Grain, noise and blur are three different things
These get used interchangeably, and they have different causes and cures.
Noise is random variation in the recorded signal, visible as speckle. It comes in two forms: luminance noise, which looks like fine sand and is relatively benign, and chroma noise, which looks like coloured blotches in the shadows and is the ugly one. This is what people mean when they say a photo is grainy, and it is what this page is about.
Blur is a loss of definition from movement or missed focus. A blurry photo is smooth and soft; a noisy photo is sharp and speckled. They can appear together in a dark room, where a slow shutter produces both, but they are separate faults with separate answers. Blur is covered in how to fix blurry photos, and the honest summary there is that most of it cannot be repaired.
Film grain is a texture people deliberately add, because it reads as photographic, softens digital harshness and gives a flat image something to hold on to. That is a look, and it is chosen. If that is what you came for, see how to add grain to photos, which is the opposite of this page.
The distinction matters because the fix for unwanted noise is more light, the fix for blur is a faster shutter, and the fix for wanting grain is a slider.
The five causes of unwanted grain
| Cause | How to recognise it | Fix |
|---|---|---|
| Not enough light | Speckle worst in shadows and midtones, the whole frame affected | Add light, open a curtain, move nearer a window, use Night mode |
| Digital zoom | Noise plus a smeared, painted look, worse than the same scene at 1x | Walk closer and shoot at 1x, then crop if you must |
| Lifting shadows in editing | Clean photo becomes blotchy only where you brightened it | Lift less, or fix the exposure at capture instead |
| A very small sensor | Noise even in decent light, on older or budget phones and front cameras | Use the main rear camera, which has the largest sensor |
| Compression and re-sharing | Blocky rather than speckled, worst in smooth areas such as sky | Post the original file once, not a screenshot or a saved copy |
The front camera deserves a specific mention, because a large share of grainy photo complaints are selfies taken indoors at night. Front cameras have smaller sensors and no optical stabilisation on most phones, so they hit the noise threshold far sooner than the rear camera does. Turning a lamp on or standing nearer a light does more than any setting.
Two of those five are self-inflicted rather than unavoidable, which is the good news.
Why small sensors struggle in low light
A camera sensor counts photons. In bright light it counts a great many, and the random variation between neighbouring pixels is tiny relative to the signal. In dim light it counts very few, and the random variation becomes a significant fraction of what was measured. That variation is the noise, and it is a property of light itself rather than a defect in the camera.
ISO is the amplification applied afterwards. Raising ISO does not make the sensor more sensitive; it multiplies whatever was recorded, signal and noise together. That is why a high ISO photo is noisy: the noise was always there, and amplification made it visible.
Phone sensors are physically small, so each pixel collects fewer photons than a pixel on a larger camera sensor would. This is the whole reason a dedicated camera outperforms a phone at night, and no software removes the difference, although computational photography narrows it considerably by merging many frames.
That merging is what Night mode does. It takes a sequence of exposures and combines them, and because the noise is random while the scene is not, averaging several frames cancels a great deal of the noise. It also explains Night mode's weakness: anything that moves during the sequence turns into a smear, so it works on a static room and fails on a person mid-laugh.
What to do with a grainy photo you already have
Noise reduction always trades detail for smoothness. The skill is in stopping before skin turns to plastic.
Reduce chroma noise first. Coloured blotches are far more objectionable than fine luminance speckle, and removing them costs almost no detail. Lightroom's Color noise slider and Snapseed's Details tools both handle this. Push colour noise reduction reasonably hard and luminance noise reduction gently.
Keep some luminance noise. A completely denoised photo looks waxy, particularly on faces. A little fine speckle reads as photographic and is far less distracting than smeared skin.
Reduce the display size. Noise is scale-dependent. A photo that looks unacceptable at full screen frequently looks fine at feed size, and cropping in or exporting smaller genuinely helps rather than merely hiding the problem.
Consider black and white. Monochrome removes chroma noise entirely, and luminance grain in a black and white photo reads as a deliberate texture rather than a fault. This rescues more dark, noisy night photos than any denoising tool.
Be sceptical of AI denoise. Generative denoisers invent plausible detail: fine for a social post, misleading for anything that must be accurate. The same caution applies as in how to enhance photo quality.
Preventing it at capture
Every fix above is damage limitation. These prevent the problem.
- Add light before you change any setting. Open a curtain, turn on a lamp, move to the lit side of the room, or ask someone to hold a phone torch just out of frame. One extra stop of light halves the noise.
- Use the main rear camera at 1x. It has the biggest sensor and the fastest lens on the phone. The ultra wide and the front camera are both noticeably noisier in dim light.
- Do not use digital zoom in low light. Zooming beyond the optical range crops into the sensor, which means fewer photons per output pixel and visibly more noise. Walk closer instead.
- Hold still and let Night mode run. Brace against a wall or a table. The longer capture only works if the phone and the subject stay put.
- Expose for the subject, not the scene. Tap the face, and if the photo is dark, drag the exposure up at capture. Brightening at capture is cleaner than brightening in editing.
- Use flash rather than accept a very dark frame. Direct flash is unflattering at close range and unpopular, but a well-lit flash photo has almost no noise, and it is often the better of two bad options. The technique is in flash photography tips.
Grainy video, screenshots and old photos
Three neighbouring cases with slightly different answers.
Video. Noise in video is more visible than in stills because it shimmers between frames. The causes are identical, but the fix list is shorter: video cannot merge multiple exposures the way Night mode does, so light matters even more. Reducing the frame rate helps a little in very dim conditions because each frame gets a longer exposure.
Screenshots and re-shared images. What looks like grain here is usually compression artefacting rather than sensor noise. It appears as blocking in flat areas and fuzzy halos around edges, and no noise reduction tool fixes it because the information has been discarded rather than buried. The answer is to obtain the original file. This is the same mechanism described in why do my photos look bad on Instagram.
Old digital photos. Cameras and phones from ten or fifteen years ago had small, noisy sensors and crude noise reduction, so their low light photos are grainy and there is nothing to recover. Treat the grain as part of the photograph's age rather than as a defect, and convert to black and white if it bothers you.
Scanned film. Real film grain is not noise and should not be denoised. Smoothing it removes the texture that makes the scan look like film.
Questions people ask
Why are my photos grainy?
Because not enough light is reaching the sensor. The camera amplifies the weak signal to reach a usable brightness, and amplification multiplies the random variation in the reading along with the picture, which you see as speckle.
Is grain the same as blur?
No. Noise is speckle from too little light and a sharp but gritty image; blur is smooth softness from movement or missed focus. They often appear together in a dark room, but they have different causes and different fixes.
Why are my night photos grainy even with Night mode?
Night mode merges several exposures to average out the noise, which only works if nothing moves. If you or your subject shifted during the capture, the phone falls back on a shorter, noisier exposure, or produces a smeared result.
How do I remove grain from a photo?
Reduce colour noise firmly and luminance noise gently, since colour blotches are the ugly part and removing luminance noise entirely makes skin look waxy. Exporting smaller and converting to black and white both help more than people expect.
Why are my selfies grainier than my other photos?
The front camera has a smaller sensor than the main rear camera and usually lacks optical stabilisation, so it reaches the point of needing amplification much sooner indoors. Turn on a lamp or move nearer a window before changing any setting.