Workflow: XDCAM 1080i25 + HDV 25 to NTSC–Shoot me now!

[David Eells]

[Andy Mees] “Working with XDHD and HDV only really becomes problematic if you actually want to deliver in that format (and you don’t right?)”

No. What’s important to me is that I can cut quickly, the pictures look good, and I don’t have to wait an hour to reconform every time I want to output a version for the EP. Which will happen at unexpected intervals.

[Andy Mees]

Hi David.
Well what more can I say? As I already noted, for me, pushing XDHD and HDV around in a native XDHD timeline gives me no problems at all. Its fully responsive and everything works in realtime. If something does need to render then it renders quickly as ProRes (ie doesn’t need to conform). With added hardware (Kona, Decklink, MXO) I can output via HD/SDI in realtime. In the box I can output a low quality SD viewing copy via firewire to a DVD burner or whatever in realtime. Or I can export a pristine quality ProRes copy via Export > Quicktime Movie w/ settings set to ProRes, Self-Contained. etc etc I think the real point though is that while all this may work for me or others it doesn’t necessarily mean its your best workflow. I think you need to sit down and run through some workflow tests for yourself in order to satisfy your concerns. In the meantime, what you also need to concern yourself with perhaps is how your EP will view the ad hoc viewing copies you will need to create … if thats a DVD then how will you approach the working format (PAL) to NTSC conversion? and how needing to create NTSC viewing copies may affect your workflow, or more importantly the time needed for that step (or the quality trade off acceptable for making it happen faster)
Andy

[David Eells]

Hi Andy

I appreciate your input. For the system I’m using at the moment(dual Xeon 2.66), the video is BARELY realtime. The actual rig I’m supposed to work on is a G5, so who knows how that will work. It’s possible we could rent a Kona for output to a line monitor, if we can justify the cost – not sure on that one.

I got called in on the end of an HDV project once, and outputting was a nightmare.

The EP has access to a all-standard DVD player, or could view from a laptop -I’m not anticipating having to convert until the end of the process.

So I have an opportunity to concert the footage if I act now. After Monday it’s probably not an option.

But anyway, thanks for giving me your perspective on this.

[Andy Mees]

[David Eells] “For the system I’m using at the moment (…) the video is BARELY realtime”

Is this referring to the performance you’re getting with your footage as is in the ProRes422 HQ timeline or within a native XDHD timeline? If the former, how does it fare in the latter?

To be honest, I’m sensing that you’ve kind of already made up your mind on the workflow that you want to use (transcode all your sources to ProRes422 HQ) and are really looking for someone else to confirm that its a good one / the best one? If thats so then go ahead, its a good one. Not what I’d do (I’ve covered that pretty much I think!) but pre processing the footage in order to conform to a single DI codec is a perfectly goo d workflow.

If that is the way you’re headed then I think you’ll need much more storage than you appear to have at present (and this may already be the cause of your primary performance issues). If your native sources already total 2.2 TB and your Burly RAID has only a 2.5 TB capacity then your current storage is running nearly 90% full. You’re not going to be getting the best performance from drives, even a RAID, that is that full. Now take all those low bit rate compressed HD sources and recompress them to a higher bitrate codec such as ProRes422 (HQ or otherwise) and you’re not going to be able to shoe horn all that footage onto your drives. More food for thought perhaps…

[David Eells] “I got called in on the end of an HDV project once, and outputting was a nightmare. “

I feel your pain mate, especially if you had no prior experience of handling such a project or guidance when it happened, but since that time, have you learnt anything new about dealing with HDV projects or would you approach such a beast in exactly the same way as you did last time?

[David Eells]

When I say barely realtime, I mean that it goes very soft, but it plays. If I put the ProRes HQ footage in an XD timeline, or XD footage in a ProRes HQ timeline, the result is the same. HDV is happier in the XD timeline.

The footage breaks down like this:

Pro Res HQ media already ingested: 402 GB
XDCAM HD: 318 GB
HDV 720p: 189 GB
DV Pal: 150 GB

I see what you mean on storage-
I have additional 1.8TB raid that I could use, and I could possible get more.

I’m not determined to convert tons of footage, just a little worried about what I might get locked into.

The last HDV project I was referring to was just something where they burned out the editor and brought me in at the end – I knew nothing about the project, and the previous editor was, shall we say, disgruntled, and not inclined to help.

So what I learned from HDV was to stay away if possible!