Sensitivity equivalence:

So now, the more contentious issue of sensitivity or ISO equivalence.

Here we have all four cameras set to the same f-number, the same ISO and shutter speed (F2, ISO 200, 1/1.3sec). We're using a low light scene because this is where equivalence is most likely to be able to predict performance. By comparison, in good light, the large sensor's ability to commit more surface area to each region of the image will tend to give cleaner images, without any dependence on aperture.

Full Frame
Four Thirds
1" Type

You won't be surprised to see that the camera with the largest sensor does better. The question is, does it do better in proportion to how much larger it is?

Equivalent exposures

Here we have the cameras set to 'equivalent' apertures (so that the images look the same in terms of depth-of-field, as well as framing), and the ISO setting required to produce a similarly bright image from the same amount of total light. To calculate this, we've used the ratio of the surface areas (though you could also use the square of the crop factor, if you preferred).

In this case we used a shorter exposure and higher ISO to ensure that the noise was visible in the final images.

Lens Selected equivalent aperture F-numbers equivalent to F5.6 on full frame Closest available F-number ISO required to use 1/1.3 shutter speed Nearest available ISO Difference from ideal
Canon EOS-1D X f/5.6 f/5.6 f/5.6 2975 3200 +0.1EV
Fujifilm X-A1 f/5.6 f/3.7 f/3.6 1268 1250 +0.1EV
Panasonic GH4 f/5.6 f/2.8 f/2.8 774 800 0EV
Nikon 1 V3 f/5.6 f/2.1 f/2.0 400 400 +0.1EV
Full Frame
Four Thirds
1" Type

As you can see, while the noise levels are pretty similar, they're not quite the same (a simplistic measure of standard deviation from a common patch suggests the noise isn't solely related to sensor size, either).

There are a number of factors in play here, most likely dominated by differences in sensor performance, but with other complications such as differences in transmission between lenses. This is why we tend to state that, although equivalence can be used as a guide to low light performance (and a very good guide in most instances), it can't be used to completely predict the differences.

Why should I care?

The idea of equivalence gives you the tools to understand the difference in performance between different camera formats in terms of depth-of-field. It also lets you relate the role of F-number and sensor size to how much light the camera will have access to, which is an indicator of likely low-light performance.

Ultimately, though, it's unlikely to make much difference to your day-to-day shooting: you can't try to flood a small sensor with extra light in order to get better images, because it'll just overexpose. But it's a useful thing to understand, when considering the size/price/image quality/depth-of-field trade-offs that you make, whenever you buy a new piece of camera equipment. There's no universally ideal 'sweet-spot' to that trade-off: but understanding equivalence can help you work out which balance is best for you.

Key points:

• In low light, equivalent aperture can help you get some idea of noise performance and where the aperture of one camera can help offset any sensor size difference.

• Differences in sensor and lens characteristics mean that, while it gives a very good guide, equivalence doesn't give a precise prediction of performance.