Most photographers grow up believing a simple rule: keep ISO as low as possible, because raising it adds noise. That rule is true — but it is also incomplete. On many modern cameras, brightening an image by raising ISO in the camera and brightening it later with the exposure slider in Lightroom or Capture One produce nearly identical amounts of noise. That property is called ISO invariance, and it quietly changes the smartest way to expose in low light.

What Raising ISO Actually Does

Aperture and shutter speed control how much light physically reaches the sensor. ISO does not. Turning up the ISO simply amplifies the electrical signal the sensor has already captured — picture the gain knob on an amplifier. More gain makes the picture brighter, but it also boosts whatever noise is already sitting in that signal. The question ISO invariance asks is straightforward: does the camera’s own gain stage add extra noise on top, or is the signal already clean enough that amplifying it later in software gives the same result?

The Two Kinds of Noise

To understand invariance you have to separate two sources of noise. Photon shot noise is quantum randomness — the statistical variation in how many photons actually arrive at each pixel. It scales with the square root of the light gathered, and it dominates most photographs. No ISO strategy removes shot noise; only collecting more light does (longer exposure, wider aperture, or a larger sensor). Read noise is electronic, added by the sensor’s readout circuitry and the analog-to-digital converter (ADC). It is roughly constant regardless of how much light you captured.

When read noise is significant, the camera’s analog gain (a higher ISO) can bury it under the amplified signal, leaving cleaner shadows than if you had shot at base ISO and brightened the file in post. When read noise is already tiny, that gain step adds nothing meaningful — and the two approaches converge.

What “ISO Invariant” Actually Means

A sensor is ISO invariant when its read noise is so low that raising ISO in the camera and brightening in post are effectively interchangeable. Shoot a scene at ISO 100 and push it four stops in your raw editor, and it looks essentially identical to shooting the same scene at ISO 1600 with the same aperture and shutter speed.

This is why modern sensors from Sony, Nikon, and Canon behave this way above a certain threshold — typically somewhere around ISO 400 to 800, where dual-gain readout circuitry engages and read noise drops. Canon’s sensors were historically less invariant (the 5D Mark III and 6D era showed visibly more shadow noise when pushed in post), but recent bodies have largely closed that gap.

How to Tell If Your Camera Is ISO Invariant

The definitive reference is PhotonstoPhotos.net, which publishes read-noise and “Shadow Improvement” charts for hundreds of camera models. A flat line means invariance: raising ISO adds no shadow-noise benefit. A line that keeps dropping as ISO climbs means your camera genuinely benefits from in-camera gain.

You can also test it yourself in a few minutes. Mount the camera on a tripod, choose a fixed aperture and shutter speed, and shoot one raw file at ISO 100 and another at ISO 1600. In your raw editor, brighten the ISO 100 file by four stops. If the shadows look the same, your camera is effectively invariant across that range — and you can safely trade ISO for highlight protection.

When This Changes How You Shoot

On an invariant camera you can deliberately underexpose to protect highlights and recover the shadows later without a noise penalty — effectively trading ISO headroom for highlight headroom. For scenes with bright light sources and deep shadows (a concert stage, a sunset, a room with a bright window), set your exposure for the highlights, keep ISO low, and lift the dark areas in post.

Dual-gain sensors add a useful twist: there is often a second “sweet spot” around ISO 640 to 800 where a second gain stage drops read noise again. On some bodies, ISO 800 can actually look slightly cleaner than ISO 500. Check your camera’s chart — the inflection points are where you want to be.

Limits and Caveats

ISO invariance has clear boundaries. It applies to raw files only — a JPEG has its tone curve and noise reduction baked in at capture, so the trick does not transfer. Extended ISOs (pull and push settings such as ISO 50 or ISO 102400) are digitally simulated and behave differently. And below a camera’s invariance threshold, raising ISO in the camera genuinely produces cleaner shadows than pushing a base-ISO file, so base ISO is not always the right choice when you cannot add light. Finally, none of this rescues a scene starved of light — once shot noise dominates, the only fix is more photons.

Conclusion

ISO invariance does not mean “ISO is meaningless.” It means the old advice to always keep ISO low has a modern caveat. On today’s low-read-noise sensors, protecting highlights by underexposing and pushing in post is often free, and knowing your camera’s invariance threshold — and its dual-gain point — lets you choose the cleanest ISO rather than blindly the lowest one. Look up your model on PhotonstoPhotos.net, shoot raw, and treat ISO as one more creative tool instead of an enemy.

FAQ

Is a lower ISO always cleaner? Not necessarily. Below a camera’s invariance threshold, in-camera ISO can clean up shadows; above it, brightening in post is equivalent to raising ISO. On dual-gain sensors, a mid-range ISO can look cleaner than a slightly lower one.

Does ISO invariance apply to JPEG? No. JPEGs have their tone curve and processing baked in at the moment of capture, so you need to set ISO correctly at the time of shooting. ISO invariance is a raw-file advantage.

How do I find my camera’s invariance threshold? Look up your model’s read-noise or Shadow Improvement chart on PhotonstoPhotos.net, or run the ISO 100 versus ISO 1600 (plus four stops) push test described above.