IBC Red 1 5k DCinema Screening & Q&A

I suppose it defeats the purpose a bit to have a youtube of the Red Camera's 5k test clips. But there you go. Have a look or check it here on Youtube.



As for the images, Risky Biz says Red

"didn't seem to display test footage that was all that revealing. Camera tests that are more resolution and contrast intensive would be a good next step"

Mike @ HDforindies had the following to say: Excerpts, for the full whammie check Mike's site

Shot 1:
What: an Oakley Time Tank watch, shot macro style

What makes it tough - it is an extremely shallow depth of field, it has some contrast challenges - shiny metal, detailed black surface of watch face, fine details in watch face and watch hands, shiny highlights.

What works: the extremely shallow depth of field shows off what is difficult/impossible for a smaller sensored camera to do - get this extreme depth of field. The dynamic range is broad but not impossible, but I personally watched Graeme Nattress abuse the crap out of it in post before I saw ANY noise AT ALL in the blacks. The watch face was totally blown to white at that point, the blacks were lifted very high, and THEN, FINALLY I started to see some noise in the blacks. Insanely clean as compared to other digital images I've seen.

Shot 2: Joanne blows a bubble, Ted pops it

Joanne (Red employee) is blowing a bubblegum bubble against black, Ted is in background and leans up to pop it.

What makes it interesting: Ted is in background, so get proof of 35mm depth of field - he's out of focus. Good skin tone demo on Joanne. She has dark hair, so the black on black demo of shadow detail is good. Good detail in the bubble - you can see texture and thickness inconsistencies in the bubblegum. Ted leans in to pop it and travels in/out of focal plane. Smooth contone on bubble highlight. I defy anyone to call this a "plastic fantastic" digital image.

Shot 8: Porsche 959

What makes it interesting, other than that it's a 959: It's a white 959 with lights shining on it, but including one practical light in the shot. Smoooooooooooooooooooooooooooooooooooooth gradations of white across the car, nice shadow detail while not blowing out the highlights. Subtle shadow details in background - a crewmember noticed an orange extension cable in the dim background that he figured would never show on camera....oops. The dynamic range in the shot is the big story here - the hood of the car has bright lights on it, but the underexposed lower portions of the car have good detail as well. The REAL kicker is that there is a practical light on set. Asking folks what they thought of the footage in general when they left the 4K auditorium, multiple DoPs (who are NOT affiliated with Red) pointed out that shot, and volunteered that it looked like anywhere from 12-15 stops of lattitude (their guessed range of numbers not mine let me stress) represented in that shot. I'm not a shooter so I can't vouch or give opinion on that, but these were qualified guys, and that's an impressive figure coming from their lips.

And again, for those that didn't read it earlier, the footage was NOT noise reduced - just some very, very simple color correcting - as Lucas from Assimilate put it (and he was techincal supervisor on the color correction) -

Something important for everyone to know - from the final 5K RGB images pre-color to the rendered images at IBC - with the exception of one shot (the girl blowing the bubble at the beginning of the sequence) the only color processes we did to the images were Lift (master) down, Gain (master) up, Saturation up, Lift (Blue) down. That's it. No shapes, keys, or secondaries. The RED team has done a very nice job with their debayer/matrix algorithms.

Also, here is the Red Q&A. I need a quiet place to listen to it so, will have my comments soon.