PDF to PNG: clear diagram copies are still pictures
When explaining a diagram or reviewing a mockup inside a PDF, you may need a page on which to draw arrows and comments. PNG provides that visual working copy, but it does not retain unlimited enlargement of the PDF’s vector objects.
Lossless storage begins after the page becomes pixels
PNG uses lossless compression for its image data. The page is first rendered onto a pixel grid at a chosen resolution, however, so lossless storage does not mean every vector detail remains scalable to any size.
Keep the source for later resizing or information extraction. A small picture should not become the only remaining source of a diagram you need to print large.
References: W3C — PNG lossless image storage
Choose resolution around the smallest important detail
Inspect a small number, thin line or color key. Increasing resolution may help those elements even when the page title already looks perfectly clear.
Compare versions at the same intended display size. Enlarging one while inspecting only a thumbnail of the other is not a fair comparison.
A picture is not a table or a layered drawing
Text inside PNG is not a set of sortable cells or separately editable diagram objects. Placing the image in a presentation does not make individual words editable.
ToolRoof produces one image per page here; it does not promise automatic diagram cropping or a transparent background. Check edges, colors and page sequence. Take numerical measurements from a reliable source, not a picture displayed at an unknown scale.
Example: a fine line on a teaching page
For an illustrative page, a 0.5-point line has a calculated width of 0.5 pixels at 72 pixels per inch, using 72 points per inch in this example. At 144, its calculated width becomes one pixel.
Rasterization and edge smoothing can distribute color between pixels, so this does not promise a particular count of black pixels. It explains why a fine line deserves visual inspection before using the page for teaching or display.