When you export a PDF, Contents, in the Export panel, decides how your labels are drawn inside it. There are three ways, and each is right for something.

Contents
Vector (text as graphics)	Everything is drawn as shapes, text included. Sharp at any zoom, and the PDF looks the same on every computer, because it needs no fonts. You can't select or search the text, though. This is where Labeljoy starts, and the smallest file for a page or two.
Vector (searchable text)	The text is real text: you can select it, search it and copy it out of the PDF. The fonts travel inside the file, which makes a one-page PDF bigger - but a long one much smaller, because each font is stored once instead of drawing every letter again.
Image	Everything is drawn as pictures, at the Image resolution you've chosen. The PDF looks exactly like the work area, whatever it holds, and the file gets bigger the higher the resolution.

Some text is always drawn as shapes, even with searchable text: text with an outline, text bent round a circle, emoji, the readable line under a barcode, and the shadow of a text. And in every mode, pictures and shadows go in as pictures.

Not sure? Keep the first one. Pick searchable text when someone needs to find or copy words in the PDF, or for a document of many pages. Pick Image if something doesn't look in the PDF the way it does on screen.

Contents is saved with the document, and it's one setting for all it exports: change it for pages and label sets and elements follow. It only shows up while Format is PDF - picture formats are pictures anyway.

In a copy of Labeljoy that isn't activated, every page of a PDF is a picture with the DEMO watermark across it, whatever Contents says. See Demo, activation and Care+.