PNG to PDF without blurry text

Blur is not added by the PDF format. It is added when an image is converted to a lossy format such as JPEG, when it is resampled to smaller pixel dimensions, or when a small image is stretched to fill a large page area.

In Original quality mode each PNG is embedded in the PDF as PNG image data. There is no JPEG step and no downsampling, so a crisp screenshot stays crisp and flat colour stays flat, without the fuzzy halos that JPEG leaves around lettering.

Transparency is kept too, stored as an alpha channel rather than being flattened into a solid block.

Smaller file re-encodes each image as PNG with the longest side limited to 1600 pixels. It is useful for photographs where a smaller file matters more than fine detail. For text it is the wrong choice: small lettering that was sharp can become soft after the reduction, and the tool cannot restore it afterwards.

Cropping to fill the page means part of the image is cut away and the rest is scaled to cover the page. If the source is a narrow screenshot on a wide page, that scaling can be extreme. Fitting the whole image inside the margins usually keeps text sharper, at the cost of white space around it.

A PDF cannot hold more detail than the image you gave it. If a screenshot was captured at a small size, no setting in this converter invents the missing pixels; the only real fix is to capture again at a larger size or at a higher display scale.

Original image size pages use the PNG's own pixels-per-metre value when the file carries one, and assume 96 PPI otherwise. That value describes how large the image is meant to be printed, not how much detail it has.