Why phone photos look bad when you zoom in, and what can be fixed

A phone sensor is many times smaller than the sensor in a dedicated camera, and its pixels are much smaller too. Each pixel catches less light. Light arrives randomly, so the less a pixel catches, the more its reading varies from the true value. That variation is noise: random grain and colour speckles, worst in shadows and dim light.

In daylight there is so much light that noise stays small. Indoors, at dusk or at night it grows fast. The phone has to do something about it, and what it does causes most of the other problems in this guide.

To hide noise, phones apply strong noise reduction. The simplest way to remove random grain is to average neighbouring pixels, which also averages away fine detail that looks like grain: hair, grass, fabric, skin pores, distant leaves. Modern phones are cleverer than a plain average, and many merge several frames. In low light they still have to choose between grain and smearing, and they usually choose smearing.

The result is often called a watercolour or oil-painting look: textured areas turned into flat patches with soft, irregular edges. It is invisible on the phone screen and obvious at 100 %. Faces look waxy and foliage looks painted.

After smoothing, phones sharpen the image so it looks crisp on a small screen. Sharpening raises contrast along edges. When it is strong, it leaves a bright outline on one side of an edge and a dark one on the other. You see these halos around branches against sky, around text and along buildings. Edges turn crunchy while the smeared areas between them stay soft, an odd mix that reads as low quality when enlarged.

Phones merge several exposures to keep both bright skies and dark shadows visible, then squeeze that range into an image a screen can show. This is HDR and tone mapping. It is why phone photos rarely have blown-out skies, and it is mostly a good thing.

It has side effects. Brightened shadows carry more visible noise, which noise reduction then smears. Strong local adjustments create halos where bright and dark areas meet, such as a glow around a skyline. When something moves between exposures, the merge can show ghosting. And the overall look can be flat, with less contrast than the real scene.

Phones save photos in a compressed format, usually JPEG or HEIC. At default settings this is mild. The bigger loss usually comes later, when a chat app or social network resizes and recompresses the photo. Each round adds JPEG artifacts: blocks, halos, stepped gradients and smeared colour. A photo forwarded several times carries several rounds of this on top of the camera's own processing. Our guide to JPEG artifacts explains each type.

When you zoom past the range of the phone's lenses, the phone crops the centre of the sensor and enlarges it. Fewer real pixels are spread over the whole photo, so detail drops quickly. Some phones use multi-frame or AI processing on zoomed shots, which can help but can also produce the painted look described above, especially in low light. Phones with several lenses switch to a longer lens at fixed zoom steps, and those steps look much better than the zoom levels in between.

A few more causes are worth knowing:

• Motion blur: in dim light the phone uses longer exposures, so a moving subject or a shaking hand smears the image.
• Missed focus: tapping the wrong area, or a subject too close for the lens, leaves the important part soft.
• A dirty lens: a fingerprint adds haze and glow around lights. Wipe it before shooting.
• Lens softness: small lenses are often less sharp toward the corners, especially ultra-wide cameras.

Several of these problems can be repaired well, if you pick the right tool and fix same-size problems before you enlarge:

• Grain, colour blotches and smearing from low light: Enhance is built for this. It removes noise and compression damage together and adds natural fine texture where the phone smeared it away. Our page on grainy night photos covers it step by step.
• Photos that lost quality in WhatsApp or another chat app: Enhance again; see how to fix WhatsApp photo quality.
• A good photo with only light grain: Enhance as well. If the photo is already clean, it leaves it essentially as it is, so you lose nothing by trying it.
• Screenshots or photos of text: Enhance can remove compression, and text it detects is protected rather than redrawn, but none of our models is for documents. A fresh capture or the original file is the better fix.
• Low resolution, heavy crops and old phone photos: clean first, then enlarge with Upscale 4×, or Upscale 8× for very small images up to 1.25 MP.

Detail the phone's noise reduction already smeared away is gone. A repair can clean the patches and sharpen the edges around them. Tools that put texture back, including Enhance, generate a natural-looking texture; it is plausible, not the exact grass or hair that was there. Fine for a keepsake, not for evidence.

Heavy motion blur, badly missed focus, blown highlights with no detail left, strong HDR halos and extreme digital zoom are also beyond repair. Small text that is unreadable stays unreadable. For those, the fix is at capture time.

Prevention helps more than any repair:

• Add light. More light is the single biggest improvement for a small sensor.
• Hold the phone steady, or rest it on something, especially indoors and at night.
• Use the phone's built-in lens steps instead of zooming in between them, or move closer.
• Wipe the lens.
• For important shots, try a RAW or less-processed mode if your phone offers one.
• Share originals, and use the HD or document option in messaging apps.

You get 3 free credits a day, with no sign-up and no card. One photo with one tool uses 1 credit; Upscale 8× uses 2. Free uploads can be JPG, PNG or WebP up to 3 MB and 8 MP, and free results carry a watermark. Pro ($9.99 a month, 200 credits) and Pro+ ($19.99 a month, 500 credits) remove the watermark and accept uploads up to 20 MB and 30 MP. Uploads and results are deleted within 24 hours and are never used to train AI. HEIC is not accepted, so export iPhone photos as JPG first.