Anyone capturing to Apple ProRes 422 codec format?

[Jon Braeley]

Now that FCP 2 has a little more support for the V1, how many of you are using or recommending capturing to the Apple ProRes 422 codec format from tape?
I hear that this will make for a smooth workflow and no rendering waits, especially useful if you go back and forth in motion etc. This codec is in 4:2:2, does this have any effect on the quality?
Thanks everyone.

[Lee Berger]

I tested the Capture to ProRes 422 and it worked. Keep in mind that it's a capture now workflow. You cannot batch capture. You play the tape and start the capture. This means that you cannot recapture if you loose your media. I think a better workflow would be to capture in HDV and then use Compressor to transcode to Prores. Also keep in mind that with FCP Studio 2 you can render in ProRess HDV and XDCam HD sequences.

After capturing directly to ProRes 422 and testing on a ProRes sequence I still had green bars over dissolves. I'm using a Dual 2.0 G5, 4 Gigs of Memory, and XServe RAID. More realtime playback might be possible with a newer machine. Anybody tested with a MacPro?

[Robert Ducon]

I tested with a Mac Pro and an HV20 via the HDMI connector - worked great. The computer and camera don't 'talk' to each other - the camera just output raw video, the computer just captures what ever it sees coming down the pipe.

The Mac Pro was flawless - ProRes is so efficient (read: small file data rates/file sizes) that I could captured using a single 250GB internal HD. For safety, I set up a software RAID using two 250GB hard drives for a total of 500GB.

I'd expect the V1U/FX7 to work the same.

[Lee Berger]

Originally Posted by Robert Ducon

The computer and camera don't 'talk' to each other - the camera just output raw video, the computer just captures what ever it sees coming down the pipe.

I assume you used the Blackmagic Intensity Card to capture via HDMI. I believe you can use firewire for control.

[Jon Braeley]

I am at the start of a new project and started to captured in regular 108060i HDV - maybe 2 hrs worth so far. I may switch to ProsRes.
Coincidentally, in Miami today there is an Apple event titled - Apple Pro Res Video Workflows Seminar, and I will be going to this.
If anything enlightening turns up at this I will post back.
Thanks.

[Jon Braeley]

Pro Res is certainly an impressive codec. The only way I can see an advantage with the V1, is if you are ingesting from the HDMI via say, a Blackmagic Intensity card. The you get uncompressed video and capture into the ProRes codec.
Would this be correct? Still some gray area here. anyone using the intensity card with the HDMI of the V1?

[Steve Mullen]

Originally Posted by Jon Braeley

Now that FCP 2 has a little more support for the V1, how many of you are using or recommending capturing to the Apple ProRes 422 codec format from tape?
I hear that this will make for a smooth workflow and no rendering waits, especially useful if you go back and forth in motion etc. This codec is in 4:2:2, does this have any effect on the quality?
Thanks everyone.

Given that one typically applies Motion effects to VERY short segments of video, it is far better to edit native HDV and then render only these tiny segments. FCP 6 automatically -- if you ask it to -- renders to ProRes and not HDV.

Bottom-line there is no advantages and many disadvantages to capturing to ProRes.

[Steve Mullen]

Originally Posted by Jon Braeley

Pro Res is certainly an impressive codec.

I discovered something by accident. I exported to Compressor 1920x1080 XDCAM EX from FCP to HD MPEG-2. I also exported 1920x1080 XDCAM EX from iMovie 08 to ProRes 422. Compressor imported the ProRes and output HD MPEG-2.

The iMovie output was a tiny tiny bit less detailed. Was this caused by iMovie or ProRes? Repeated the test using 8-bit uncompressed from iMovie. Now there was no difference between FCP and iMovie.

This test convinced me that there is a quality benefit to uncompressed. However, I need to run the test using ProRes HQ. I suspect it will look as good as uncompressed.

If so, then there is a real-world advantage to using HQ.

[Jon Braeley]

The travelling Apple Pro Res seminar was in town last week. It is certainly impressive when they demonstrate using the split screen, where ProRes footage is on one side and uncompressed on the other, and you cannot spot the difference.

[Steve Mullen]

Originally Posted by Jon Braeley

The travelling Apple Pro Res seminar was in town last week. It is certainly impressive when they demonstrate using the split screen, where ProRes footage is on one side and uncompressed on the other, and you cannot spot the difference.

Do you know if they used ProRes 422 HQ?

When I re-ran my test with HQ -- I could not tell HQ from UC.

So I'm using HQ in the future.

[Paul Frederick]

A great way to test "honest to God differences" is to line them up in AE, and apply the "difference" transfer mode to the one on the top. Any pixels that are not "perfect" will appear. If there are differences at the pixel level, you'll see them! Start by lining up the same clip, one over the other, set the top one to "difference" and you'll see just a black screen, then delete the top one and replace with the "pro-res" one and you should see some pixel's pop out.... (you're clips need to be lined up EXACTLY to the frame or it won't work.) The more compressed images should show more pixels.

Also you may need to apply an adjustment layer over both and boost the brightness in order to see the pixels better.

It works well and I'd love to hear your results!

[Fred Olen Ray]

We have undertaken the recapture of an entire feature film's worth of footage (26 tapes) after struggling with the previously adhered to work flow. We had an outside source capture our footage... their work flow was to capture at 108060i HDV and the put the 24pA files into a ProRes 23.98 timeline. Bad move. The pulldown was not really removed and the footage was VERY aliased looking in any lower light scenes. Still captures were disappointing.

We then tried a test of recapturing the same footage straight from the camera using the ProRes easy set-up which has been a life saver. The footage is coming in at 23.98, no pull-down, no stutter, and the image quality has righted itself and now all is beautiful. And no HDV files to deal with at all. Only ProRes... but it's slow... takes about 2 hours for a 1 hour tape... but worth every second of it.

Yeah, there's no batching, or real logging for that matter, but the files DO come in sub-clipped and all we had to do was re-name them appropriately.

Now let's see how the DaVinci handles the footage!

[Jon Braeley]

Steve, Yes I was told they used HQ, but this was for comparison against uncompressed HD. I understand that ProRes (SQ? Standard for lower bit-rate) compares nicely with HDV footage.
I am going to make my own test. I logged the first 3 hrs of 20 hour shoot in 108060i. I will now log using ProRes HQ and standard.

For this project, ProRes will be the right choice I feel, as I am using quite a lot of outside graphics, titles (soem in motion) and still images from historical archives - all with variable quality. I really do not want to face quality difficulties and long render times in HDV.

[Jon Braeley]

The footage is coming in at 23.98, no pull-down, no stutter

Fred, just saw your post. Are you saying that regular 108060i can be ingested straight to ProRes at 24p?

That's interesting!

[Fred Olen Ray]

Originally Posted by Jon Braeley

The footage is coming in at 23.98, no pull-down, no stutter

Fred, just saw your post. Are you saying that regular 108060i can be ingested straight to ProRes at 24p?

That's interesting!

It was shot initially on the V1U in the 24pA mode (which then gets spit out at 60i with a 3:2 pull-down), but capturing it via the ProRes codec removed the pull-down and delivered a gorgeous 23.98 quicktime file! No further converting needed!