DV > FCP > Online workflow

[Jeff Carpenter]

If you capture by firewire then what is on your hard drives is exactly the same as what is on the tape. If you’re taking those hard drives to the post house then you’ve done all you can do. They’ll take over from there, you’ve given them the best possible quality.

The ONLY issue I can think of is this: Will they be simply taking your project and openning it up in their Final Cut system, or will you be exporting a final video that they will then import into something else like an Avid?

If they’re using Final Cut, don’t worry about a thing. Feel free to add some filters and effects on your own…they can change them or erase them if they don’t lke them. As long as they’re using your project they have access to the original, max-quality video files.

IF, however, you are going to export your final project for them to re-import, you should not do ANY color correcting at all. Any filter you apply will slightly alter the video which will then be altered AGAIN when they color correct it. You can lose quality this way. But again, this only matters if you’re exporting your project as a single video file or laying it back to tape.

[Bill Dewald]

[vanBaal] “considering everything was shot on DV, is my edit already “online” (full resolution)”

Yes, but… Anything you add to the footage, be it filters, text, etc. will be compressed with the DV compressor.

In that case, exporting an uncompressed sequence is the way to retain maximum quality.

[Alexander Gao]

should the dv be captured in uncompressed, then, or still just ntsc dv?

Alexander Gao

“When the revolution happens, I’ll be leading it.”

[Mitchji]

[thewanggao] “should the dv be captured in uncompressed, then, or still just ntsc dv?”

Hi,

Capture is not necessary but rendering out uncompressed to reduce passes through the DV codec will be good.

Another option would be rendering out as PhotoJPEG@75%. See this thread:
https://forums.creativecow.net/cgi-bin/new_read_post.cgi?univpostid=886339&forumid=8&postid=886430&pview=t

Best Wishes,

Mitch

[Lu Nelson]

Whoah — hold on.

First, yes you can get somewhat better quality by recapturing from tape. If you originally captured your footage with good machine control (using an RS-422 connection and not firewire control), then your timecode is accurate enough that you could consider recapturing your clips to an online sequence (any higher res YUV codec from DV50 up to 10bit uncompressed). You recapture from a high end DVCAM deck (DSR-1800 or 2000) via SDI. These decks have hardware signal control, error control, and chroma upsampling.

If you captured originally with firewire control (i.e. using a firewire cable for both capture and deck control) your timecode may not have the same offset on every clip and if you recapture you’ll find your hundreds (or thousands of edits) are all out by a frame more here, a frame less there, etc. It’s a nightmare and you better not do it. So what then?

As the first response suggested, you should find a post house that can deal with your DV footage files directly from disk. Ideally, they are simply color-correcting on a higher-end FCP workstation, or perhaps using Final Touch HD..something that can handle your FCP timeline and files directly. Explain that going back to tape is not frame-accurate.

I would add: beware of trying to “export” your sequence to a higher res — there are pitfalls. There’s a lot of excitement on this board about going to PhotoJPEG, but this is an RGB codec, and unless you color correct FIRST, that color-space change is going to clip off all the white values in your DV footage (which is YUV based).

LMN

[Graeme Nattress]

Lu, since when was PhotoJPEG only RGB. It is both YCbCr and RGB depending upon quality setting. At 75% it’s Y’CbCr and at 100% it’s RGB.

Graeme

– http://www.nattress.com – Film Effects and Standards Conversion for FCP

[Lu Nelson]

Uh, well that’s a surprise. I’m not sure I understand. I am talking mainly about the problem of having superwhite values (which one nearly always does) in your DV footage.

Are you saying that if you were to batch export for example, a series of raw uncorrected DV clips in your bin — or, alternately: use media manager’s “recompress” function to convert your DV footage to PhotoJPEG 75% (we’re talking about the same codec that at 35% is the basis of Apple’s OfflineRT setting, right?), that all the superwhite information in your DV clips would not simply be clipped off? That PhotoJPEG would actually allow for superwhite values? My understanding is that they would be lost; but that if you did the same operation and converted your footage to DV50 or 601 Uncompressed they would not as the color/luminance values remain the same.

Please correct me if I’m wrong.

[Graeme Nattress]

Yes, it supports Super White at the 75% setting – just tried it to be totally sure.

Graeme

– http://www.nattress.com – Film Effects and Standards Conversion for FCP

[Andrew Van baal]

Thanks for the responses everyone.

Capturing via Firewire is really my only option as far as the offline is concerned; I don’t have a DSR deck or a SDI card and I doubt the client would spring for either.

The image looks good (enough) for me captured via Firewire in FCP, but I do want to maximize its potential; how significant is the difference in quality between a Firewire DV NTSC 48kHz capture and a component capture from a DSR deck? Which leads me to another question: how important is the deck in maintaining the original quality of the tape image? Are all Firewire decks the same in that regard or is there some variation?

The post house that’ll be doing the online (Outpost Digital in Santa Monica; anybody worked with them?) is going to work directly from my FCP sequence/hard disk files, as you recommended. So that’s good.

I’m concerned about the timecode issue, though. Not so much for this project, but let’s say I’m doing an offline edit for a non-DV project that has to be conformed. Firewire timecode is completely unreliable for this?

AvB
FCP Editor

[Daryl K davis]

I’ve found that it isn’t so much the firewire timecode is the concern, but more so the firewire deck control.

We now always capture off the DV deck using RS-422 deck control becuase we were finding a 1 or 2 frame discrepancy when we used firewire deck control. Then if a project is to be re-captured uncompressed in on-line from original DV tapes using SDI and a capture card, or even if you were off-lining with DV clones of the original D-beta or Beta SP camera tapes and now need to conform from the camera originals, we’ve found it critical to use RS-422 deck control for sake of frame accuracy. Also that 1 or 2 frames off drives audio post people nuts.