It's hard to tell for sure from the summary, but the pink skin tone and over-saturation criticism seems to apply to shooting in rec 709, which of course nobody does.
Alan Roberts evaluates cameras for performance standards which aren't meant to relate to cinema aesthetics. 35m Film stocks from the 70s, 80s and 90s likely wouldn't qualify at Tier 1 or 2 because they'd be far too noisy, as well as inaccurate color.
Colour comments aside, I did appreciate his findings regarding sharpness and raw versus ProRes and controlling noise. But testing a camera against broadcast specs isn't going to be the final word and he concedes that in his summary that includes both the burger and the fries!
My take reading this article which is only re-reporting Alan's test, is that a lot of his comments about the colour and look of the camera comes from using the VIDEO encoded space, not FILM. VIDEO has always been BM's own look for a pre-graded image, and it appears to me that the comments about the colour are in the context of comparing it to an expectation that VIDEO is REC709....it isn't.
Also, it would appear to be somewhat confusing as to if all his testing was done with windowing / crop mode ON, or even that the full sensor / supersampling of the 4.6K field was used.
We don't know which firmware was tested. I know that things changed a bit post 3.0 for example.
Be great to see the ACTUAL test results rather than an article about the test. But we need to remember that these tests are done for a BROADCASTER's point of view.
I also find it interesting to note that it was a surprise that ProRes, as a side effect of compression, mitigates some of the aliasing, something many of us have known for some time.
No commercial I've done has even been anything other than LOG originated then graded later. I think it's only broadcasters, and event type work that needs it to actually look good right out of the box without anything else happening.
I agree, all but news and a few outliers maybe will want LOG footage anyway.
I wonder, does the new firmware allow you to record with a user LUT, or only display the result of it on the screen?
Evaluating raw DR using 4:1 is probably a mistake. He should have used lossless raw. The way BM does lossy raw compression shaves off a bit of DR.
Doesn't the UM46 support baked in LUTs? I thought it did. In fact, I specifically asked a BM rep this exact question at NAB last year and I was told it did.
EDIT: to be more specific, I was told the new 4.0 OS would support baked in LUTs, which wasn't out at the time.
|« Previous Thread | Next Thread »|