Lesson 9: ISO — Focal Point

Chelsea London © 2017 | Fujifilm X-E1 | 27.0 mm | ƒ/2.8 | 1/2000s | ISO 6400

Introduction to ISO

In this lesson, we will tackle the last of the three exposure controls: the ISO, also sometimes called sensitivity (though this is a misnomer). Once you have mastered shutter speed, aperture, and ISO, you will know 90% of what you need to know to create (technically) good images that reflect your vision.

ISO is the only one of the exposure controls where there is a fundamental difference between film and digital. In film, ISO is a physical property of the film you are using, and the only way to modify it is to change to a new roll. With digital you can easily change ISO between shots simply by turning a wheel, hitting a button, or digging into a menu. This action allows for adaptation to the current light conditions. For those who shot film a long time ago, you may have used different words for sensitivity: ASA or din. The first is exactly the same as our current ISO - it simply changed names when it became standardised. The latter uses another logarithmic scale and is completely outdated.

Like shutter speed and unlike aperture, ISO is a linear value. This makes it easier to determine a stop - simply a doubling of the ISO value. Let’s assume you keep the same shutter speed and aperture, are shooting at ISO 800, and want one stop of underexposure, you would go to ISO 400. If you want one stop of overexposure, you should go to ISO 1600.

A very common misconception is that increasing the ISO in digital cameras increases the sensitivity of the camera/sensor. Unfortunately, this isn’t possible, as sensor sensitivity is a physical property baked into its construction. Changing ISO changes the amount of amplification applied to the initial voltage produced by a photosite on the sensor when a photon hits it. Or, in simpler terms, adjusting ISO basically alters how much the camera boosts the initial signal created by light hitting the sensor. The base ISO of a sensor (generally ISO 100) would have no additional amplification applied to it, while ISO values higher than this have amplification applied to it.

Sean Makin © 2018 | Nikon D610 | 24.0mm | ƒ/1.8 | 13s | ISO 6400

Noise

You have heard that increasing ISO makes your images noisier, right? This is another common misconception that appears to be true if you look at two images taken at significantly different ISO settings in different lighting environments. But the real answer is a little more nuanced and technical.

Noise in the simplest form is related to light, or lack thereof. Because sensors are not perfect there are many different noise sources including read noise, dark current, hot pixels, photon noise, shot noise, fixed pattern noise, and color mottle. We won’t explore these terms any further as this goes well beyond the limits of a beginner photo class. When we increase the amount of light on the sensor, the number of electrons hitting the photosites on the sensor increases. This, in turn, increases the amount of current produced by the transistor. The increase in the signal produced by these transistors increases the signal-to-noise ratio (SNR), meaning we can swamp the noise produced by the sources listed above and end up with a cleaner, less noisy image.

Test scene with fixed shutter speed, exposure is balanced during postprocessing to achieve equal brightness. No noise reduction or sharpening applied.  Click here for the full-resolution image.

The idea that increasing ISO causes more noise arises when you compare pictures taken in various lighting conditions, with the same shutter speed and aperture settings, to achieve a well-exposed image. For instance, if you use settings like 1/250 and f/4, and take one photo during the day at ISO 100 and another in the evening at ISO 1600, you're letting in four stops less light. This reduction in light means there's less signal, resulting in a lower signal-to-noise ratio and more noticeable sensor noise in the images. This can be seen in the example below.

Test scene where shutter speed is varied to achieve balanced exposure. No noise reduction or sharpening applied . Click here for the full-resolution photo.

Let’s consider a different situation where you take two photos in the same lighting conditions. In the first photo, your settings are 1/250 shutter speed, f/4 aperture, and ISO 400. However, this isn't fast enough to capture fast movement. So, you decide to increase the shutter speed to 1/1000, keeping the aperture at f/4. To keep the image brightness the same, you also increase the ISO to 1600.

Now, the question is, which change in settings actually leads to more noise in the picture? Surprisingly, it's the 2-stop increase in shutter speed. This is because it lets less light reach the camera sensor, and that reduction in light is what causes more noise in the image, not the higher ISO.

Chelsea London © 2018 | Fujifilm X-T10 | 56.0mm | ƒ/2.0 | 1/60s | ISO 6400

What does this mean for me as a photographer?

Setting aside the technical details, as photographers, we often find ourselves restricted by a minimum shutter speed (to avoid blur) and a chosen aperture size (based on our lens and desired depth of field), which determines the amount of light we can capture. When we want a well-exposed image, we might need to increase the amplification of the signal on our camera's sensor, (e.g. increasing ISO).

It's useful to associate higher ISO with more noise, as a higher ISO is typically needed in darker environments, where less light is available, leading to more noticeable noise. Because of this, photographers usually have a set of ISO values for their cameras, including a base ISO, the first ISO where noise becomes apparent, the highest acceptable ISO for good quality (which is crucial), and the maximum ISO they're willing to use in emergency situations.