Best capture method?
 

On a Z1, shooting HDV and doing post on a SD system,
what should give better image quality:

- capture via FW to DV avi
- capture via YUV out to an YUV uncompressed avi.

Probably the best way is to capture via FW the HDV stream, edit on HDV or an intermediate format and in the end export to "SD PAL"
But at the moment, I have to edit in PAL directly.

Thanks
Carlos

Have you thought about capturing HDV, resizing and editing the resulted SD footage? Just an idea...

What are you using to edit?

Capturing uncompressed takes an enormous amount of hard drive space.

I usually downconvert in the camera, edit in SD, then recapture the edited project in HDV. Then go from there to produce my final DVD. But if my computer were faster, I would edit in HDV from the start, perhaps with an intermediate codec if I had the hard drive space.

Vito, how are you frame-accurately capturing your HDV for onlining? (which tool?)

Avid Xpress Pro

Just select a sequence and choose batch capture. Turn off downcovert in the camera. Time code is still the same, so it recaptures it just fine.

I should add that in Xpress Pro you can switch project formats back and forth between SD and HDV. Switching to HDV before recapturing is necessary, I think. It's been a few projects since I've had to do this, so I'm not sure of this point.

Also, you can mix resolutions on the timeline. I'm working on a wedding that has mostly SD on the timeline, but a portion in HDV that I captured first to output a demo. No problems with it so far.

By the way, Douglas, I saw your seminar at WEVA two summers ago. Enjoyed it and learned a lot.

I think my two favourite parts were the very beginning where you said, basically "if you don't monitor your sound, you're a hack." The silence after that comment was deafening.

The second was your tip to go to Walmart, buy a $2.50 scrap of fake fur, and build your own Rycote. Hehe...

Vito, maybe I've got a hiccup. For some reason, I can't get frame-accurate recaps when I've got several short (timelapse) pieces. I've quit searching for the answer because drivespace is cheap, but...
I'm usually 1-2 frames off when there are lots of short 8 frame sequences.

Glad you enjoyed the sessions and thanks for the kind words.

Are you trying in Xpress as well? I thought you used Vegas.

Anyway, how are you capturing in the first place? Is each 8 frame timelapse a separate logged clip? Or are the clips captured as one long clip, and chopped up in the editing?

If the latter, you might have more luck recapturing the clip instead of recapturing from a sequence. Recapturing from the sequence means a lot of shuttling to get each little clip, giving more chance for error. If you recapture the long clips instead, it takes more drive space, but very little shuttling.

Just offline the clips (delete the media but not the master clips) and select them for rebatch instead of a sequence. The sequence will be back online after the clips are recaptured.

In Xpress, there's also a "Optimize for batch speed" setting in the capture settings that lets the camera/deck continue to roll across adjoining clips. That might help.

My last suggestion would be to lengthen the preroll in the capture tool, and make sure your deck configuration is showing the right deck or camera as opposed to "generic dv device."

I don't know if any of these suggestions would help in Vegas, cause I don't know it...

Ciao,
Vito

Yes, Express and FCP, and of course, Vegas. I'm recapturing sequences, I guess I never considered recapturing entire clips, simply because sequences are well...what the app *should* capture accurately, but none of the apps do. The T/C in HDV isn't the same as in DV or other formats, but if you're going from absolute time/header time, then of course it should accurately function. I've just not gone that route. I'll give it a go.
My knowledge of Vegas is granular; I don't know Avid X nearly as deeply as Vegas.

Amen to that, brother. Maybe I've just been lucky so far.

But I'm not sure it's the app that's the problem. It's more likely the deck or camera. I think frame accurate control in DV over firewire is difficult to achieve.

I wasn't aware of that...does that mean the camera treats the timecode on tape differently when downconvert is on? I always thought it would be exactly the same.

Hope recapturing clips works for you...

I still have to edit with a TARGA 3000 + Premiere 6.5 + Win2K and monitor with a SONY PVM-20L4.

The FW port on the Targa never worked properly (thanks Pinnacle!), so I have to capture via YUV or SDI and use as a codec the DV or YUV uncompressed.

With YUV uncompressed I get a cristal image.
With DV codec I can notice the artifacts, so for me is not an option.
Sometimes I capture in MPEG-I frame at 50Mbps. Good image, small files but a large number of crashes on Premiere.
So I work 95% with YUV (around 72G / hour)

I can't use the TARGA 3000 to capture via FW but I can capture a regular DV stream using a FW port from the PC.

The main question is about the image quality from the Z1.

During shooting, the YUV out is uncompressed and the DV out is compressed (in SD PAL). But in PLAY mode, the video signal recorded on tape is compressed in HDV, so what should give me a better picture quality ?:

- the DV out and captured in a DV codec
- the YUV out and captured in YUV mode.
(both in PAL)

--------------------------------------------------------------------------
About frame accurate recaps.

I have a large number of years editing online and the last time I dit NOT have frame accuracy was with good old U-matic - more than 15 years ago.
Since that age, frame accurate is not an option.

When I capture from Betacam (SP or Digital via RS-422 9-pin) I ALWAYS get frame accurate in's and out's.
For large projects I capture from Betacam to DV, edit and then re-capture the used clips in YUV.
NEVER a frame lost, obviously!

With the Z1 and the FW output I've had a terrible nightmare.
I've tryed a couple of external boxes to extract a 9-pin signal from the FW cable and control the camera via 9-pin (my only option)
ALWAYS with several frames out of the place.
I really don't know why.
Is the Timecode inside the FW a "different" signal from the Betacam timecode?
Is it because of the Long GOP of the HDV ?

For some reason SONY does not have made an HDV player with a 9-pin port for remote!...

I also would love to capture and recapture from HDV with frame-accurate.
Is it really possible?

I'm loving the day that I will upgrade my edit station to a full HD without all this mess.
Is it a MATROX AXIO ??
What do you feel?