Sean Oneil
Forum Replies Created
-
[bogiesan] “This is total nonsense and bizarrely ego-centric. First, every second of your footage is not usable and, second, even if every second was perfectly exposed and crucial-to-the-film, you’re not releasing a 200-hour film and, third, even if you were funding the release of a 200-hour epic, no one would watch it for five weeks, and, fourth, do you people think you invented documentary production? How do you think doc producers have been editing beautiful work for the last 100 years?”
I disagree. Sifting through 200 hours of footage while it’s in linear tape form is in most cases a waste of time for an editor. A student or unskilled person should be hired to ingest everything first, assuming you have time to do it and you have better things to do while its happening (like working on the story, or working on another project). It’s far more efficient to pull selects after its been captured.
The old school way of doing things (which I’m quite experienced with) existed because capturing 200 hours of footage would have been expensive and impractical (if not impossible) in the past. That’s not the case today. Hard drives are dirt cheap. 200 hours of footage is nothing. I haven’t made a VHS tape for someone in years.
All that said, I agree with you on her situation. It sounds like a bad idea. Unless she is getting paid well and by the hour, I’d tell the director and producer to do the monkey work themselves or find someone else to do it.
Sean
-
I read about this thing in Engadget a while back.
Faster or not, the fact you can encode h264 in the background without tying up any resources is an interesting thing to consider.
Sean
-
[Shane Ross] “Can’t be getting more than 100MB/s…that is the limit of PCIexpress.”
What are you talking about? PCIexpress is 500MB/s per lane. And ExpressCard slots are 250MB/s.
-
In college I was taught that turning decks on and off does the most damage. The ICs being warmed up and cooling off over and over again is what will kill it.
That said, I think this is kind of a dated concern. Probably better to turn it off.
But as far as analog decks like BetaSP, I think that you should make sure they’ve been warmed up before you use it (30min should be plenty) because it could affect the video levels.
-
[Shane Ross] “Export an EDL from FCP. Open Cinema Tools. Cancel any query it has about opening or accessing a DATABASE. Go FILE>EXPORT>EDL…30 to 24.
Import that EDL into FCP. Might have issues with speed changes, but that is what I have done.”
Just curious. Does that work if the 30fps footage was drop-frame?
-
The 1200 does not have your daddy’s “crash record”. If you are using Firewire output, it will read the TC from the FW and lay it down perfectly. So there’s really no downside except that you have to make the whole tape at once.
-
Just a quick disclaimer. When you remove pulldown using a capture card like a Decklink or Kona, you lose the ability to batch re-capture selections of your footage. It will assume that the in point of every clip is the A frame. Not good. This may or may not have been addressed by Apple, BM or AJA – but as far as I know it hasn’t.
So unless you are capturing at online quality, don’t do it unless you are prepared to re-capture ALL of your source footage. Thanfully with FCP6’s ProRes, the days of capturing at offline quality are hopefully over.
-
[KMimsey] “So what would the right way to do this from the beinning be? Edit at a different frame rate? Capture differently?”
If you shoot at 24fps, you should edit at 24fps whenever possible. Bad cadance causes a lot of problems nowadays. People have LCD televisions and watch clips on the internet. So the idea of cutting 24fps footage at 29.97 is a really bad one nowadays.
But when your source footage is different framerates, this is where things get very tricky.
Ideally you should be able to capture everything at its native framerate. And then you can edit everything on a 29.97 seuqence. Because the 23.98 footage is captured at 23.98, you won’t have any cadance breaks.
The problem is that FCP 5 is not meant for this. When you take ftg captured at 23.98 and put it on a 29.97 timeline, not only do you have to render it, but it doesn’t add new 3:2 pulldown to it. It’s really stupid. Instead it just duplicates frames. This causes a stuttering effect that is noticable during pans and such.
This is why the other fellow in this conversation (Gramme) provides an “Add Pulldown” filter. But despite how great and useful it is, it requires a lot of extra steps.
FCP 6 is designed to mix formats, including different framerates (Avid can’t even do that). So I can only assume they fixed this annoyance. So once you get FCP 6, the workflow should ALWAYS be to remove pulldown from any ftg that has it – regardless of what framerate your delivery is going to be.
With HDV its easy. You just choose the 1080p24 Easy Setup. With SDI formats, it’s a lot trickier. You have to identify the A frame prior to capture inside the Blackmagic System Prefs under the VANC menu (probably similar process with Kona). A properly downconverted Digibeta will have the A frame always starting at the even 00 TC number. I know the big facilities here in LA check for that now during QC sessions. But you can’t rely on it. You should always double-check.
-
[KMimsey] “Any tips on finding the A frame? I’ve had a really difficult time with that so far.”
Open the clip in QT Player. QT Player doesn’t deinterlace for you, so you can manually check the first 5 frames to see what the order is.
Doing a reverse pulldown to footage after it’s been edited at 29.97 is an inredibly painful process. Doing each clip individually in CT is what I’ve done before. But there are alternatives.
One is to hire a facility that has a Teranex. These do a good job at this. But it’s very expensive and you will be stuck moving into a high-end HD tape to tape workflow (D5 or HDCam).
Another possiblity is a freeware Mac program called JES Deinterlace, found here:
https://www.xs4all.nl/~jeschot/home.htmlIt can do a 3:2 removal whilst automatically detecting cadance breaks. It’s actually really cool and it can create a new clip for every cadance break. It’s not a magic bullet solution thought. It will make an occasional mistake. Sometimes you’ll come up a field short. Also, it’s destrictive in the sense that you lose original TC. But it can get 95% of the work done for you.
-
[Graeme Nattress] “I think that somewhat defeats the point of the magic new codec agnostic timeline though. HDV plays back just fine – it’s it’s rendering that’s a pain. Playback of HDV on a ProRes timeline should be fine, and then any rendering goes to ProRes, which is also fine.”
One of the reps at the Apple booth told me that new multi-format timeline feature will not perform very well with HDV, and that I would want a quad or eight-core Intel Mac to do it. Even then she reccomended not doing it at all.
Of course, she could have been refering to using an HDV timeline, which would make complete sense. So you could be right. HDV on a ProRes timeline could work really well. The only issue is keying and CC.