Pan and Zooming in on RED footage

I'm starting my first RED project and I've read up on the workflow and such but can't seem to find an answer to this. I'm editing in ProRes 422. This is only being finished for web use, so no film or going back to the R3D files or anything.

There's a shot where I need to zoom into the subject and it's my understanding that Red footage allows you to do this without losing much resolution. What's the best way to go about this?

(I'm on FCP 6, ny the way)

Thanks!

You shot a Web spot on Red? Must be nice to have that kind of budget to throw around.

Yes, it's true that Red has more resolution that any video format. Depending on how the camera was set up, you might have 2K, 4K or 4.5K footage. Any of that can be blown up, obviously, because there are many more pixels there than you're going to use. Knock yourself out.

Jeff,

I went to the shoot today to just be there as "Hey this guy is the editor." I didn't know much about the shoot other than the concept. I was expecting them to shoot on DV and saw the footage on the monitor and my jaw literally dropped that they were using the RED. Obviously it looked amazing...though it would've been nice to have a heads up.

They had a red rocket card on set so all of the footage has been converted to ProRes 422 for me.

I guess my question is, when I bring the footage into my sequence, will it display the image full frame/100%. Or will it be a percentage of that allowing me to zoom in? Or should I set my frame size to 1080? (I assume 2k footage is bigger than that)

Thanks!

-Matt

The resolution advantage on the RED makes for nicely sharp pro res 1080p or 2k, however the true resolution advantage is when you have access to a 4k decode. If you want to be really able to push a shot in, beyond what you can do with your HD decode, you'd need to get that R3D decoded to 4k and use that.

Graeme

Quote

I guess my question is, when I bring the footage into my sequence, will it display the image full frame/100%.

By default, Final Cut makes your clip fit your timeline. (Actually, by default Final Cut asks you when you make your first edit if you want to change your timeline to match your source material; if you say no, then it makes your clip fit your timeline. Otherwise, it's vice-versa.) If you drop a 2K clip onto an NTSC timeline, it will be scaled down a lot. This is a non-destructive operation; if you want to change the clip's scaling or reposition it, go right ahead. You'll need to render before you can view the results at full quality, but that'll be quick.

Quote

Or should I set my frame size to 1080?

You shouldn't change your timeline just because the source footage is higher resolution than you expected. You should keep your timeline set to whatever you want your mastering format to be. Since you said you were expecting DV, that'd obviously be NTSC.

One thing to note well, though: Do not set your timeline to DV. If you'd started with DV material, like you said you were expecting to, then the correct choice would have been to use a DV timeline, so you can work in real time all the way through the project. But DV is extremely low-quality compared to ProRes 422. It's even low-quality by Web video delivery standards, believe it or not.

Quote

(I assume 2k footage is bigger than that)

The Red One can shoot either 16:9 or 2:1. If they were set up to shoot 2K 16:9, your frames will be 2048x1152. If they shot 2K 2:1, they'll be 2048x1024. As you can see, 2K is slightly bigger than HD. Not big enough to give you a lot of latitude to repo if you're working in HD.

Something to watch out for: Did your DP roll at 24 frames a second, or 23.976? The Red One can do either. Since you're delivering only over the Internet, it really doesn't matter which you use, but make sure your timeline frame rate matches your source material's frame rate.

Thanks Jeff, that makes sense. They shot at 23.976. I'll make sure the settings match.

Thanks again!

-Matt

If they did the downconvert on set on a RED rocket, you got the best debayer possible for your edit files. What you didn't get is access to the 4k or 2k resolution to blow the footage up if needed. You got locked in by the downconvert.

The sensible process would be to convert the R3D files 1920x1080@23.976, although the fact that it's headed for the web means they could have converted to anything. If you don't have access to the R3D files, you're going to be limited to blowing up what you have and the usual margins apply. Anything above 5% blow up will be noticeably softer on a good monitor but you could get away with more than that on a web delivery.
If you had access to the R3D files, I would cut your sequence as it is and see where you might need to blow up footage. When you get near the end, you can recompress the selected shots using RedCineX or ClipFinder (my suggested choice) and do the blowups there and bring them into FCP.

An alternate method would be to do the edit as you are already doing it, then when you're almost done link the shots that need to be blown up to the full resolution R3D proxies and do the blowups in your FCP sequence. This would be slow, even with an OctoCore Mac but if you just have a few shots to do might save you some hassle.

ak
Sleeplings, AWAKE!

Oy, that's a good point, Andrew. I totally jumped to the conclusion that the footage was converted to ProRes at native resolution. If that wasn't the case, then yes, you're definitely going to have to go back to the original camera mags, Matt.

Even though the jargon is all different, Red workflow is exactly like film workflow. Your camera mags are your negative. The ProRes files you have are film scans. If the scans were done at high resolution, then you have some latitude to move them around and zoom in. But if they were done at low resolution, like SD telecine transfers, then you need to re-scan the neg to get enough resolution to repo your shots. Because re-scanning the neg is time-consuming ? i.e., expensive ? it makes sense to only re-scan the frames you actually need.

There are various ways to automate this, but that gets into offline-online workflow and is a subject for another discussion.

Yeah, the footage is not at native resolution, it's been scaled down. I have access to all the R3D files.

It's really only one or 2 shots that need a zoom in.

So should I go ahead and edit normally and once I'm done, use redcinex to go back to the original files which will bring in those selected shots at full resolution?

Yes. Edit as you normally would. Once you're done and you know which shots need resizing,
you can either:
1) Open up the R3D versions of the files in RedcineX and crop and scale them to the way you want and export to your editing resolution
2) Open up the R3D versions of them in RedcineX export the full res versions and do the resizing in FCP
3) Drag the Full size proxy file from the R3D folders into FCP and do your resizing there. I would render it out as a self contained QT once you've done your work, the proxies can really slow down a system if it isn't tweaked for them i.e. fast CPU, lotso'RAM, fast hard drives and fast drive interface.

ak
Sleeplings, AWAKE!

Is the proxy the file with the underscore _P?