Michael et al,

I seem to have run into a problem capturing 23.976p footage from a DV50 tape over firewire in a software only MC v 2.5.3. The DV50 tapes are dubs from a DBeta which was a downconvert from a 1080p/23.976 HD SR master (in turn created in online from telecine film footage).

A couple of issues:

1. There doesn't seem to be a 1% slow down switch in the capture tool as there is when capturing analog or SDI (in DNA systems). The captured audio is thus progressively out of sync as the clip length increases.
2. The captured video seems to retain motion in ALL frames (no identical frames) regardless of the pullin setting. I tried capturing clips with 4 different pullin settings (A-D) and the resulting footage is all the same; interlaced frames appear at the same timecodes on all four clips!

Some questions:

1. I assume that 23.976p DV50 can be captured over firewire just like DV25. Is this correct?
2. Is it possible that the cadence was broken in the DBeta -> DV50 transfer by not properly ganging TC?
3. Assuming the above is true, shouldn't I still be able to capture 23p by finding the NEW A frame TC in the DV50 tape?
   
4. If the answer to 1 is yes, is there any reason the DV50 capture would be less realiable than a DBeta capture over SDI (embedded audio) using a Mojo SDI?
5. Does the fact that I'm seeing motion in ALL frames indicate that I'm indeed capturing 23p properly. If so, why do I seem to be capturing incorrect (interlaced) frames.
6. Is there any reason for cadence to be broken (ie improper pulldown removal) within a clip if I capture a continuous pulldown tape in 5 min junks provided I set my pullin and TC correctly (ie A on :00 logs) for each clip (assuming the issues above were not a problem!)
   

Thanks in advance,

David.

In a 23.976 project, there are no tools for +/-1% as in a true 24p. It is assumed that your sources on tape are in sync. So I would suspect that the dailies are out of sync and is not a result of the capture in Avid, but at time of transfer. The process for DV50 is the same for DV25.

If you are seeing interlace artifacts, it is most likely due to one of the dub processes getting to your DV50 dailies. A one field delay will cause errors. Try changing your "A" frames to "C" frames and see if that clears up the capture. If not, try B and D. 

Michael 

Michael,

thanks for your reply. Indeed, I assumed the problem was upon transfer. The DV50 capture is actually of a master rather than dailies but that is irrelevant; however, I suspect the transfer was done via YUV with LTC timecode fed from the DBeta source into the DV50 deck. I understand that SDI transfers are +-0 frames while LTC/video gen is +1 frame wrt SDI. So, if the DV50 deck is fed SDI as well as LTC, the deck will lock +1 frame off and the cadence is shot.

But... as I mentioned, I tried all possible pullin options with no apparent correction. The only difference between the different captures is that each had interlacing in frame pairs one frame ahead from the other. This is what puzzles me.

Assuming the transfer was done in regen rather than locked to the source TC, shouldn't I be able to reconstruct the 23p by hitting the right pullin frame (after a little trial and error) regardless of which one it happens to be? Also, why am I not seeing duplicated frames if I'm off in the pullin?

24p still puzzles me sometimes...

Thanks again, David.

 There is a 5th relationship that exists between 24 and 30 - if you tried all 4 relationships; A, B, C, and D for timecode ending in :00 try this one:

Keep the pullin as "A" but change the timecode to be -1 frame ending in xx:29. This will create the
"X" frame relationship.

Michael