- Posts: 519
- Joined: Sun Jan 17, 2016 6:56 pm
I upgraded my A camera from the Ursa Mini Pro G2 to the Ursa Mini Pro 12K a couple months ago. The reason I made the switch was for the full sensor scaling in BRAW and the higher frame rates, as these provide more options that I can offer clients. It turned out to be the right choice as despite the pandemic production drought, my clients have been quite pleased with the added flexibility and in a couple cases the 240fps was just enough that I didn’t have to rent a different camera to deliver an overcrank effect that a couple different directors wanted.joe12south wrote: 3. I genuinely want to learn what advantages to color reproduction the 12K offers, if any. Truly. This isn't a "gotcha" test. I want there to be improvements. I love that BMD pushed sensor tech in the way that they did. What I have said is, to date, I haven't seen a clear advantage with the limited amount of footage I've seen.
But, of course, with that new sensor tech in there I’ve been curious about that same question — is the color quality on the 12K meaningfully different than the G2?
Where I (fully subjectively) see a difference is while grading. The images from the 12K appear to have more malleable color deeper into the shadows, noise reduction seems to work much easier than on the G2 (smaller amounts of noise reduction, results in a cleaner image), and I can shift the hue, sat and luminance of particular colors further before the image appears to break. I also appear to get more subtle differentiation of hues in the shadows (and perhaps the highlights too). These differences seem more noticeable in HDR than in SDR (which makes sense as the available image data is being stretched over a larger range on the display, and it is why Dolby recommends 12bit color depth for their HDR spec, or at least that is my understanding).
Now, of course, those are fully my own subjective impressions while grading images I've shot on the 12K. So I want to be clear here — I’m not saying these differences that I think I see are objectively true.
But, if I was setting out to design a test to look for objective color differences in the 12K sensor data, I’d be investing most of the time into trying to figure out an objective repeatable method to compare those variables.
Also, since the launch BMD has presented the 12K as being useful for VFX work. I don't really do enough of that work to even begin to suggest which image attributes matter most for it (though I've been told that control of image noise is quite important) so I think that could be another area to mine for possible objectively testable differences.
EDIT: Just gonna add here, been spending part of today grading more files from the 12K. The images from that camera viewed on a calibrated 4K HDR display simply look damn good. I can't explain it in objective terms and who knows if the difference would get obliterated by YouTube compression. But, viewing the files in front of me on a calibrated monitor. Wow. They look absolutely fantastic.