Image to PDF Without Losing Quality

What “without losing quality” honestly means

There are three different things it can mean, and most converters quietly pick the weakest one.

A baseline JPEG loses nothing at all here. PDF's DCTDecode filter is JPEG, so the compressed data your camera wrote can be placed into the document unchanged. No decode, no re-compression, no downscale, no quality slider in that path. The file inside the PDF is the file you started with, byte for byte.

A PNG loses no pixels, but it is not the same bytes. PDF has no PNG filter, so the image is decoded and stored with Flate compression instead. That is lossless — every pixel keeps its exact value, transparency included — but if someone tells you a converter embeds PNG “byte for byte”, they are describing something the format does not allow.

Everything else is re-encoded exactly once. HEIC, WebP, AVIF, TIFF, GIF, BMP and SVG have no representation inside a PDF, so they must be decoded and written again. Photographs go to JPEG at 92% quality, because a lossless page from a 12-megapixel HEIC would be around 30 MB and unusable. Graphics, screenshots and anything with transparency go to PNG. You can force lossless for everything under More options.

The report you get afterwards

Every conversion lists each file with the route it actually took. That matters because the route can change: asking for a smaller file, or setting a size limit, deliberately gives up the untouched path, and a badge that guessed from the file extension would then be lying to you. The report is read from the conversion itself.

Three JPEGs that cannot take the fast path

• Progressive JPEGs. PDF's DCTDecode is defined for baseline JPEG. Progressive streams render inconsistently in print pipelines built on Ghostscript, so those files are re-encoded as baseline instead of being embedded and hoping for the best.
• CMYK JPEGs, which come out of print workflows, are converted to RGB.
• Mirrored photos. A rotation can be expressed as a PDF transform; a mirror cannot. Four of the eight EXIF orientation values are mirrored, and those files are decoded so they come out the right way round. See why images come out sideways.

What we do not carry across

Colour profiles. If your photo is tagged Display P3 or AdobeRGB, the numbers survive but the tag does not, so a wide-gamut image can look flatter on screen. The pixels are unchanged; the interpretation is. The full explanation is here, including when it matters and when it does not.

If your PDF looks soft rather than merely different, that is usually a resolution problem rather than a compression one — our sister site has a guide on blurry PDFs.