Black Magic PocketHD and 4K production camera

I agree with Harry about the concerns you need to have with rolling shutters. It needs to be an "obsession", if by that you mean paying close attention to it.

BM claim the effect will be minor on the Pocket, but we'll see.
 
I agree with Harry about the concerns you need to have with rolling shutters. It needs to be an "obsession", if by that you mean paying close attention to it.

BM claim the effect will be minor on the Pocket, but we'll see.

Minor as in slightly better than the BMCC due to reading out a smaller area of the sensor in the same time frame.
 
It should be understood that John Brawley measured the 4k and found it had about 1 1/2 stops less dynamic range than the 2.5k BMCC. I still want the 2.5k version. The IQ in hard sunlight depends on it.
 
Yeah, it can be confusing..when there are other cameras that have gone for CCD, one of the primary reasons being global shutter. What BMC decided to do was use CMOS again...but with global read out, at a sacrifice of 1 stop dynamic range.
 
It should be understood that John Brawley measured the 4k and found it had about 1 1/2 stops less dynamic range than the 2.5k BMCC. I still want the 2.5k version. The IQ in hard sunlight depends on it.

What would be brilliant is if there was an option to switch the sensor read between global and rolling shutter mode. I see no reason why this couldn't be done in a firmware upgrade, if the sensor supports this option.

That way you could use the global shutter for pans and action work and switch to rolling shutter when capturing maximum dynamic range is the priority.

A drop of 1.5 stops is pretty significant. In the real world that probably puts the camera at around 11 stops. Not a disaster, but I think it will be noticeable. There is something about footage that hits that +13 stop mark. It suddenly stops looking like really good video and more organic, like film which gets a solid 14 stops. Compare footage from a RED One or Sony f35 to the Alexa or new Sony bodies (F3/F5/F55/F65).
 
What would be brilliant is if there was an option to switch the sensor read between global and rolling shutter mode. I see no reason why this couldn't be done in a firmware upgrade, if the sensor supports this option.

That way you could use the global shutter for pans and action work and switch to rolling shutter when capturing maximum dynamic range is the priority.

A drop of 1.5 stops is pretty significant. In the real world that probably puts the camera at around 11 stops. Not a disaster, but I think it will be noticeable. There is something about footage that hits that +13 stop mark. It suddenly stops looking like really good video and more organic, like film which gets a solid 14 stops. Compare footage from a RED One or Sony f35 to the Alexa or new Sony bodies (F3/F5/F55/F65).
While I suppose anything is theoretically possible, if the 4K camera's sensor is the CMOSIS CMV12000, then it has no rolling-shutter mode. And it's rated at 60 dB without using any of its three HDR modes, or about 10 stops maximally. The Apertus Axiom, if and when it's released, will use the same sensor and is planned to have an adjustable dynamic range, from 10 to 15 stops. Don't get too excited, though. Apertus admits that the CMV12000's pseudo log response curve -- the HDR mode they'll be using -- has trade-offs (see here).
 
What would be brilliant is if there was an option to switch the sensor read between global and rolling shutter mode. I see no reason why this couldn't be done in a firmware upgrade, if the sensor supports this option.

That way you could use the global shutter for pans and action work and switch to rolling shutter when capturing maximum dynamic range

+1
 
While I suppose anything is theoretically possible, if the 4K camera's sensor is the CMOSIS CMV12000, then it has no rolling-shutter mode. And it's rated at 60 dB without using any of its three HDR modes, or about 10 stops maximally. The Apertus Axiom, if and when it's released, will use the same sensor and is planned to have an adjustable dynamic range, from 10 to 15 stops. Don't get too excited, though. Apertus admits that the CMV12000's pseudo log response curve -- the HDR mode they'll be using -- has trade-offs (see here).


from the site given above:

Smart Dynamic Range

Imagine your camera had a knob for dynamic range like the one it has for shutter speed. You can turn it up or down just as you need it. Sounds crazy doesn't it? That's exactly what this sensor can do. While in general the notion is "the higher the dynamic range the better" you need to be aware that you sacrifice details (digital bandwidth) in shadows/midtones for gains in latitude when you increase dynamic range. This could be undesired in situations like when dealing with fine skin tones or when the scene does not require a high dynamic range at all e.g. when shooting well lit indoor scenes. Smart Dynamic Range for us means that you can dynamically influence the response curve for your particular situation in a range from 10 - 15 stops of dynamic range.


High Dynamic Range

The image sensor supports three different High Dynamic Range (HDR) modes.

One mode alters the sensors photosite response curve from a linear default behavior to a shape that is closer to a logarithmic response. This happens before the ADC so no digital data is sacrificed for the additional dynamic range.

The second mode uses different exposure times for even and odd pixel coloumns. So two exposure brackets can be captured at the same time for sacrificing horizontal resolution.

The third mode is possible due to the high speed nature of the sensor allowing alternating exposure times with each recorded frame and doubling/tripling/etc. frame rate result in double/triple/etc. the amount of data to be captured.

The response curve processing happens before the ADC so no digital data is sacrificed for the additional dynamic range.
 
Last edited:
Sorry, I should have been more specific. We tested the F35 at about 10-11 stops, but you are correct. Regardless the camera does produce a very organic looking image. I edited my original post.
 
Last edited:
Hmm.. I compared Cineform debayering to Resolve debayering and found Cineform only marginal softer.
Which of the debayering algorythms of Cineform have you used?
And much more important, are you sure the little "C" ikon on the computer of your VFX guys wasn't set to "Auto".
It makes a huge difference.

I'm with Frank. Haven't seen such a huge quality difference either.

In my experience, CineformRAW at quality 5 = Filmscan1 is extremely close to raw DNGs, except for the most critical work like green screen etc.

I get CineformRAW from BMCC raw DNGs using my RAW 2 CF RAW utility. It also merges audio automatically into CineformRAW video file.
 
Back
Top