Computational Photography: What the Phone Decided for You
Every photograph from a phone is an argument about what the scene should look like. We shot the same frames on four of them and measured how far each moved from the raw capture.
Tech Daily Review Desk
The shift
Computational photography made the sensor the uninteresting variable. What separates phone cameras now is the processing pipeline, and it is applied before you ever see the image.
A phone photograph is the output of a pipeline: several exposures aligned and merged, noise reduced, tone mapped, sharpened, and skin treated separately from everything else. None of it is optional and most of it is not disclosed. Comparing phone cameras is therefore comparing editorial decisions made by engineers, which is a different activity from comparing lenses.
Why it matters
Because you cannot undo it. A pipeline that lifts shadows and warms skin produces a photograph many people prefer and a record of the evening that is not quite what happened.
How the testing worked
Four phones tripod-mounted side by side, firing on the same remote trigger at the same instant, saving raw and JPEG together. Deltas were measured between each phone's own JPEG and its own raw file on the same sampled regions — a grey card, a shadow patch and a skin patch — so every figure is that phone against itself rather than against a reference we chose.
What we recorded
| Phone | Shadow lift | Saturation delta | Sharpening | Skin warmth shift |
|---|---|---|---|---|
| Phone A | +1.8 EV | +14% | Heavy | +180K |
| Phone B | +1.1 EV | +6% | Moderate | +40K |
| Phone C | +2.4 EV | +21% | Heavy | +310K |
| Phone D | +0.7 EV | +3% | Light | −20K |
Reading the results
The phone people most often call "the best camera" made the largest changes to the scene, not the smallest.
Shadow lift is where the pipelines diverge most, and it is the edit that most changes the mood of an image.
The most restrained pipeline produced files that looked flat out of the phone and held up best to editing afterwards.
What to be careful of
Three things worth holding on to when you meet these figures somewhere else:
- "more natural" in a review usually means closer to the reviewer's taste, not closer to the scene
- Shooting raw sidesteps most of this and loses the multi-frame stacking that makes phone cameras good
- Pipelines change with firmware; a camera comparison has a shelf life measured in months
What we would buy
Look at other people's photographs from a phone before you buy it, ideally of faces in mixed light. The pipeline is the camera, and the pipeline is a matter of taste in a way that a sensor size is not.
If you edit, shoot raw and JPEG together rather than raw alone. The JPEG gives you the stacked, denoised result the phone is good at; the raw gives you room to disagree with its decisions about shadows and skin. Keeping both costs storage and nothing else, and it is the only way we have found to get the benefit of the processing without being stuck with it.
Our verdict
Pick the pipeline you like rather than the sensor you were sold, and look at raw files if you intend to edit. The processing is the camera now.
Behind this piece
Every product covered here was bought at retail and lived with for at least a fortnight before a word was written. Battery figures come from a scripted day rather than a manufacturer estimate, and sustained performance is measured warm — after half an hour of load — because that is the state a device spends most of its life in.
Where a firmware update changes a result, the piece is re-tested and the change noted rather than quietly corrected. No review unit arrives with conditions attached, and no score is agreed in advance with anybody.
Where to go next
If you are weighing this against the rest of the cameras we have covered, the category archive lists every review in date order. The closest comparisons we have published are DJI Pocket 3 Review: Close to Faultless and DJI Pocket 3 Review: The Vlogging Camera to Beat.
For the reference material behind these figures, the specification pages carry the measured numbers, and the common questions collect what people ask once something is in daily use.
