Capturing expectations w/CineForm
 

I had to recapture a clip from my FX1 because the initial capture was very noisy (not sure why, but the top half of the initial capture had weird jumpy overlay noise stuff on it, when viewed outside of PPro). The second capture went well, looks good, etc. However, the second capture is several frames longer than the initial capture. I specified the in/out points directly and used the same timecode values for the capture.

Is it expected that, when specifying in/out points, that repeated captures might have varying lengths?

I may retry the capture to a different file this evening--perhaps the first capture was just hosed.

Thanks,
Matt

What kind for firewire capture do you have on the computer side?
I was having this kind of trouble until I went out and got a generic firewire pci card that used the oxford chipset.
Before that I was using the firewire port on creative sound card which I learned from cineform is known to have problems.

I highly recommend using Scene Detect on HDV. Otherwise, you will try to cut where there is no I-Frame. Let the software do the work of sorting out the clips for you.

I don't believe it is possible to be frame accurate. You should be able to get withing 15 frames easily enough.

And I use the Audigy 2 firewire port with no problems. But I also capture using HDLink and then run through the clips to see what is worth keeping and what is not. Then I transcode to the Cineform AVI just the clips worth saving.

That is the same card I was using. I had to open a ticket with cineform to solve the issue (Creative Audigy ZS I think). I was getting artifacts and blockiness in my captured material that did not show up on tape. Now I have an Adaptec card and no more artifacts on the captures.

I have a ticket open right now with a problem I have getting Marked IN/OUT point captures to work and I am wondering how the I frame mpeg issue mentioned abouve might impact that. Scene dectect works when I press the TAPE button in the prem pro capture window but that is the only reliable capture method that I can currently use. Otherwise I end up with nothing, no file at all after the capture "ends".

Steve,
Thanks. I generally do use scene detect. The clip I'm talking about was just one clip (about 6 minutes). What I'm concerned about is if I have to reconstruct my project, the frames would be off a bit from capture to capture. This is, of course, not a problem with DV because DV is frame accurate. It seems that HDV is not so specific, however. I imagine burning a finished project back out to HDV tape would have similar issues on recapture. I just need to be aware of the issue when/if I have to recapture/rebuild the project--I've had to do that a couple times when I discovered a DVD error after I had already cleaned up my hard drives. :( Damn QA department...

I guess for logging, I'd need to use the camera's footage instead of logging from the capture...

Thanks,
Matt

Hi Matt, it should not matter, because the timecode matches up, no matter if it gets 20 extra frames or 25, sometimes it gives you more or less, but the timecode stays the same. In essence if your clip was messed up, but the timecode was intact, you could make your project, edit it then you could recapture the offending segment, do "unlink media" from the bad clip and then relink to the good clip, it should just put it in its place. Don't know if that helps or not.

Mike

>>>I had to recapture a clip from my FX1 because the initial capture was very noisy (not sure why, but the top half of the initial capture had weird jumpy overlay noise stuff on it, when viewed outside of PPro).

One of my two computers does exactly that, regularly. It has a slower processor and slower hard-drive than my other machine, and I assume its one of those factors that's the cause of the poor capture. Both machines have the same type of firewire interface.