Best workflow for V1U and FCP?

[Transmission2]

Im about to start to loading footage for a documentary and I want to make sure I have an appropriate workflow in place. Can someone tell me if this makes sense or please give me alternative solutions.

1. Load footage via FCP's HDV 1080i60 easy setup.
2. Do assembly edit in HDV.
3. Create EDL and export sequence as either uncompressed 8bit or 10bit to do color correcting and effects.

Will the uncompressed codec give me more options for color correcting? Will the color space be improved? Will the render times be longer for uncompressed than HDV? Should I export as DVCPro HD to improve render times? Sorry for all the questions....I just need some advice :)

I am editing for a finished piece which will run about 1.5 to 2 hours.

thanks in advance!!

[magi1500]

If you export your movie as 8 bit uncompressed (10 bit would probably be a waste because HDV is only 8 bit, unless your piece is chock full of motion graphics that you plan to do as 10 bit), it is going to take up boatloads of drivespace (you'll need a terabyte or so for a couple of hours) and to actually be able to play it, it will need to be a speedy RAID (capable of 200 MB / sec +).

I did a bunch of experiments turning HDV into DVCPRO HD and it degraded the quality. DVCPRO 100 is obviously a better codec than HDV, but something about the conversion process made my HDV footage look like crap. Not to mention you'll lose resolution. V1U's HDV is 1440X1080 native, while the 1080i60 version of DVCPRO HD is only 1280X1080.

Render times for HDV are hella long because of the long GOP compression (to decide what a frame should look like, FCP has to go back to the nearest I frame and figure out all those frames to get to the current frame, every time... that is a lot of math.) But if you're using 8 bit uncompressed without a RAID, your render times will take a while because you're talking tens of GBs worth of renders every time.

That said, while I don't have a lot of experience with color correction myself, the pros I know always export the final as uncompressed and work with it that way. I know the guy that cuts the music videos for American Idol, for example. They shoot that on DVCPRO HD. He ingests and cuts DVCPRO 100 native. When he is done cutting, he exports the whole thing as uncompressed 8 bit and delivers it that way. His computer can't play the files (he doesn't have a RAID), but the post house that does the CC can, of course. But his finished piece is only a few minutes, yours will be considerably longer.

Also, would you have to use an EDL, or could you just use Media Manager to export only the clips used in your sequence (with a few seconds of handles), recompressed as 8 bit uncompressed?

Either way, your workflow to start is sound. Ingest and cut as HDV 1080i60 (because you shot 30p, if you'd shot 24p, it would be different). Do not export as DVCPRO HD.

This will get you through the edit stage. Talk to whoever is going to do your CC about whether they would prefer two hours of HDV (around 25 GB) or uncompressed 8 bit (probably at least 800 GB).

Forgive my verbosity.

:)

[Transmission2]

Sounds good Magi1500. Thanks for taking the time to respond to me. Im going to stick with my plan an probably go uncompressed near the end of this project.

Looking at your stunning 24p makes me second guess my choice about the 30p. I wish final cut supported a 24p workflow right now. I have a delivery in may for this portion of the movie and I didn't think I could risk waiting for an fcp update, nor was I interested in running every clip through cinema tools. It would probably take way too much time???

I wonder if I decide to shoot 24p down the road... how would it cut in with the 30p, aesthetically speaking?

[Elton]

I edit HDV native in FCP, (HDV 108024p usually) and simply copy/paste a finished timeline into a custom Sheer codec (www.bitjazz.com) lossless timeline and do final grades and export from there. It still creates a huge QT file, but it is non-destructive and less than half the size of uncompressed, but the exact same quality. A 2 drive sata RAID can keep up with it.

If you have the SATA RAID, it is definitely faster than going through the "conform to HDV" process (rebuilding GOP's) and much, Much, MUCH better quality for a final delivery. It essentially allows you to finish without any further degradation to the already heavily compressed HDV image.

Finish in Sheer QT, export to a large external firewire drive (probably need a 500 GB Lacie for a feature), and have the option of exporting to D5 at a higher end post house, or do high quality encoding of h.264 for HD DVD/Blu Ray...or simply encode your DVD from there. Sigh.

[magi1500]

transmission, as far as how would the 24p look in your 30p sequence...

The plus side is you'd be able to capture HDV 1080i60 like normal (because the 24p is laid to tape at 60i) and, if you knew it was going into a 30p sequence and you'd NEVER need it to be in a 24p sequence, you could just drop it in with the interlaced frames and not worry about pulldown or rendering of any sort. It would have the feeling of footage shot on film and transferred to video. It would cut fine with your 30p stuff, but I'd save it for something you specifcally wanted to have a more cinematic or dramatic effect.

p.s., pet peeve of post houses: DO NOT USE A LACIE DRIVE. No offense, but they suck, and they have been the running joke of the three post houses I've worked for.

Elton, this Sheer codec sounds interesting. I'll check it out, too.

[Transmission2]

thanks a lot magi. i think at this point im going to stick with the 30p because ive already shot 12 tapes worth. when i go out on the road i may decide to shoot 24p for all the concert stuff and hopefully it will look fine with the stuff im shooting now.

as for the lacie drives...i totally agree. two of them fried at the post house i worked at so I stay away from them. Im using G-Raids right now. They seem pretty solid.