You learned in the exposure course that high ISO settings bring more noise, and how to reduce it. That is a useful rule, but the full story is more interesting, and it changes how you shoot in low light. Most noise actually comes from collecting too little light, not from the ISO setting itself. This lesson explains where noise really comes from, how cameras differ at high ISO, and how to find the highest ISO you can comfortably use.
Noise, briefly
Digital noise is random variation in brightness and color that appears as grain or speckles, most visible in dark, smooth areas. You met luminance and color noise in reducing noise and protecting image quality. This lesson looks at the causes from the sensor's point of view.
The two main sources of noise
| Photon shot noise | Read noise | |
|---|---|---|
| Comes from | Light itself: photons arrive randomly, so counts vary naturally | The camera's electronics when the signal is read and converted |
| When it dominates | In most normal and low-light photos | In the deepest shadows and very dark exposures |
| How to reduce it | Collect more light | Better sensor technology; appropriate ISO |
Shot noise is part of nature: even a perfect sensor would have it. The more light you collect, the smaller it becomes compared with the signal. This is why the amount of light reaching the sensor matters most, a key idea from camera sensors and digital image capture.
Light, not ISO, is the main cause
When you raise ISO in a dark scene, you usually also use a faster shutter speed or smaller aperture, so less light reaches the sensor. The photo has more noise mainly because it collected less light. ISO then amplifies that weaker signal to look bright.
Tip:Practical result: to reduce noise, first look for ways to collect more light, such as a wider aperture, a slower shutter speed (if the subject and support allow), more light on the subject, or flash. Raising ISO is the last step, not the cause of the problem.
| Option | Trade-off | Lesson |
|---|---|---|
| Wider aperture | Less depth of field | aperture |
| Slower shutter speed | Risk of shake or motion blur | shutter speed |
| Add or move to more light | May change the mood | understanding flash |
| Then raise ISO | Amplifies the remaining noise | ISO |
ISO invariance
Many modern sensors have very low read noise. On these cameras, a RAW photo taken at a low ISO and brightened later in editing looks almost the same as one taken at a higher ISO with the same aperture and shutter speed. This is called ISO invariance.
Remember:It is useful in very high-contrast night scenes, where a lower ISO protects bright lights from clipping while shadows are brightened later. But it applies to RAW files, it varies between cameras, and many sensors with dual conversion gain are not invariant across the ISO where they switch modes, as explained in dynamic range and highlight/shadow information. For most photography, simply exposing well is still the best approach.
Comparing high-ISO performance fairly
- Compare at the same final size: viewing a high-resolution image at 100% exaggerates its noise compared with a lower-resolution image.
- Larger sensors collect more total light at the same exposure, so they usually look cleaner, as in sensor formats.
- Sensor generation matters: newer designs often outperform older, larger ones.
- Processing matters: camera JPEGs may hide noise with smoothing; compare RAW files for a fair test.
"Bigger pixels always mean less noise" is an oversimplification. When images are compared at the same size, total light collected by the whole sensor matters more than the size of individual pixels.
Finding your usable ISO
- Photograph the same low-light scene at increasing ISO settings, correctly exposed each time.
- Process them as you normally would, including your usual noise reduction.
- View at the size you normally use, such as a full screen or a print size.
- Note the highest ISO you are happy with for different uses: web, prints, client work.
- Use it as your Auto ISO limit, as explained in ISO.
Modern noise reduction, including AI-based tools, can remove much more noise than older methods, covered in noise reduction. This raises your usable ISO, but cannot recover detail that was never captured.
Common misunderstandings
- "ISO causes noise": lack of light is the main cause; ISO amplifies it.
- Underexposing to avoid high ISO: brightening later usually gives as much or more noise.
- Judging noise at 100% zoom only: compare at real viewing size.
- Fearing high ISO so much that photos blur: a sharp, slightly noisy photo beats a blurry clean one.
How professionals handle low light
Event, wedding and concert photographers use fast lenses and good technique to collect as much light as possible, then use whatever ISO keeps shutter speeds safe. They know their camera's comfortable limits from testing, use modern noise reduction, and accept some noise when the moment matters more than perfection. Low-light shooting strategies are covered further in low-light exposure strategy.
Practical examples
A dim restaurant portrait
- Situation
- At f/5.6 you need ISO 12800, and the photo is very noisy.
- What to do
- You switch to a 50mm f/1.8 lens and shoot at ISO 1600.
- Why it works
- The wider aperture collects about three stops more light, so far less noise.
- Result
- A much cleaner portrait at the same shutter speed.
City lights at night
- Situation
- At ISO 3200 the street lights clip, but lower ISO makes the photo look too dark on the camera.
- What to do
- On your ISO-invariant camera, you shoot RAW at ISO 800 with the same aperture and shutter speed and brighten shadows later.
- Why it works
- The lower ISO protects the bright lights while shadow noise stays similar.
- Result
- Detailed street lights and acceptable shadows.
Key points
- Noise comes mainly from photon shot noise (too little light) and read noise (electronics).
- Collecting more light reduces noise most; raise ISO only after aperture, shutter and light.
- ISO-invariant sensors let you brighten low-ISO RAW files with similar noise, but it varies by camera.
- Compare cameras at the same viewing size; total light collected matters more than pixel size.
- Test your camera to find your usable ISO, and use modern noise reduction wisely.
Frequently asked questions
Does high ISO cause noise?
What is ISO invariance?
Which cameras are best at high ISO?
What is the highest ISO I should use?
Conclusion
Noise comes mostly from collecting too little light, not from the ISO dial itself. Collect more light first, raise ISO when needed, compare cameras fairly and know your own usable limits. Next, you will learn about a special kind of noise that appears in long exposures: hot pixels.
Finished reading? Track your progress through Sensors and RAW.