Downsampling and saving over the original, leaving no route back to full size for print or a future relist.
Resolution and scaling
Downsampling
Also called downscale, reduce resolution or shrink image. Here is what it means, when it changes what you export, and what people get wrong.
Resolution and scalingRuns on your deviceUpdated 10 August 2026
The short answer
Downsampling is reducing an image's pixel dimensions, for example turning a 4000 x 4000 photo into 1000 x 1000. A good downsampler averages many source pixels into each new one rather than discarding pixels. It is lossy and irreversible: what is thrown away cannot be brought back by enlarging again.
Updated
The short version
Downsampling at a glance
| Quick fact | Detail |
|---|---|
| Also called | Downscaling, resizing down |
| Halving both dimensions | Removes 75 percent of the pixels |
| Best filters | Area averaging, Lanczos, bicubic |
| Worst filter | Nearest neighbour, which aliases badly |
| Reversible | No |
| Usual side effect | Looks sharper at the smaller size |
Slide the table sideways to see every column.
How does a good downsampler decide each new pixel?
Going from 4000 px to 1000 px is a factor of four, so each output pixel covers a 4 x 4 patch, sixteen source pixels. A proper filter such as area averaging or Lanczos reads all sixteen and produces a weighted blend, so every original pixel contributes something. That is why the result still looks like the scene.
Nearest neighbour instead picks one of the sixteen and discards the other fifteen. Most of the image is simply never consulted, which is where shimmering edges and broken thin lines come from. The filter choice matters far more when shrinking than when enlarging.
Why do downsampled photos often look sharper?
Because real detail gets concentrated. A slightly soft edge spread over four pixels in the original occupies one pixel after a 4x reduction, so per-pixel contrast rises and the edge reads as crisp. Sensor noise averages out at the same time, which cleans up shadows. This is why a 24 MP file exported at 2000 px square usually looks better than a photo shot natively at 2000 px. A small amount of output sharpening after the resize, applied last, is standard practice.
What is aliasing, and why does fabric trigger it?
Aliasing happens when detail in the original repeats more finely than the new pixel grid can represent. The two patterns interfere and produce moire: wavy colour bands that were never in the room. Fine knitwear, herringbone, pinstripes, mesh and printed halftones are the usual culprits, which makes it an everyday problem for clothing sellers.
Proper filtering prevents most of it, because averaging the whole patch removes the fine frequencies before they can alias. If it still appears, resizing to a slightly different target size, or a very light blur before the reduction, usually clears it.
What does downsampling do to file size?
It is the strongest lever you have. Halving both dimensions removes 75 percent of the pixels, so a 4000 x 4000 file at 16 MP becomes a 2000 x 2000 file at 4 MP, and the JPEG typically lands somewhere around a quarter to a third of its former size. Compare that with pushing JPEG quality down, which introduces visible blocking long before it delivers the same saving. When you are trying to get under Etsy's or Walmart's 1 MB ceiling, resize first and only then adjust quality.
What people get wrong about downsampling
Each one is a real failure mode, not a style preference.
Shrinking with nearest neighbour and blaming the camera for the shimmering pattern that appears on knitwear.
Dropping JPEG quality to hit a file size limit when a resize would have got there with a cleaner image.
Sharpening before the resize rather than after, so the halo gets shrunk into a hard fringe.
Questions people ask
Downsampling, answered
The follow-up questions people search for once they have the definition.
Does downsampling lose quality?
It permanently loses detail, but the image usually looks better at its new size. Detail that no longer fits is averaged into the pixels that remain, which raises apparent sharpness and reduces noise. The loss only shows if you later need the larger version back.
Can I undo a downsample?
No. The discarded pixels are gone from the file. Enlarging afterwards estimates replacements from what survived, and a learned upscaler invents plausible texture, but neither restores the original measurements. Keep the full-size master and export copies from it.
What is the best resize algorithm for shrinking?
Lanczos or area averaging for photographs. Both read every source pixel in the region being collapsed, which suppresses aliasing. Bicubic is a reasonable middle ground. Avoid nearest neighbour entirely unless you are working with pixel art or label maps.
Should I sharpen after downsampling?
A small amount, as the very last step before export. Resizing softens edges slightly, and light output sharpening restores the bite. Sharpening before the resize is counterproductive, because the halo you created gets compressed into a hard, visible outline.
Same cluster
More from the glossary
Neighbouring entries, so every term in the set is one hop from every other.
DPI
DPI means dots per inch: the number of ink dots a printer lays down across an inch of paper.
Diffusion model
A diffusion model generates images by learning to undo noise.
EXIF
EXIF (Exchangeable Image File Format) is the block of information a camera writes inside a photo file: shutter speed, aperture, ISO, lens, date and time, orientation, colour space and often GPS coordinates.
From the studio
Tools that do this
Each one runs on your own device, so the photo stays where it is.
Sources
Where the facts on this page come from. Every link opens the specification, standard or documentation the claim was read at.
Keep reading
Elsewhere on Photomop
Platform image rules
Image size, file type and background rules for 22 selling platforms, each one sourced.
Read itConvert image
JPG, PNG, WebP and AVIF, with an honest answer about HEIC and alpha.
Read itAll the tools
The whole toolbox in three families, with what each one needs before it can run.
Read itOn-device privacy
The pipeline in five steps, what leaves your device, and how to check it yourself.
Read it
Now do it to a photo
Photomop is a photo studio that runs on your own device. Resize, crop, compress, convert, batch edit, remove a background and change one all work in the browser tab you are reading this in, at full resolution and with no watermark. The on-device page shows you how to check that the photo stays put.