A Quick Detour into Signal Amplification

Last time we talked about going to higher and higher ISO's. How does that work exactly?

Roughly put, the sensor at it's most basic level is a device that records whether or not a photon hit it, and if so, how many did? (We'll talk about color a little later.)

Basically when a bunch of photons hit the sensor element, it converts it into a buncha electrons. This charge is then amplified, and converted into a voltage. Measuring the voltage tells you how many photons must've hit the device in the first place.

The sensor is nothing more than a vast array of sensor elements, and the way it records colors is just by packing a tight array of red-green-blue sensors very close together, and then reverse-figuring out what the color must've been. Quite primitive in a certain way.

The key word in the last mumbo-jumbo is "amplification". A small trickle of electrons has to be converted into a flood of electrons, and then measured. This is where all the problems begin.

Firstly these sensors are all sitting on the same substrate of silicon (typically.) So a stray photon (or electron) could easily make its way onto the sensor even if it were not present originally. There are other quantum-mechanical effects that would make this happen even if things were "ideal".

So when you "amplify" the number of photons, you not only amplify the "real" photons that you wanted to but you also amplify the "noise" photons that you had no intention to.

Electrical engineers have a succint expression for this — When you amplify the signal, you also amplify the noise.

If you amplify enough, you will no longer be able to distinguish between the signal and the noise, or in plain words, your picture will be filled with garbage photons. It will look spotty like someone sprinkled random colors all over it.

There are ways to address this, of course, and that's what engineers are hired for but just "cranking the ISO" is not really a solution to your photography problems.

And that's why, muchachos y muchachas, you need to read this blog!

ISO, You Saw, We All Saw!

The last piece in the number of photons puzzle is the actual "photograph" itself.

A photograph is nothing more than a recording of photons. In the bad old days, you had film. These days you have a digital sensor.

The term ISO dates back to film. It has been co-opted in the modern digital world.

Effectively, it tells you the film's sensitivity to light. (In the modern world, digital sensor's sensitivity to light.)

The designers were well aware of the fact that light strikes the eye logarithmically so twice the ISO meant twice the speed.

Now digital sensors are basically the same but they allow going up to far higher ISO speeds than bad old film ever could.

Theoretically, going to higher and higher ISO's means that you are effectively increasing the "number of photons". However, practically it doesn't work that way.

That's for a future post though.