UMP Noise/Blueish color

So here in our office we use a Ursa Mini 4.6k and an Ursa Mini Pro. Lately, I've been noticing that my Ursa Mini Pro doesn't handle low light as good as the Ursa Mini 4.6k. So I wanted to see how they were stacking up together at 1600 ISO and discovered that the Ursa Mini Pro looks absolutely horrible compared to the UM46 and I want to know if this is normal? I'm pretty sure this camera has slowly been getting worse. Below I split screened the two cameras at 1600 ISO with lens cap on to see what kind of noise they produce. The Ursa Mini Pro isn't even black, it's a purplish blue. And yes, I black calibrated both cameras before this test. I brightened the image a little in post so you can see what's going on, but non brightened versions are here:

https://drive.google.com/drive/folders/ ... sp=sharing

UMP VS UM46.jpg (717.42 KiB) Viewed 2120 times

Here were the settings:

1600 ISO
4500k
Tint 5
1/48 Shutter Speed
Default Sharpening off
UHD Prores 422
Rokinon 85mm lens (Lens cap on) Iris wide open

So my question is, is something wrong with my camera? I don't remember it being this terrible compared to the Ursa Mini 4.6k when I got it. Blackmagic? Thoughts?

How can I fix this? Because it looks blue compared to the Ursa Mini 4.6k

I don't have the UMP but isn't this was Black Shading Calibration is for?

Edit: Oops. Missed that one crucial line of text.

Phillip, did you also do a comparison at ISO 800 and ISO 400? It would be best to do these tests in uncompressed raw. If you want to see patterns easily, also capture test results with and without False Colour. The camera is not a ‘no light’ camera where all sensors will perform as you might expect.

I found during my URSA Mini 4.6K Camera testing of low light scenes with the lens cap off, it’s best to shoot in raw at ISO 800 and raise a little in post if you must. You want to perform tests on both cameras after each has been powered on for several minutes too so that both sensors are likely at the same operating temperature.





Sent from my iPhone using Tapatalk

I would wait 10-15min, then do a black calibration, with lens cap on the camera, no lens, to reduce any chance of a light leak from the lens, and do this at 400, 800 and 1600, since the camera does not apply any real gain, this should not change the results for a black, no light sensor check.

That said, I had a Nikon V1whch had nice blacks displayed when the lens was caped, but the V3 gave me a very noisy looking screen in the EVF, like you are getting. But when properly exposed, both csmeras had a similar image, no noise. So doing a lens cap test does not always really show much, except a stuck pixel.

You would be better served shooting a dark color wall, with a color checker included in the frame, properly expose, as a Rick suggests, and then compare the results.
Cheers

So here's a couple stills of DNGs I shot at 1600, all same settings as before, except now in Uncompressed Raw at 4.6k. and against a low lit background.

Here's the Ursa Mini 4.6k

UM46 DNG Still (1).jpg (564.44 KiB) Viewed 2070 times

And Here's the Ursa Mini Pro

UMP DNG Still (1).jpg (584.17 KiB) Viewed 2070 times

You can see blue streaks running through the black case in the background on the UMP, and the image from the UM46 looks much different.

I punched in on the black case and took a screen shot, and the black looks blue and streaky on the UMP and not at all on the UM46. You can see those shots here as well as a DNG from both cameras.

https://drive.google.com/drive/folders/ ... sp=sharing

You do know, shooting Raw at 1600 is no different from shooting Raw at 800, except the mid level curve is a little different, as thr image is shifted towards the high side. You need to shoot the tests at 800, Native ISO to compare the two cameras. BM cameras do Not apply analog gain when you change the ISO setting, it just shifts the mid level curve.
Cheers

Phillip, thanks for taking the time to do the extra tests, but you may have forgotten to test at ISO 800 and 400 with exposure adjusted which will inform you about the behaviour of the sensor.

I’ve stopped using ISO 1600 for low lit scenes. To me 1600 may have a place in very brightly lit scenes but generally you can set the camera at ISO 800 and adjust other elements of the exposure. You may also find you need to bring down the shadows or mid tones in post to clean up the low lit portions of the image.

I gave up expecting the camera to do everything you wished it could do when you don’t have a well lit scene. I am determined to learn how to get there in post though and that can be very satisfying when you get there. When Blackmagic calls these cameras “cinema” cameras that’s just a much shorter way of calling it the Blackmagic “sometimes the camera can get what you want and sometimes you have to do the work but it’s worth it” Camera!


Sent from my iPhone using Tapatalk

All the sensors have a slightly different temp/tint even when set to the same numerical settings. For better or worse, those temp + tint numbers in camera are relative rather than absolute. It's normal behavior and it's true of other brand cinema cameras as well. Rather than using the same numerical settings, you should use the auto white balance with a grey card to bring the two cameras in line, or you can adjust the temp/tint manually in camera while viewing the image on a parade scope, or do it in post in the raw tab of Resolve.

If the black balance really truly refuses to line up between the two cameras for similar expsoures despite your best efforts to match them, then you may want to contact BMD support.

Just wanted to give an update to this post and to give a shout out to Blackmagic for massively improving my Ursa Mini Pro. I ended up sending it in to them because My camera was showing horribly different results than our Ursa Mini 4.6k, as I explained in the original post....but here is the comparison again that I originally made....

UMP VS UM46.jpg (717.42 KiB) Viewed 1782 times

I emailed them about it and they said they could probably improve the quality...So I sent it to them about 2 and a half weeks ago, just got it back today, and here's a comparison of what it looked like at 1600 ISO with the lens cap on from before, to what it looks like now (and yes all black shading had been done prior to the BEFORE test)

After Camera Fix.jpg (614.54 KiB) Viewed 1782 times

So yea now it looks great! I did an 800 ISO test with the lens cap on and ramped up the gain in post and it is really hard to see any noise anywhere. I think they might have also replaced my LCD screen, because it use to look really terrible, and depending on the angle you looked at it from it would change colors to some nasty yellowish color. Now it just looks like the 4.6k LCD screen!

So thanks Blackmagic! I'm super happy now!