MiniDV tape archive problem

rdthrush wrote on 5/4/2003, 7:35 PM

I am converting old home movie tapes (8mm/VHS) to MiniDV tapes with a Canon ZR40 camcorder. I then upload the DV tapes with the Vegas4 Capture tool to do some light editing, transitioning & titling. I expected to archive the MiniDV tapes as well as the .veg and .dar files used to produce the dvd. However, in testing this assumption, I find that recapturing these tapes (with a different tape name) does not result in identical files. I initially had the "Enable DV scene detection" enabled (Options -> Preferences -> Capture). I disabled that option and tried again and still did not get identical captures.

<http://www.thrush.com/pix/vv4-media-bin-2.jpg> has a screen dump of the media bin that shows more detail about the captures. The first clip is from the original capture. The subsequent files were captured from the same MiniDV tape and with the same ZR40 used to create the tape. Although some file sizes are identical, the files do *not* compare as identical.

There were no dropped frames in any of these captures.

Am I making some incorrect assumptions about these MiniDV tapes? Why aren't the recaptures identical?

Comments

Zorro2 wrote on 5/4/2003, 10:00 PM

I am not following you - why the need to 'recapture'?

rdthrush wrote on 5/4/2003, 10:28 PM

I don't have unlimited disk space. Before removing the 'capture' files (to free disk space), I wanted to verify that the archive tape could be 'recaptured' (reloaded) and was identical to the original file.

Jason_Abbott wrote on 5/4/2003, 10:37 PM

I've several times had Vegas recapture media for projects from the original tapes, without problem. Just right-click the event in the media pool and select recapture. I've never had to pay much attention to the .avi files so I'm not clear on the approach you're trying to take.

SonyDennis wrote on 5/5/2003, 2:10 AM

I applaude you for running this test. I would do the same thing myself before trusting a new system.

However, I don't think a binary compare is a valid test of whether the files are the "same" or not. What really matters is the audio and video data being the same. Since the video capture application probably puts in different metadata (perhaps a timestamp of when the capture was done? I don't know) the binary data probably differs each time you capture the same clip.

Try putting the "before" and "after" clips on stack tracks, and comparing frames (by muting the top track) and audio waveforms, to make sure they are the same. If you wanted to get fancy, you could also use the "Difference" transfer mode on the video, with a contrast boost to accentuate any differences, and a phase-inverted audio track render (should produce silence) for the audio test. A tip if you're doing it this way: Shift+Up arrow zooms audio waveforms.

///d@

rdthrush wrote on 5/5/2003, 7:19 AM

Thanks for the excellent tips. I think my initial problems are understood. The video and audio tracks compare per your suggestions for at least the first 1hr 2min. There are some differences beyond that which I expect may be due to camera (Canon ZR40) inconsistencies at the very end of the tape. Since I don't use the last minute or so that isn't an issue for me.

Apparently, a raw binary comparison is not appropriate for comparing captures. Is there a tool that can be used to compare the data bits encapsulated in the DV headers?

The initial problem was that a recapture (w/ DV scene detection enabled) resulted in differing start/stop timecodes and lengths. I then disabled scene detection and recaptured the same tape to several single 1hr (~14GB) files which were quite different when compared as binary files.

BillyBoy wrote on 5/5/2003, 9:34 AM

You can also try this simple and quick Rube Goldberg method.

Start with two video tracks. Then drag and drop the video portion only to the timeline for both original and the second copy.

Obviously the two clips should appear the same length on the timeline. Now to randomly check if they are the same throughout grab a hold of the opacity control at the very top edge of the top track, drag slowly down and the video on the second track should bleed through at the same place beneath. Check as many spots as you feel necessary to confirm they are the same moving the opacity slider up and down. You may see very slight varations in which frame shows where comparing the top and bottom tracks, just jiggle one track or the other sideways a fracion and they should line up, unless of course you're a fanatic and go ape over a 30th of a second or something like that. <wink>