hello eveyone
i always read that smaller sensor has poor low light performance compared to bigger sensor (assuming sensor technology is of the same generation) which seems logical because the surface to collect light energy is smaller
can someone explain why in this test the low light performance of the pocket camera with a 12.48x7.02mm
sensor on a f1.8*0.71= f1.28 is better (less noise) than the 5d mark3 on a f1.4 lens (almost similar aperture) but 10x smaller sensor ?
the ratio of sensor surface is 36*24/(12.4*7) = 9.9 should not the image on the 5d mark 3 be 3.2 stop brighter (ln2(9.9)-(1.4-1.28)) and thus have less noise as explained here (http://www.clarkvision.com/articles/does.pixel.size.matter/#Introduction)
i also read that it is pixel size that determines the s/n ratio and not the sensor size . so pocket cinema camera has a pixel width of 12.48mm/1920 = 6.5um and 5dmark3 has a pixel width of 36mm/5760=6.2um so it should have similar noise ratio. but then if the 5dmak3 is multisampled by a factor of 3 ( 5760/1920) on each axis,when we average the 9 pixels, the noise level must be dived by 9. so why does this test shows bmpcc in low light with less noise than the 5dmk3 ? was there a noise reduction done in post process for the bmpcc and not for the 5dmk3 ?
thank you