Question @dfort, would the 5D Mark II image look more competitive and have reduced noise if we compared FHD 3x3 1856x1248 vs the BMPCC4K.
No. The comparison here with the 5D2 using raw 10-bit 2560x1200, 1x1 sampling should be more competitive than 3x3 sampling. In addition, you'll get aliasing with 3x3 sampling. Things will get much better when (if?) we can get 1x3 sampling on the 5D2.
To make a long story short, What is the pixel durability of a FULL FRAME Sensor Pixel VS a Super 35 Crop Sensor? I've may have missed this discussion in the quality of a FULL FRAME ML FHD RAW sensor pixel durability.
Pixel durability? Had to Google that one and found some phone drop tests. What are you talking about?
As far as the full frame vs. a Super 35 crop sensor -- the BMPCC 4k has a Micro 4/3 sensor. The 5D2 in 2560x1200, 1x1 sampling mode is probably using even less real estate than Micro 4/3. The Micro 4/3 crop factor is 2.0, right? I believe that crop_rec setting is something like a 3x even though you need to put the camera in 5x zoom mode. In any case, you aren't using the full sensor.
However, this does bring up a good point. Why does a full frame sensor have a more pleasing "look" than the smaller sensor cameras? The cine lens makers have been jumping on the full frame bandwagon lead by the camera manufactures --
Alexa LF, the
RED Monstro,
Sony VENICE,
Panavision DXL2 and -- ok, let's get back to the ML universe. The new crop_rec settings will give you greater resolution but there's something about the look of a full frame sensor that goes beyond pixel peeping.
