A converter cannot know which details identify the subject. It treats a catchlight in an eye, a speck of background noise, and a logo corner as color samples. Your job is to decide which of those details deserve pixels.
Too much local detail
Noise competes with the silhouette and increases color changes during placement.
Deliberate shapes
A clear outline and controlled highlights remain readable at actual size.
Choose pixel dimensions before converting colors
Resolution determines how much information survives. If you reduce colors first and resize later, the resize step can introduce blended colors that are outside the palette. Set the final width and height first, then quantize the result.
| Approximate width | Best for | Main risk |
|---|---|---|
| 24–40 px | Symbols, small icons, simple faces | Tiny text and interior detail disappear |
| 48–80 px | Avatars, mascots, compact logos | Curves may need manual cleanup |
| 96–160 px | Detailed characters and illustrations | Placement effort grows quickly |
| 160+ px | Team projects with planned coordination | Versioning and maintenance become essential |
These ranges are starting points, not platform limits. A simple emblem can look excellent at 24 pixels while a portrait may need several times that width. Test two or three sizes and choose the smallest one that preserves the subject.
A reliable image-to-pixel-art workflow
1. Start with a clean source
Use the highest-quality source available, but crop it tightly. Remove watermarks you do not have permission to remove or reproduce; use artwork you created or are allowed to use. Increase separation between the subject and background before downscaling.
2. Protect the silhouette
The outer shape is usually more important than interior texture. If the subject blends into the background, add a deliberate outline or remove the background. For transparent artwork, clean the edge so it does not contain semi-transparent halos.
3. Resize at the intended aspect ratio
Do not stretch an image to fill a target box. Lock the aspect ratio, resize, and add transparent padding only when the placement plan requires a specific rectangle. Existing pixel art should use nearest-neighbor scaling.
4. Convert with the smallest useful palette
Open the Wplace converter, disable colors that are unavailable or undesirable, and run the conversion. A smaller palette can make the image easier to coordinate and repair, even if a larger palette gives a closer photographic match.
5. Inspect at two zoom levels
At high zoom, inspect broken outlines, isolated pixels, and transparency. At 100% zoom, check whether the subject remains recognizable. A technically clean image can still fail if its expression or symbol is lost at normal size.
6. Make a deliberate cleanup pass
Remove single-pixel noise, reconnect broken lines, simplify uneven diagonals, and standardize repeated shapes. Manual cleanup is what turns a generated conversion into a purposeful piece of pixel art.
When to use dithering
Dithering alternates palette colors to simulate an intermediate tone. It is valuable for broad gradients, photographic shading, fog, and soft lighting. It is usually less helpful for lettering, QR-like patterns, thin outlines, flags, or small logos.
Try dithering when
- • a large gradient becomes a distracting hard band
- • a photograph loses depth after palette reduction
- • the final image will be viewed from a distance
Avoid dithering when
- • text and symbols need solid, repeatable edges
- • the image is so small that patterns read as noise
- • placement speed matters more than tonal detail
Final review before export
- Every visible color exists in the palette you intend to use.
- The width and height match the placement plan.
- Transparent pixels are truly transparent.
- No important line is only one accidental pixel away from breaking.
- The image is recognizable at 100% zoom.
- The exported PNG has a clear revision name.
After export, continue with the Wplace template guide to record coordinates and prepare the overlay. The palette reference is useful when you need to verify an individual RGB or HEX value.
Try it yourself
Compare settings with the same source image
Generate a clean version and a dithered version, then judge both at actual pixel size.
Open the image converter