There are a couple of areas where PDF handles graphics a little differently than other output formats.
Help+Manual will always try to fit everything you put in your topics within the margins of the PDF page if possible. Images in tables will be scaled to fit in the table cell containing them and images on the page will be scaled to fit on the page.
If you want to avoid this scaling effect you need to plan your images so that they are small enough to fit in their containers without scaling (table or page).
When you scale an image in the Help+Manual editor it is normally physically resized when you publish it to an output format in Help+Manual. This is handled differently in PDF. Scaled images are not resized. Instead, Help+Manual exports the full original image file with a scaling command. The image is thus displayed at the correct size relative to the page in the PDF viewer, but you get better quality when the user zooms the viewer and images will also have better quality in printouts of the PDF page.
Doing this also increases the size of the PDF file, of course, because the larger images are stored in larger files. You can reduce the impact of this by using SVG vector images where possible, because these scale to large sizes at full quality without needing more data in the file.
Normal bitmap graphics are converted to compressed BMP images in PDF output. PNG images and SVG vector graphics are supported natively and will be included in the PDF as original PNG and SVG files if possible.
PNG with alpha transparency
PNG images with transparency will now show what is beneath them in the transparent areas instead of a simulated white background.
SVG vector images support depends on image complexity
SVG images are tested for their complexity. If they contain gradients, which cannot be reproduced in PDF, the SVG image is exported as a bitmap. If there are no complex gradients, H+M renders the SVG natively.