Hi all, I've been lumped with some footage to sort out and wondered if anyone could help, I'm relatively new to using Red (normally I work with a Sony ex3, I've had 3 days of research before the RED job was to start...) and was wondering what I can do about a couple things. We shot at 2K for some product adverts in studio with limited light, but with my limited skills it's still looking okay (the client has no problems with the rushes at least).
We do have noise, and to reduce it I've been using Neat Image Denoise which does a sterling job, however I would like to know if there's anything I can do prior to output from RedcineX (if that's the best option) to improve image quality.
My goal is to take the footage into after effects, fix what noise is unavoidable, add an extra layer, output uncompressed (animation/prores HQ/other???), take it into Color for a bit of clean up, and then a final export, probably from FCP. Hopefully this isn't too convoluted and makes sense.
Such questions to come may have obvious answers, but I have been dropped off into the deep end here, so be nice, and imagine you're explaining Red to a petulant, slightly slow child ;-)
What can I do to reduce noise caused by low light/crap light/other unknown variables without reducing sharpness later? The denoise software will be kinder to the footage later if there's less grain to start with.
What is the best quality output for taking footage into After Affects? Limited testing has shown tif sequences to be fairly nice, but I'd love to hear that there's a better alternative.
Is there anything I can do for banding that doesn't involve copious noise? (as I'm trying to reduce that!)
I've got until first thing Wednesday (UK time) to sort out 5 8-second files, so any advice will be most graciously received.
I'm sure I'll have additional numpty questions to ask as problems arise, but I'll stick to these for now.
I sadly wasn't directly involved with the shooting, but I believe it was shot with an even histogram, it was shot on the Red MX, I believe (I will confirm this as soon as I can), and my preliminary results involved me making a minor adjustment on the curve, outputting as a tif sequence, importing in AE for the required extra footage to be overlayed, then outputting as an uncompressed file for Color correction. Sorry if this isn't what you mean but I'm a bit behind on technicalities and am playing catchup...
My advice to reducing noise is to do it as the last possible pass on the video. There is no reason at all to not finish RED footage directly from the R3D as every capable post software reads the R3D data directly now.
Taking your problem into an APPLE workflow solution would be this.
Relink to online R3D media.
Grade in color, apply FX etc.
Send back to FCP
Add NeatVideo to the problem shots.
That seems to be a fairly straightforward method, with one key issue for me. I assume I bring in the QT reference file for the first edit, then how do I go about reconnecting with the r3d file? As I said, if there's a spectrum with numpty to expert, you can guess which end I'm at...