High ISO Noise and pixel area

[daveleo]

I ran into a buddy with a Nikon D600 (a "full frame" sensor), while I had my Fuji X100 on my shoulder. I gasped at the size and weight of his setup (it is HUGE!), he scoffed at the size of the X100. The discussion moved away from weight, cost and size and onto high ISO noise. He claimed the D600 was unbeatable on this. I said the X100 was excellent.

My question now is technical. * Assuming that all other sensor design characteristics are perfectly equal *, isn't high ISO noise simply related to pixel size (area)? The D600 sensor area is 2.35X that of the X100 sensor, but it has only 2X the pixel count, so it's pixels are about 19% larger than the X100 pixels. Thus (I am concluding) the D600 inherently has something on the order of only 19% less noise than the X100.

I hate these pixel debates, but he bugged me too much to forget about it.

Thanks for listening, and remember that you'll really hurt my feelings if you tell me that I'm wrong. 🙄

 

[willie_901]

Dave,

It's helpful to think about the signal-to-noise ratio, SNR, instead of the high ISO noise. The sensor read (electronic) noise level is essentially constant at all ISO settings. What decreases as ISO increases is the signal level. The read-noise level increase due to ISO electronic amplification is much less than the sensor read noise. Increasing ISO does not increase the noise. This is the case for the X100 and D600, and for many contemporary cameras with CMOS sensors. There are exceptions (M8/M9 and some Canon DSLRs).

Increasing ISO decreases the SNR because the meter calibration changes so we use faster shutter times and, or narrower apertures. Less light reaches the sensor when the shutter is open. After the shutter closes the signal is amplified so the brightness matches the meter's prediction. Shutter and aperture determine exposure (signal level) and ISO affects post-acquisition brightness.

More sensor area means more signal. This is the primary advantage of the D600.

Pixel size does play a role too. Larger pixels usually have more SNR on an individual pixel basis. But this only part of the story.

Here is a quote from Cambridge In Color.

"This is not always the case however, because the amount of background noise also depends on sensor manufacturing process and how efficiently the camera extracts tonal information from each pixel (without introducing additional noise). In general though, the above trend holds true. Another aspect to consider is that even if two sensors have the same apparent noise when viewed at 100%, the sensor with the higher pixel count will produce a cleaner looking final print. This is because the noise gets enlarged less for the higher pixel count sensor (for a given print size), therefore this noise has a higher frequency and thus appears finer grained.

Another component in the signal level is the lens transmission factor. The T-factor compares for light loss due to optical design and lens glass differences between lenses. So two lenses with the same focal lengths and maximum apertures can deliver different amounts of light (signal) if they have different T-factors.

I owned a X100. There is no doubt the D600 SNR is greater. But newer sensor and ADC electronics, and surface area are probably more important than pixel size.

For my everyday carry I would never use a D600. Its weight and size means I would leave the camera at home much more than the X100. At the same time, if I was shooting sports or theatre I would use the D600.

The issue isn't just SNR, the issue is what's the best overall tool for the task at hand.

 

[daveleo]

Hey, thanks for a (typically) excellent answer. I felt that I was oversimplifying the issue. Time to expand my understanding of the topic, (for no GAS-driven reason - I love the gear that I have).

Your answer is very much appreciated (as usual).