# D600 High ISO in DX

Started 7 months ago | Questions thread
 Like?
 Re: pixel pitch and SNR In reply to Bobn2, 7 months ago

bobn2 wrote:

Leo360 wrote:

bobn2 wrote:

noirdesir wrote:

This is based on explanations by bobn, you'll have to ask him for details.

--
Bob

"In general for a planar structure such as a pixel, capacitance is proportional to area, therefore the read noise scales down with sensor area, as long as pixel area scales with sensor area, and that scaling is performed by uniformly scaling the pixel."

Which is exactly the opposite of what noirdesir stated citing you.

Leo

No it isn't. The article there is talking about the effect of different size ﻿sensors﻿ (I am perplexed why you, Roger and all the others of your persuasion cannot distinguish between the size of a sensor and the size of a pixel, and understand that if you want to observe how things change as a parameter changes, you need to change only that parameter). So the stated assumption for that discussion is that the sensor area falls, but pixel count remains the same, which means pixel size must get smaller ('as long as pixel area scales with sensor area, and that scaling is performed by uniformly scaling the pixel'). So, it discusses the effect of pixel size on read noise. The full quote is this:

The read noise is the total of all the electronic noises in the conversion chain for the pixels in the sensor array. To compare it with photon noise, it must be referred back to its equivalent in photoelectrons, which requires the division of the noise measured in volts by the conversion gain of the pixel. This is given, for an active pixel sensor, by the voltage at the input (gate) of the read transistor divided by the charge which generates that voltage, CG= Vrt/Qrt . This is the inverse of the capacitance of the read transistor gate (and the attached floating diffusion) since capacitance C= Q/V. Thus CG=1/Crt.﻿

In general for a planar structure such as a pixel, capacitance is proportional to area, therefore the read noise scales down with sensor area, as long as pixel area scales with sensor area, and that scaling is performed by uniformly scaling the pixel.

﻿This makes it clear that read noise scales down with pixel size.

--
Bob

Now my head is spinning and I feel dizzy. Please, correct me if I wrong. The larger the pixel the larger the read-transistor capacitance and, hence, the smaller the Conversion Gain (CG). And smaller CG means MORE read-noise (the same voltage fluctuations result in more noise electorns). Then, disabuse me from my misunderstanding, why you (and Wikipedia) say "read noise scales down" rather then "scales up" with pixel size??

Leo

Complain
Post ()
Keyboard shortcuts: