640x480, 960x720 or 1440x1080 might do in a pinch).
Then re-encode using a common, known-good, lossless encoder, saving NOT as a MOV but as mkv or mp4.
Thank you, _Al_ and Cornucopia. I will go in that direction. Process to 59.94p, but keep it 720x480.