XDCAM HD Workflow

hello group,
is anyone having major issues with using XDCAM HD aside from "FCP 7.0.3 frequent crashes"? i'm getting ready to start 6 one hour episodes shot on the ex3. i am thinking about using the xdcam codec and seq rather than my usual prores because of storage space. we'll be working from a terrablock which uses fiber channels to communicate to the edit bays. final delivery will be 1080i (1920 x 1080) @ 59.94. any golden nuggets of wisdom will be helpful as this is a new format for me to use. i am particularly interested in workflow. we will have assistant editors to ingest the footage, so i'd like to give them the best approach. we also will have an editor doing the online, but i think i want to avoid doing any upresing for the online because it seems a bit hairy with file naming and relinking, etc.
fcp 7.0.3
osx 10.5.8
2 x 3.2 ghz quad core power macs w/ 8-12 gb RAM
magic bullet looks
sapphire

thanks!

adam

See the thread 7.0.3 crashes. They seem to be having a lot of problems with XDCAM.

All the best,

Tom

>See the thread 7.0.3 crashes. They seem to be having a lot of problems with XDCAM.

thanks tom.
i have been reading that thread, but i wanted to know if anyone else has experience using xdcam hd besides that poor, poor editor

is xdcam hd just terrible to use?? or are we seeing one unfortunate case here?

Use the XDCAM footage, but don't use an XDCAM sequence. Use a ProRes sequence. Don't use the ProRes RENDER setting...use an ACTUAL ProRes sequence. You will get a green render bar, but you can work, and it is FAR more stable than XDCAM.


www.shanerosseditor.com

Listen to THE EDIT BAY Podcast on iTunes
[itunes.apple.com]

Adam,

I have been following the other discussion...

It seems that the problem may be related to using different frame sizes, frame rates, many tracks of video, and certain filters/effects. The fact that much of the footage is XDCAM EX may be a contributing factor, as well as FCP RT and preference settings, and HW configurations.

I would suggest that you create some realistic sample scenarios and see how things work (testing, testing, testing). You may not run into the same issues as Mike and JW (who reported their issues in the other ongoing discussion). For example, Ben King had to put 21 tracks of XDCAM footage in a sequence to finally get FCP to crash...

The best thing to do may be to create a complete one-hour sample piece, doing all of the things you normally would, in order to validate your workflow. The edit can just be an assembly edit (i.e., you don't need to do a "real" edit for content), unless you have the time to use the first one-hour piece as a testbed case.


-Dave

awesome input guys, thanks. all of my footage will be from the ex3. we will probably have 5-7 layers of video max. i'm going to create an apple prores sequence and run some tests. i'll keep you posted on the progress.

adam

acronan Wrote:
-------------------------------------------------------
> awesome input guys, thanks. all of my footage
> will be from the ex3. we will probably have 5-7
> layers of video max. i'm going to create an apple
> prores sequence and run some tests. i'll keep you
> posted on the progress.
>
> adam

Yes, do what Shane said. Adding to that, I would use a ProRes 422 or ProRes 422 (HQ) sequence preset but with a full raster frame size (1920 x 1080), rather than a 1440 x 1080 frame size (as your text and other items will look better on export).


-Dave

HQ is too much for the 8-bit XDCAM footage. It was designed for 10-bit footage, like AVCIntra, RED. Zero gain for using that setting over ProRes 422.

And OF COURSE use a Full Raster setting...why would you use anamorphic? That's only for HDV converted to ProRes via firewire.

Crashing due to mixing formats and frame rates, while using a GOP sequence setting like XDCAM or HDV is VERY common. Avoid GOP format sequence settings at all cost.


www.shanerosseditor.com

Listen to THE EDIT BAY Podcast on iTunes
[itunes.apple.com]

Quote

Crashing due to mixing formats and frame rates, while using a GOP sequence setting like XDCAM or HDV is VERY common. Avoid GOP format sequence settings at all cost.

It is a good general rule of thumb. But personally I've never encountered any of the issues that seem to rear their heads with this issue unless asking unfeasibly crazy things of FCP. I think people assume way too much of a 32bit package that still relies on code that was officially depreciated in 2005! Don't get me wrong I prefer FCP to Avid and Premier because it actually does rock. Hopefully the next version will blow the competition away again and we can whinge that our 64bit FCS hasn't been updated for 2 years

XDCAM (or EX or HDV) workflow is however best (as Shane said) on a ProRes 422 timeline.



For instant answers to more than one hundred common FCP questions, check out the LAFCPUG FAQ Wiki here : [www.lafcpug.org]

XDCAM has given us more headaches than any other format - it's ease of ingest is far outweighed by painfully long render times in standard XDCAM Time lines (especially HD flavors) It is a very nice clean looking picture - but the file structure requires hellishly long conform times when you output. (a 10 minute program in high bit rate XDCAM HD took nearly 40 minutes to render out to a QuickTime file for playback at event nearly causing us to miss a deadline.)

I agree a ProRes timeline (not just render) makes life much easier - the next version of Final Cut would be a dream if it would allow log and transfer to directly wrap the files to pro-res to edit directly instead of XDCAM flavors.

My opinion is somewhat different, slightly.

Firstly, avoid working in long GOP whenever you can. It is troublesome, renders take longer, and so do your exports. That is if you don't meet with random crashes. You will have to eventually finish in a good quality I-frame codec, as XDCAM is a bad mastering format.

Between ProRes HQ and ProRes, both are 10 bit codecs. The only difference is in compression. ProRes HQ applies less lossy compression than ProRes (SQ). So use what you can best afford, but don't kill yourself over it. Without a good monitoring setup and good squinting, you are unlikely to notice the difference when comparing between the two.

For raster size, your source is 1440x1080. You do not get better quality by adding pixels. However, text and graphics will look sharper if you render to full raster. This also depends on your delivery format. If you are capturing via HD SDI, you capture at 1920x1080, because SDI always sends an uncompressed full raster HD image through the pipe.



www.strypesinpost.com