JPG to Pixel Art Converter

What JPG compression does to your image

Before you convert a JPG to pixel art, it helps to know what the format already did to your image. JPG is lossy: to shrink file size, it slices the picture into 8×8-pixel blocks and throws away the fine detail inside each one — the detail your eye is least likely to miss at full resolution. It also stores color at lower resolution than brightness, a shortcut called chroma subsampling. The result is faint block seams, soft ‘ringing’ halos around hard edges, and color that smears a pixel or two past where it should stop.

At normal viewing size, none of this matters. But converting a JPG to pixel art runs color quantization — the image is reduced to a small palette, and the converter picks the best few colors to represent everything. Quantization cannot tell an artifact from real detail. A ringing halo earns its own palette color; a smeared edge becomes a fringe of wrong-tinted pixels; the block grid can survive as visible texture in flat areas like skies and walls. That is why a JPG that looks fine full-size can come out noisy as pixel art.

Settings that hide compression artifacts

You cannot remove JPG artifacts, but you can make them irrelevant when you convert JPG to pixel art. Every artifact is small — a pixel or two of ringing, a subtle tint inside a block — so any setting that averages more source pixels together, or gives stray tones fewer palette slots to occupy, hides them.

• Go chunkier. A smaller output size means each pixel block averages more of the source, and the compression grid disappears into the average. If 128 px shows noise, 96 or 64 px usually will not.
• Cut the color count. At 16–24 colors, smeared in-between tones merge into their neighbors instead of earning their own slot. This is the single most effective fix for a rough JPG.
• Rethink dithering. Floyd–Steinberg keeps gradients smooth but treats noise as detail worth preserving, which can amplify blockiness near edges. Atkinson dithers more gently, and turning dithering off entirely suits bold, flat looks.
• Lock a retro palette. A fixed palette snaps every pixel to a known color, so compression tints simply cannot appear in the output.

JPG photos vs JPG graphics

Not every JPG suffers equally. Photos survive compression well: their soft gradients and natural texture hide artifacts, and JPG was designed for exactly this content. When you convert a JPG to pixel art and the source is a decent phone photo, the compression penalty is barely visible.

Sharp-edged graphics are the opposite case. Logos, text, and flat illustrations saved as JPG collect ringing halos along every hard edge — precisely where quantization will pick them up and turn them into fringes of off-color pixels. If a PNG version of the graphic exists anywhere, use it and take the PNG to pixel art route instead; lossless edges convert dramatically cleaner. Fall back to the JPG only when it is genuinely the only copy you have, and lean on a low color count when you do.

When to find a better source

Some JPGs are not worth fighting. Every save re-runs the compression, so an image that has been downloaded, re-shared, and re-saved a few times carries stacked generations of artifacts — meme rot. Screenshots of JPGs are similar: the screenshot captures the artifacts of the original, then often gets compressed again on upload.

Warning signs that a better source will save you time: blocky patches visible in flat areas at full size, mosquito-like specks around edges, and colors that look washed out compared to how you remember the image. A reverse image search often turns up a larger, earlier copy. Two minutes of hunting beats twenty minutes of tweaking sliders on a hopeless file — no jpg to pixel art conversion can restore detail the compression already destroyed.

Step-by-step: clean pixel art from a phone photo

Camera JPGs from a recent phone are high quality, which makes them the easy case for jpg to pixel art. This workflow produces the cleanest result:

• Crop tight. Fill the frame with the subject — a face, a pet, an object. Every pixel block spent on background is a block not spent on the thing you care about.
• Boost contrast and saturation slightly. Phone photos skew flat; a small push separates shapes so the palette lands on your subject.
• Convert at 96–128 px with 16–24 colors. Keep Floyd–Steinberg dithering on for skin and sky — the smooth gradients in a photo are what dithering handles best.
• Check edges at full zoom. If they fizz, switch to Atkinson dithering or drop a few colors. If flat areas show texture, lower the pixel size a step.
• Download the PNG. Do not screenshot the preview or re-save to JPG — either would soften the pixels you just sharpened.