OK so far I’m 90% there with them going with the Alexa LF. Will I still have issues if they want to push in from a head-to-toe wide to a mid? I’m guessing that will depend on the lens they use. Not too familiar with a standard Alexa shoot.
That sounds like a big push-in. When you consider that your frame is going to be wider than it is tall, to go from extra wide shot to a mid like that might be 50% push-in. If you’re finishing at HD, then you might be fine but if you’re delivering UHD, I’d be worried about it going soft.
I get a lot of material from the Mini LF. I think it’s a really great camera and the Prorez 4x4 should be just fine. You can punch in the equivelent of 230% in respect to HD, and still maintain a minimum 1 to 1 pixel at HD. That said, however, if focus is off just a gnat’s ass . . . ets, it will show badly.
Whilst true, this doesn’t necessarily mean that you can take two shots, punch in 230% on one of them, and they look good next to each other.
The apparent sharpness from a camera and lens happens when it’s all super sampled down the HD from higher res. I’ve never done a big ass blow up/less of a scale down and it not be noticeable.
The only way to tell is to do a test on set on the first take and get everyone to sign off then. Do not let them make this your problem.
Perhaps I didn’t say that properly. An LF image is 4448x3096. To fit that in an HD raster would be to reduce it 43.1654% The inverse of that (what it would be to scale it back up so that the end result is just a center/crop of the original) is 231% or so. I don’t advocate blowing any image up 230% and expecting any quality. Depending on noise, grain, and focus, and other things, a shot that is punched into might stand out quite badly, but technically you have not uprezed your original image. As you say, do a test and see what people are good with. In the world of TV commercials I hear few complaints about image quality on punch ins.
Well I just found out that they are going with the Panasonic GH5. The rational is that it’s a corporate video being streamed and it’s just a talking head with powerpoint type graphics on the screen behind them.
I’ve insisted that they shoot all the sizes they need. I’m not advising on blowups.
Am I right in asking them to shoot to an external recorder that will give me slightly better quality and Prores is also nicer to work with in Flame.
I would go with an external ProRes recorder and Rec709. Sounds like you have to do a lot in a little amount of time so taking Complicated grading out of the equation will help you.
Also I would say that shooting against a good green BG will help you the most of anything. Get them to use some of that money they’re saving by using a GH5 to rent a stage with a green cyc wall. Or, at the very least, a backdrop far enough back from the talent so there’s no shadows on it
The green screen they are going with is vinyl and I was given a sample of the type of setup they will do. Very smooth and flat. Here’s a screen shot.
To be hones I’ve keyed with a lot worse that this having RED footage with a shit lot of noise. I’ll be onset to make sure I get the best results. I might even take my Flame on set.
John- Got to let you know, I had a similar green screen edit with G5 footage 2+ years ago. 4K internal recording…don’t remember their settings. It was like using h264 footage for keys. Hair was terrible. I didn’t have Neat. Edit took twice as long as it should have. Use an external recorder.
Capture a couple of clean plates of the green screen between takes. Then you can use an IBK style key which could make a lot of the detail work easier.
Also please report back on how it all went. I’m very curious about the prosumer camera + external recorder results.
So I asked the DP to record to an external recorder and this is what I got back:
“Like I mentioned before, I highly recommend against using them. I find them unreliable in that they often drop frames, fail to record correctly or add additional noise to the image.
"The GH5 records in 10 bit 4K at 400mbps All-Intra H.264 codec. The difference between internal recording and Pro-Res external recording is imperceptible and the unreliability of external recorders makes internal recording a clear winner”
My old highschool was recently closed down due to the presence of cancer causing chemicals. These chemicals were not a new arrival; they were used in the CONSTRUCTION of the building. They’ve been there for DECADES.
…AND I TURNED OUT FINE!
The DP is defending a cancerous building because he doesn’t yet have cancer. Just because an H264 hasn’t personally burned him doesn’t mean it’s not a goddamn biohazard of a work format.
This is the sort of DP you get when they shoot with a toy camera.
also, that stat is bullshit PR. I just tested how big 24 10-bit DPX’s are at 4k (of pure color noise). They’re 800 megs. A prores4444xq is 200 megs. There is NO WAY IN HEELLLLLLLLL that the camera is writing a 400meg-per-second h264 file. I’d imagine that what comes off the sensor might be that level of throughput, but it’s not what gets written. Not even close.