How to Make Pixel Art From a Photo: A Step-by-Step Guide
Turn any photo into clean pixel art with free tools: pick the right grid size, cut the palette, fix stray pixels, and make a color-by-number chart.
by jmApp · Updated 2026-11-16
To make pixel art from a photo, crop the photo tightly, shrink it to a small grid such as 32×32 or 64×64 pixels, and cut it down to a palette of 8 to 32 colors. Then fix stray pixels by hand and enlarge it with nearest-neighbor scaling so every pixel stays a sharp square. Free tools like GIMP, Photopea and Piskel can do every step.
The automatic part takes about five minutes. The hand cleanup afterward is what makes it look like real pixel art instead of a blurry, low-resolution photo. Below is the full process, with the exact settings to use.
What size should pixel art from a photo be?
Size decides how much of the photo survives. A typical phone photo is around 3000×4000 pixels. If you shrink it to 48×48, each new pixel stands in for a block of roughly 60×60 original pixels. Faces, text and thin lines disappear fast, so pick the grid based on how much detail you need to keep.
| Grid size | Good for | Colors to aim for | What you lose |
|---|---|---|---|
| 16×16 | Icons, simple objects, a single fruit or logo shape | 4–8 | Almost all detail; only the silhouette remains |
| 32×32 | Pets, characters, game-style sprites | 8–16 | Facial expressions, small textures |
| 64×64 | Portraits that still look like the person, small scenes | 16–32 | Fine hair, background clutter |
| 128×128 | Landscapes, detailed portraits, wall art | 32–64 | Very little, but it starts to look less like pixel art |
If you are unsure, start at 48×48 or 64×64. It is easier to judge a result you can recognize, and you can always try smaller later.
Which free tools can turn a photo into pixel art?
You do not need to buy anything. These cover every step:
- GIMP (Windows, Mac, Linux): free desktop editor. Handles cropping, resizing, color reduction and pixel-by-pixel editing.
- Photopea: runs in a web browser and uses menus very close to Photoshop. Good if you are on a Chromebook or a work computer where you cannot install software.
- Piskel: a free browser-based pixel editor. It is weak at converting photos, but great for the cleanup stage because it is built for drawing pixel by pixel.
- Online pixelators: many sites let you upload a photo and pick a block size and color count. They are fast for a preview, but you usually get less control over the palette, so plan on cleaning the result in GIMP or Piskel.
- Lospec palette list: a free library of ready-made pixel art palettes. Useful when you want a retro look instead of the colors the software picks for you.
If you already pay for Photoshop, the same steps work there with nearly identical menu names.
How do you convert a photo to pixel art step by step?
Here is the process in GIMP. In Photopea or Photoshop the menus are the same or very close.
- Pick the right photo. Choose one with a clear subject, a plain background and strong light. A dog on a grass lawn works. A dog in a busy living room does not, because the clutter turns into noise at small sizes.
- Crop tight. Use the crop tool to cut everything that is not the subject. Crop to the shape of your final grid, usually a square.
- Boost contrast and saturation. Go to Colors > Brightness-Contrast and push contrast up a bit, then raise saturation slightly in Colors > Hue-Saturation. Shrinking averages colors together, so a photo that looks normal now will look gray and flat at 32 pixels wide.
- Shrink to your grid size. Open Image > Scale Image, set the width to your target (say 48 px) and keep the chain icon locked. For shrinking, a smooth interpolation such as Cubic gives cleaner averaged colors than None.
- Reduce the colors. Go to Image > Mode > Indexed, choose “Generate optimum palette” and set the maximum number of colors, for example 16. Set dithering to None for a clean, flat look.
- Zoom in and clean up. Zoom to 800% or more, pick the Pencil tool at 1 px, and fix the problems covered in the next section.
- Enlarge for sharing. Switch back to Image > Mode > RGB, then Image > Scale Image again. Multiply the size by a whole number (48 px × 10 = 480 px) and set interpolation to None. This is the setting people miss most often. Any other option blurs your crisp squares.
- Export as PNG. File > Export As, and save as .png. JPG compression adds faint smudges around every pixel edge.
In Photopea, the resize step is Image > Image Size, and the crisp enlarge setting is called “Nearest Neighbor.” Color reduction is under Image > Mode > Indexed Color.
How many colors should you use?
Fewer than you think. Real pixel art usually looks better with a small, deliberate palette than with every shade the photo contains. A 32×32 pet portrait often reads best with 8 to 12 colors: two or three shades for the fur, one or two for the eyes and nose, and a few for the background.
A few rules that hold up well:
- Give each material its own ramp. Skin, hair and clothing each get a dark, mid and light shade. Three steps is usually enough.
- Make the background simple. One or two flat colors keep the subject readable. A solid color behind a portrait almost always beats a pixelated version of the real background.
- Try a fixed palette for a style. In GIMP’s Indexed dialog you can choose “Use custom palette” instead of letting it pick. A retro console palette from Lospec instantly makes a photo look like a game screen.
- Avoid near-duplicates. If two colors look almost identical when zoomed out, merge them. Every extra color makes the image busier and harder to edit.
How do you fix pixel art that looks muddy or noisy?
The converter does the heavy lifting, but it cannot tell an eye from a shadow. These are the fixes that take a result from “blurry photo” to “pixel art”:
- Delete orphan pixels. A single pixel of a color surrounded by a different color reads as dirt. Repaint it to match its neighbors unless it is a deliberate highlight, like the shine in an eye.
- Redraw the eyes. At 32×32 an eye might be one or two pixels. Place them by hand. A one-pixel shift changes the whole expression.
- Add an outline. A one-pixel dark outline around the subject separates it from the background. Use a dark version of the neighboring color rather than pure black for a softer look.
- Clean up jagged lines. Curves look smooth when the steps follow a steady pattern, like 3 pixels, 2 pixels, 1 pixel. Uneven runs (3, 1, 3) look broken.
- Check it at actual size. Zoom out to 100% every few minutes. Something that looks wrong at 1600% often looks fine at real size, and the other way around.
For a 48×48 portrait, expect 20 to 40 minutes of cleanup the first time. It gets much faster once you know what to look for.
How do you turn your pixel art into a color-by-number?
A small, limited palette is exactly what a color-by-number grid needs. Each pixel becomes a cell, each color gets a number, and the picture appears as you fill it in. This makes a great printable activity for kids or a personal gift.
- Keep it small. 24×24 to 40×40 works on a printed letter-size or A4 page. Bigger grids make the cells too tiny to number by hand.
- Limit the palette to 6–12 colors. Number them 1, 2, 3… and write the key at the bottom of the page with a swatch of each color.
- Show the grid. In GIMP, View > Show Grid with the spacing set to your enlargement factor (10 px if you scaled up 10×) lines up with every pixel. Take a screenshot or use Filters > Render > Pattern > Grid to burn the lines into a copy.
- Write the numbers. On a printed blank grid, write each cell’s number by checking your finished image. Leave the most common color, often the background, as blank cells to save time.
- Test it yourself. Color a corner before giving it to someone. Two colors that looked distinct on screen can be hard to tell apart in markers or pencils.
If you would rather do the coloring part than build the chart, try a few pictures in Fine Artist (free on iPhone and iPad and Android). You pick a color, tap the matching numbers or swipe across a whole area, and the pixel art appears cell by cell. Its designs come from the app’s own growing library of animals, food and characters rather than your photos, but coloring a few is a quick way to see how artists group colors and shade with only a handful of tones. That makes your own conversions easier to clean up.
What are the most common mistakes?
- Enlarging with smooth interpolation. Your squares turn into soft blobs. Always use None or Nearest Neighbor when scaling up.
- Scaling by odd amounts. Enlarging by 250% makes some pixels two wide and others three. Stick to whole multiples like 4×, 8× or 10×.
- Leaving dithering on. Dithering scatters dots of mixed colors to fake extra shades. It can work for skies, but on faces it looks like static.
- Starting with a busy photo. No setting rescues a cluttered background. Crop harder or pick another shot.
- Saving as JPG. Use PNG so the edges stay clean.
Once you have done one photo start to finish, the next one goes much faster. Keep a copy of the palette you liked, reuse your grid size, and save the cleanup for the parts that matter most: the eyes, the outline and the shape of the subject.
Frequently asked questions
Can I make pixel art from a photo on my phone?
Yes. Browser tools like Photopea and Piskel work on phones and tablets, though a mouse or stylus makes the pixel-by-pixel cleanup much easier. Many mobile photo editors also have a pixelate effect for a quick preview.
Why does my pixel art look blurry after I enlarge it?
The editor used smooth interpolation when scaling up. Undo, then enlarge again by a whole number like 8× or 10× with interpolation set to None or Nearest Neighbor.
What is the best size for pixel art made from a portrait?
64×64 is a good starting point for a recognizable face. 32×32 works for a stylized look, but you will need to redraw the eyes and mouth by hand.
Should I use dithering when reducing colors?
Usually no. Dithering adds scattered dots that look noisy on faces and objects. Turn it off and add shading by hand where you need it.
Is converting a photo to pixel art cheating?
Not really. Many pixel artists use a photo as a starting point and then redraw large parts by hand. The cleanup, palette choices and outlines are where the real work happens.