Turning a photo into an emote that does not turn to mush

Last updated: 2026-08-18

Drop a selfie, a pet photo or a scan of your own drawing straight into an emote and nobody can tell what it is at 24px. Here is how to cut the detail down with a monochrome conversion so it still reads when small.

What happens when you drop a photo straight in

A selfie, a pet, a phone snap of a drawing — all of them carry a lot of detail. What lands in chat is a 24–28px square, and neither the tonal range nor the fine detail of a photo fits into that.

A shrunken photo does not read as a small photo. It reads as a grey blob: the expression and the silhouette both collapse, and nobody can tell what the emote is.

The fix: strip the detail down to a single colour

To survive at small sizes you have to cut the colour count and the tonal range hard. Split the image by brightness and paint only the dark side in one colour — the same idea as a rubber stamp or a screen print. The fine detail goes, but the silhouette and the broad light/dark shapes stay, which is what makes an emote recognisable when small.

In PochiSta, switch on "Monochrome stamp" in the image tab. The result is not baked into the image — it is stored as settings, so moving the sliders back returns you to the original photo.

1. Remove the background first

Cut the background out before converting. If it is still there, it gets painted in the same colour as the subject and the boundary between them disappears.

The "Background removal" button picks its method automatically, so a photo shot against a white wall or a plain backdrop is done instantly.

2. Use the threshold to decide how much gets painted

The threshold is the brightness at which a pixel switches from painted to not painted.

  • Lower: only genuinely dark areas survive. Lines get thinner, and a bright subject can disappear entirely
  • Higher: more of the image is painted. A backlit or dark photo turns into one solid block

Aim for the least amount of paint that still shows the shape. Watch the actual-size preview and lower it until the silhouette is only just readable.

3. Start with two tone levels

Tone levels control how many steps of shading you get. Two levels (a plain black/white split) looks the most stamp-like and holds up best when small.

  • 2 levels: painted or not. Rubber-stamp style, and it survives being combined with text
  • 3–4 levels: mid tones come back and some depth is preserved — useful when you want an expression to read
More levels means the result stays closer to the photo, but the tendency to turn to mush comes back with it. If you cannot tell the levels apart in the actual-size preview, use fewer.

4. For line art and stamps drawn on paper

A drawing or a stamp photographed or scanned from white paper has a background that is white, not transparent. Turn on "Make light areas transparent" — the paper drops out and only the lines remain in your chosen colour.

If shadows made the paper look grey in the shot, those shadows get painted too. Lower the threshold, or reshoot somewhere brighter.

5. Choose the colour for where it will be used

The default vermilion reads like a seal and stays visible on both light and dark backgrounds. Black is strong on light themes but sinks into a dark chat.

When in doubt, add the sticker-style white outline after converting. It keeps the silhouette clear on either theme.

Checklist

  • Did you remove the background first?
  • Is the silhouette readable at actual size (24–28px)?
  • Has it collapsed into one solid block?
  • For line art, is the paper white gone?
  • Did you check on both dark and light backgrounds?

Related articles

You can put this article into practice with PochiSta

Export straight to each platform's recommended size and file-size limit, optimized automatically. No sign-up, completely free, and your images never leave the browser.

Make an emote