URSA 12K banding

Sigma 18-35 @F4.2 - 5.6

ZEBRA @95%

Well, I guess I've got some shooting to do tomorrow to test my 2 cameras with OLPF.

Yesterday I had the chance to continue testing this issue comparing the behavior of the URSA 12k OLPF vs BMPCC 6k G1, using same lens the SIGMA 18-35 @28mm F4.2 -5.6 and same external ND, in this case the HOYA ProND 16 or 64.

In may opinion it has nothing to do with the codec selected. It is possible to see banding on the URSA12K side screen.

How I tested.

First test:

URSA settings: ISO 800 - 180 shutter - 5600K 10 Tint WB - Codec Constant Bitrate 8:1 - Zebra set to 95% - use as a reference the BMD GEN5 Film to extended video lut - HOYA ProND 64 (6 stops)

BMPCC6K G1 settings: same as the URSA model using same lens an external ND

To expose I just adjusted the ND value and then lens iris to get the best looking image posible at both cameras displays. I mean best contrast without getting clipped shadows but avoiding a washed-out image.

Then download the BRAW files into resolve and check.

Once in Resolve, applied BMD GEN5 Film to extended video LUT and adjusted exposure / contrast / SAT.

Result: I can't see banding in the sky in both cameras BRAW files.


Second test:

URSA settings: ISO 400 - 180 shutter - 5600K 10 Tint WB - Codec Consant Bitrate 8:1 - Zebra set to 95% - use as a reference the BMD lut Film to extended video GEn5 - HOYA ProND 16 (4 stops)

BMPCC6K G1 settings: same as the URSA model using same lens an external ND

Same method used in first test to expose.

Once in Resolve, applied BMD GEN5 Film to extended video LUT and adjusted exposure / contrast / SAT.

I you use any of the filmstock LUTS available in Resolve it is easier to see it. I also tested with Filmbox plugin and Filmconvert plugin leaving the default settings for GEN5 Film and it is also easier to see the banding.


Result: I can't see banding in the sky on the BMPCC6k, but there is little banding in the URSA 12K OLPF, really a hint almost not visible


Third test:

URSA settings: ISO 200 - 180 shutter - 5600K 10 Tint WB - Codec Constant Bitrate 8:1 - Zebra set to 95% - use as a reference the BMD GEN5 Film to extended video LUT - HOYA ProND 16 (4 stops)

BMPCC6K G1 settings: same as the URSA model using same lens an external ND

Same method used in first test to expose.

Once in Resolve, applied BMD GEN5 Film to extended video LUT and adjusted exposure / contrast / SAT.


Result: I can't see banding in the sky on the BMPCC6k, but there is clear banding in the URSA 12K OLPF.


In my opinion is related to the amount of light that reach the sensor. If the exposure is more to the right the banding gets worst, even the exposure is far away from clipping at 95% Zebra.

By lowering the exposure (only with the lens iris or increasing ND value) the banding is not visible or it is not possible to detect it in the URSA 12k BRAW files.

In the BMPCC6k G1 camera, I was only able to see a hint of banding in blue sky (almost invisible) setting the ISO @100 and then expose the same way I did in the other tests.


I hope this info helps.

@diego So there is no banding at ISO 800, but as you lower ISO to expose the same scene, banding is introduced? Seems like banding starts to reveal at ISO400 and is evident at ISO200 in your tests... I can try to see if I can reproduce this in my tests.

Hi David,

Basically yes. The lower the ISO value you choose you will need to allow more light to enter thru the lens, otherwise the image you get will look really dark on camera's screen. This will push the exposure to the right. I my case, depending how much I push the exposure to the right (ZEBRA @95% no clipped highlights at all) the banding is less or more visible.

What it is also tricky is that it seems only to happen when you have a gradient blue sky. If the sky have more gradient the banding gets worst.

Here in Barcelona we have nice blue sky almost every day in this sesson

In the following pictures you can see same scene exposed with 3 different ISO values (125, 400, 800). No exposure adjustments in post.

It does seem like the banding becomes more noticeable as ISO lowers and exposure shifts to the right. Have you tested with different OLPF strengths, or is it consistent across the board? Also curious if a slight underexposure and lift in post could help mitigate it.
I’m actually planning to travel to Barcelona for a company project to capture photos of the city using my two cameras with OLPF, so this info is really helpful for getting the best results. I also came across this guide https://gowithguide.com/spain/barcelona, which should help me locate beautiful spots to capture. Excited to put all this knowledge to use in those clear Barcelona skies

Perhap's you have the same problem that me on my PC with the 12k LF.
Uncheck the "Use GPU for Blackmagic decode" in preference remove the banding for me

Uhm it’s not clear if you develop with lut or use color management or cts.
I often saw problem on smooth gradient using lut, the reason that I never want to use lut for de logging operation.


Inviato dal mio iPhone utilizzando Tapatalk

Hi guys!
This is an update. STILL no solution from BMD. They always ask for more test files. Sent them 2 times already. They confirmed there's banding but they need more tests (lower exposure, more lower exposure....). Doesn't make sense. Well, we'll see what will be the final result.

I was thinking of getting Ursa 12k Cine, but whit this spectacle in mind I'm not so sure....