How to Use This Tool
Choose an image and download the PDF. Nothing is uploaded — the file is assembled byte by byte in your browser.
A PDF has no pixels
This is the part that catches people. A PDF page is measured in points, at 72 to the inch, and an image is placed into a rectangle of that page. The image's own pixel dimensions do not appear anywhere as a size — only as a resolution.
So the effective DPI is a property of the placement:
DPI = image pixels ÷ placed size in inches
A 2,480-pixel-wide image across the full width of an A4 page (8.27 inches) is 300 DPI, which is print quality. The same image placed across an A3 page is 212 DPI, which is noticeably softer. Nothing about the file changed.
The panel above shows the figure for your image and page. Roughly: 300 or above is print quality, 150 is acceptable for a document, and below 100 will look soft in print while still being fine on a screen.
The JPEG goes in unchanged
PDF supports the DCTDecode filter, which means a JPEG can be embedded exactly as it is — the compressed bytes are copied straight in. No decoding, no re-encoding, no additional quality loss, and the PDF ends up about the same size as the original file.
PNGs are different. PDF cannot embed PNG data directly, so a PNG must be decoded and re-encoded. This tool converts it to JPEG at the quality you choose, which means:
- Transparency is lost, because JPEG has no alpha channel. It is flattened onto white.
- Screenshots and line art suffer, since JPEG produces fringing around sharp edges. A PNG of a document is often better kept as a PNG.
Why PDFs cannot be concatenated
After the objects comes a cross-reference table listing the exact byte offset of each one, and a trailer pointing at that table. A reader uses the table to jump straight to any object without scanning the file.
It also means the offsets are only valid for the exact bytes in front of them. Insert anything, or join two files, and every offset after the change is wrong — which is why merging PDFs requires rewriting the structure rather than appending, and why a truncated download usually will not open at all.
A scan is not a document
A PDF containing a photograph of a page has no text in it. You cannot search it, select from it, or read it with a screen reader, because as far as the format is concerned it is a picture.
Making it searchable requires optical character recognition, which recognises the characters and adds an invisible text layer behind the image. That is a separate operation and not something this does. If the original is available as text, converting from that will always produce a smaller, searchable and more accessible file.