Use case Photography

Enhance and binning: enlarging a noisy photo

When to set binning on Enhance: what each square takes out, of grain and of fine detail, on a reception at ISO 12800 and on a clean suit, and when to leave it off.

A photo taken in low light, at a high ISO, is covered in grain. If you also want it larger, for a print or a crop, Enhance can do both in one run: set binning, its simple noise reduction. This guide shows what binning takes out of a photo, grain and fine detail both, square by square, on a wedding reception shot at ISO 12800. And it shows why to leave binning off on a clean photo.

Every result here is what Enhance returned, with the settings on its label or in its caption. Drag a divider to compare, and zoom to inspect.

A bride's face at a wedding reception, her eyes, brows and hair covered in heavy coloured grain from a high ISO — A bride at her wedding reception, ISO 12800: a part of the frame enlarged by a third, with Enhance and binning at 3x3. WIPE, ZOOM & PAN to see more detail. Photograph by John Trainor, CC BY 2.0, via Wikimedia Commons; enhanced.

The same part enlarged by a third with Enhance and binning at 3x3, the grain gone from her skin and eyes, her lashes, irises and the strands of her hair sharp — A bride at her wedding reception, ISO 12800: a part of the frame enlarged by a third, with Enhance and binning at 3x3. WIPE, ZOOM & PAN to see more detail. Photograph by John Trainor, CC BY 2.0, via Wikimedia Commons; enhanced.

What Enhance does

Enhance is for enlarging. It makes a photo larger and sharper, by any scale from 1 to 4, and rebuilds fine detail at the new size: lashes, strands of hair, the weave of cloth. It keeps your colours and tones as they are. Event photography shows it on crops and prints.

Enhance reads everything in a photo as detail, grain included. Without binning, it keeps the grain and enlarges it with everything else.

What binning does

Binning is Enhance’s simple noise reduction. Pixel binning averages each square of 2x2, 3x3 or 4x4 pixels into one before Enhance enlarges. Averaging takes the grain out. It also leaves less room to enlarge: the largest scale is 4 divided by the square’s side. In return, you can send a larger photo.

Binning	Largest scale	Largest photo, paid compute	Largest photo, community queue
off	4	12 MP	3 MP
2x2	2	48 MP	12 MP
3x3	1.33	108 MP	27 MP
4x4	1	192 MP	48 MP

Choose the binning first, and the scale follows it: it goes up to the largest scale in the table, and at 4x4 the photo comes back at its own size.

Grain against detail

A larger square takes out more grain. It also takes out more of the photo’s own fine detail, and Enhance draws more of what you see back in. So the choice is how much detail a photo can give up for how clean it needs to be.

Below, the same part of a reception photo at ISO 12800, five ways: the photograph, and Enhance with binning off, 2x2, 3x3 and 4x4. All are at the photograph’s own size, scale 1, each pixel drawn twice. Photograph by John Trainor, CC BY 2.0, via Wikimedia Commons.

Five views of a bride's eye at ISO 12800, three across: the photograph, covered in coloured grain; Enhance with binning off, the grain kept as a mottle; binning 2x2, clean, with separate, sharp lashes and a thin streak drawn down through the iris; 3x3, as sharp, the streak fainter; and 4x4, the softest, its lashes and iris smoothed — ZOOM & PAN to see more detail.

Without binning, the grain stays on her skin and in the white of her eye, as a mottle. At 2x2 the grain is gone, and her lashes come out separate and sharp; a thin streak is drawn down through her iris. At 3x3 they are as sharp, and the streak is fainter. At 4x4 the streak is gone, and her lashes and iris are softer.

Five views of a groom's eye and brow at ISO 12800, three across: the photograph, grainy; Enhance with binning off, still grainy and blotched; binning 2x2, the grain gone and streaky painted strokes across his brow; 3x3, clean, with deep creases drawn round his eye; and 4x4, the softest, with the fewest lines — ZOOM & PAN to see more detail.

The groom, at the edge of the frame, shows the cost most. Without binning his skin stays grainy and blotched. At 2x2 Enhance paints streaky strokes across his brow. At 3x3 they are gone, and deep creases are drawn round his eye. 4x4 is the softest, with the fewest lines.

Five views of the embroidered lace of a wedding dress at ISO 12800, three across: the photograph, grainy, its scrolls and the glints of its beads faint; Enhance with binning off, still grainy; binning 2x2, many glints kept, with a faint mottle between them; 3x3, clean, with fewer glints; and 4x4, smooth raised scrolls lined with a neat row of evenly spaced beads — ZOOM & PAN to see more detail.

Without binning the lace stays grainy. 2x2 keeps the most glints of the beads, with a faint mottle left between them. At 3x3 the mottle is gone, and so are some glints. At 4x4 the embroidery comes back as smooth, raised scrolls lined with evenly spaced beads: neater than the photograph shows.

Five views of a pearl earring and a woman's hair against a plain yellow wall at ISO 12800, three across: the photograph, the wall covered in grain; Enhance with binning off, the wall still mottled; and binning 2x2, 3x3 and 4x4, the wall clean, with fine flyaway hairs drawn across it, more of them at 3x3 and 4x4 — ZOOM & PAN to see more detail.

Behind her earring, the plain wall stays mottled without binning, and comes out clean at every square. From 2x2 on, Enhance draws fine flyaway hairs across the wall, and more of them at 3x3 and 4x4.

In short: 2x2 keeps the most of the photo’s own detail, and can leave a little grain and some painted strokes. 3x3 and 4x4 come out cleaner, and more of what you see is drawn. 4x4 is the softest.

Enlarging twice

At 2x2, you can enlarge up to twice. Here is her eye at scale 2, with binning off and at 2x2, beside the photograph’s pixels drawn twice:

Three views of the bride's eye at ISO 12800, the same part at twice the size: the photograph, its pixels repeated, grainy; Enhance at 2× with binning off, the grain drawn as a fine, scratchy mottle over her skin and the white of her eye; and Enhance at 2× with binning 2x2, clean, the lashes separate and the iris clear, with a thin streak drawn down through it — ZOOM & PAN to see more detail.

Without binning, the grain is enlarged with the rest, a scratchy mottle over her skin and the white of her eye. With 2x2, the eye is clean at twice the size.

On a clean photo, leave it off

A clean photo has no grain to trade. Binning then only takes away fine detail, and Enhance draws something else in its place. This is a suit’s sleeve beside a wrist, from a photo taken at ISO 200:

Five views of a pale blue suit's sleeve beside a wrist, from a clean photograph at ISO 200, three across: the photograph, the cloth's weave fine and faint; Enhance with binning off, much the same; binning 2x2, the weave turned into clear diagonal stripes; and 3x3 and 4x4, bold, even stripes, with small red dots drawn on the wrist — ZOOM & PAN to see more detail.

Without binning, the cloth keeps its fine, faint weave. At 2x2 the weave turns into clear diagonal stripes, and at 3x3 and 4x4 into bold, even ones. Zoom into the wrist: 3x3 and 4x4 also draw small red dots on the skin. That is detail drawn, not kept. Photograph by Shixart1985, CC BY 2.0, via Wikimedia Commons.

Leave binning off on a photo without visible grain. To get a large clean photo through, frame the part you need instead.

Which to choose

A clean photo: binning off.
A noisy photo you are enlarging: 2x2 for the most size, up to twice, and the most of the photo’s own detail. 3x3 or 4x4 for a cleaner result, at up to 1.33 times or the photo’s own size; 4x4 is the softest.
A noisy photo you only want cleaner, at its own size: use Denoise instead. It does that much better than binning. Event photography shows it on this same reception.

Before you deliver

Compare every face at full size with the original. Lines in faces can come out deeper than the photo shows, and 2x2 can paint strokes on skin.
Check the fine things a client knows: lace, jewellery, hair. Where grain hid them, what you see is drawn.
At scale 1, every result here lines up with the photograph pixel for pixel. You can layer it over the original in an editor and blend the two.

Each result here is one run, and each tool’s price is on its card in the app.

How these were made

The photographs are from Wikimedia Commons, used as published there and credited under each. The reception is “BethsWeddingReception-3497” by John Trainor, CC BY 2.0, from a Canon EOS 60D at ISO 12800; a 1244 × 2211 part of the frame was sent. The suit is “A couple is exchanging wedding rings” by Shixart1985, CC BY 2.0, from a Nikon camera at ISO 200; a 1024 × 768 part was sent.
Each result was made on 24 October 2026 and returned as a PNG, with the settings on its label.
The figures show the same pixels of the photograph and of each result, each pixel drawn twice. The figure at scale 2 shows the photograph’s pixels drawn twice and each result at its own pixels.

Try it on your own file

No sign-up and no card — a workspace opens in your browser the moment you arrive.

Open Web App