This is somewhat related to former posts, but I'd rather start a new discussion.

An HDR capture mode would be a revolutionary feature.

And I am not only talking about the applications in CG. Having that extra range at hand when it comes to color grading and regular postfixing is an incredible advancement. Yes, your sensor is catching up with film already, in delivering 11 stops of latitude, I believe. But why stop there? Go further, and outperform film!

I agree, that multiple captures in time open up a can of worms, when it comes to capturing fast motions. But with 4k you have plenty of spatial resoltion capture 2k HDRI footage. You can use a prism beamsplitter, that is coated with incrementally increasing ND filters. A very simple solution, because on the hardware side it would be not more complicated than any other ND filter that can pop in front of the sensor. Here is a McGyver-style prototype: http://gl.ict.usc.edu/research/rtlp/index.html

Of course, it would make a lot of sense to have EXR as direct capture format then. You would get great compression ratios, a wide color gamut and plenty of color precision. You would not need to invent any of this for your RAW format, and your user had a wide open choice of direct postprocessing software. You would play well with others. And it would take away the burden of delivering the ultimate software with the ultimate camera.

And if you really consider implementing the DepthMap idea, then EXR is just right for you. It sports a Z channel, that is already supported by a wide variety of software. You could even go as far as having an on-board depth selector, that simply cuts the Z Channel at a user defined threshold into black and white. That would be an instant Alpha, that could be embedded in EXR frames, too. And streamed out to the on-set control monitor for live composites.


best regards,
Christian Bloch