ISO is a numerical value that represents how sensitive your camera is to light. A higher number like ISO 3200 makes your camera more sensitive, letting you shoot in darker conditions, while a lower number like ISO 100 requires more light but produces cleaner images. The term comes from the International Organization for Standardization, which created the unified scale used in every camera today.
How ISO Works in a Digital Camera
Every digital camera sensor captures light by converting photons (particles of light) into an electrical signal. When you raise the ISO setting, you’re telling the camera to amplify that electrical signal, making a dim scene appear brighter. The tradeoff is noise: amplifying the signal also amplifies the random electrical interference that exists in every sensor, producing the grainy, speckled look you see in photos taken at high ISO values.
Your camera’s base ISO is its lowest native setting, typically ISO 100 or ISO 200 depending on the model. At base ISO, the sensor operates with zero amplification, which means noise is at its absolute lowest and image quality is at its peak. Every step up from there involves progressively more amplification and progressively more visible grain.
The Scale: Linear and Logarithmic
The ISO system actually defines two scales. The one most photographers use is the arithmetic (linear) scale, where the numbers work exactly how you’d expect: ISO 200 is twice as sensitive as ISO 100, ISO 400 is twice as sensitive as ISO 200, and so on. Each doubling corresponds to one “stop” of exposure, meaning the sensor needs half as much light to produce the same brightness.
There’s also a logarithmic scale, inherited from the old European DIN system, where adding 3 degrees to the number represents a doubling of sensitivity. A film rated ISO 100/21° has identical sensitivity whether you read the linear value (100) or the logarithmic value (21°). In practice, almost no one uses the logarithmic notation anymore, but it’s still technically part of the standard.
How ISO Is Actually Measured
Assigning a specific ISO number to a sensor isn’t arbitrary. The standard uses a concept called saturation-based speed, which is especially relevant for digital sensors. The idea is straightforward: engineers measure how much light exposure it takes to fill a pixel to its maximum capacity without overflowing (called clipping). The formula is Ssat = 78 / Hsat, where Hsat is that minimum exposure at the sensor’s surface that pushes the output signal to its valid maximum.
Two physical properties of the sensor largely determine its base ISO value. The first is quantum efficiency, which measures how good the sensor is at converting incoming photons into electrons. A sensor with 50% quantum efficiency turns 100 photons into 50 electrons. Some high-end sensors designed for scientific imaging reach quantum efficiency as high as 95% at certain wavelengths. The second property is full well capacity, which is simply how many electrons a single pixel can hold before it saturates. A pixel that can store more electrons before overflowing has a wider range to work with, which translates to better dynamic range and lower base ISO.
Where the ISO Scale Came From
Before the unified ISO system existed, photographers in different countries used completely different scales. In the United States, the standard was ASA (American Standards Association), first defined in 1943 based on research at Kodak. ASA used a linear scale: ASA 200 film was twice as fast as ASA 100 film. In Europe, the dominant system was DIN (Deutsches Institut für Normung), published in 1934, which used a logarithmic scale where every increase of 3 degrees meant double the sensitivity.
The two systems were formally merged into the ISO standard starting in 1974, with full international adoption rolling out between 1982 and 1987. The merger was elegant: the arithmetic ASA scale simply became the linear ISO value, and the DIN scale became the logarithmic ISO value, joined with a slash. So ASA 100 film and DIN 21° film became ISO 100/21°. The linear number won out in everyday use, which is why modern camera displays just show “ISO 400” rather than the full dual notation.
One quirk of the historical transition: in 1960, the ASA standard dropped previously built-in safety factors against underexposure, which effectively doubled the rated speed of many black-and-white films overnight. A film’s actual sensitivity hadn’t changed, but its official number had. The DIN system made a corresponding adjustment to stay aligned. This is a useful reminder that ISO numbers aren’t a pure physical measurement. They’re a standardized convention designed to give photographers consistent, predictable exposure.
Signal, Noise, and What the Numbers Mean in Practice
The relationship between signal and noise is central to understanding why ISO matters. Every sensor produces some amount of unwanted noise from several sources. Read noise is the electrical interference introduced every time the sensor converts its analog signal to a digital one. Modern scientific-grade sensors have pushed read noise as low as plus or minus 1 electron per pixel, while older sensor designs produce 5 to 10 electrons of read noise per readout. Dark current is thermal noise generated by the sensor even when no light hits it, measured in electrons per pixel per second. Cooling the sensor reduces dark current, which is why astrophotography cameras and laboratory microscopes often use cooled sensors.
Then there’s photon shot noise, an unavoidable source of randomness that comes from the nature of light itself. It follows a square root relationship: if a pixel collects 100 photons worth of signal, the shot noise is roughly 10 (the square root of 100). This means that collecting more light always improves the signal-to-noise ratio. Doubling your exposure from ISO 1600 to ISO 800 (which requires twice as much light) doesn’t just halve the amplification noise. It also collects more photons, pushing the ratio of useful signal to random noise in your favor.
Pixel size plays a role too. A larger pixel has more surface area to collect light. An 11-micrometer pixel has about six times the area of a 4.5-micrometer pixel, so it gathers substantially more photons in the same exposure time. This is why cameras with larger pixels (often found in full-frame bodies) tend to handle high ISO settings better than cameras with smaller, more densely packed pixels.
ISO Beyond Consumer Cameras
In scientific and industrial imaging, sensitivity is described with more precision than a single ISO number can provide. Researchers working with microscopes, medical imagers, or machine vision systems care about quantum efficiency at specific wavelengths, read noise in electrons, and dark current rates, because these individual measurements tell them exactly how the sensor will perform in their particular application. Camera sensitivity in these fields can’t be captured by one value alone, since it depends on the wavelength of light being detected, the exposure duration, the pixel size, and the sensor architecture.
Some scientific sensors use a technique called binning, which combines the output of neighboring pixels into one larger “super pixel.” On a CCD sensor, binning happens before the signal is digitized, so combining a 2×2 block of pixels quadruples the signal-to-noise ratio. On a CMOS sensor, binning happens after each pixel has already been read out individually, so the same 2×2 block only doubles the signal-to-noise ratio because read noise has already been introduced. These distinctions matter in fields like fluorescence microscopy, where detecting a handful of photons from a faintly glowing cell can mean the difference between a usable image and a blank frame.