PNG to SVG

What happens to the file size

Expect a moderate change either way. Both formats compress lossily, so the result depends on which encoder handles your particular image better and on how much detail your particular image holds.

Two things drive the number more than the format does. The first is the content: a photograph of a landscape holds far more unique detail than a screenshot of a spreadsheet, and detail is what costs bytes. The second is the pixel dimensions, which no format conversion changes. A twelve megapixel photograph is twelve megapixels whatever container it sits in.

If you need to hit a specific size rather than simply make the file smaller, converting is the wrong tool. Our image compressor takes a target in kilobytes and finds the highest quality that fits under it, which is what upload forms with a stated limit actually require.

Which one do you want

The question is not which is better. It is what the file has to do next.

• A form or a CMS demands an .svg and you have a photo or a screenshot. Exact copy. It is the only one that will look right.
• You want a logo you can scale, recolour and edit. Traced shapes, then open the result in a vector editor.
• You are printing it large. Traced shapes, if the artwork is flat enough to trace. An exact copy prints at exactly the resolution the PNG had.
• You are not sure. Exact copy. It cannot make the picture worse, and you can always come back and trace it.

Getting a clean trace

If you have chosen tracing, the input matters far more than the settings do.

• Flat colour traces well, gradients do not. Each colour becomes a region, so a smooth gradient comes out as a series of bands.
• Start with clean edges. A blurry or heavily compressed source produces ragged outlines, because the tracer follows what is actually there rather than what you meant.
• Match the palette to the artwork. Two colours for a silhouette, four for a flat logo, eight for most things, sixteen or thirty two when the picture has more going on and you are willing to accept more paths.
• Edge accuracy trades points for smoothness. Sharp follows the outline closely and produces a larger file, smooth rounds off the wobbles and produces a tidier one.
• Treat the result as a starting point. A trace opened in a vector editor and tidied by hand beats any automatic trace on its own.

Your filename is kept

The converted file keeps the name you gave it, with only the extension changed. Nothing is appended, no brand is inserted and no counter is added. These files get uploaded straight into forms that sometimes check the name, and an unexpected one is one more thing that can be rejected for no good reason.

There is no watermark either, at any size, on any number of files. A watermark on a free converter exists to sell you the paid version, and we would rather not have that conversation with you at all.

Converting several files at once

Drop as many files as you like. Each one is converted separately and appears in the list with its original size, its new size and the difference between them, so nothing about what happened is hidden from you.

Download them one at a time, or use Download all to take the whole set as a single zip. The archive is built in your browser too, from the files already sitting in memory, so it costs one click and no waiting. Images are already compressed, so the zip stores them rather than compressing them again, which would only make the file larger.

If a file cannot be read, that row says so and the rest carry on. One awkward file in a folder of fifty does not stop the other forty nine.

What happens to the metadata

The conversion decodes the picture and encodes it again, so the embedded metadata does not travel with it. That includes EXIF data: the camera model, the exposure settings, the date, and on a phone photograph the GPS coordinates of wherever it was taken.

For most people that is a benefit rather than a loss. A photograph taken at home and uploaded to a public listing carries the location of your home inside it unless something strips it, and very few sites do. Converting the file removes it as a side effect.

If you actually want the metadata kept, for a photography portfolio or a copyright claim, this is the wrong tool and you should keep the original alongside the converted copy. Colour profiles are handled the same way: the image is converted into the standard sRGB space that browsers and the web use, which is right for anything published online and wrong for a print workflow that depends on a wider profile.