A picture can look amazing as a digital poster but print out soft, aliased, or jagged. In this case, the problem isn’t the image itself; it’s the size that’s used. When a raster image is enlarged, the existing pixels must be spread over a larger area, meaning you’ve got fewer pixels to print. DPI, dots per inch, is a commonly used measure of resolution (though a piece of software may call image resolution by pixels per inch). A file with a great deal of pixels will allow a larger physical print, whereas the same would not work for a small web image, for example.
However, a pixel count listed in the metadata of the original image doesn’t necessarily tell the whole story. What matters in prepress is the effective resolution of the image in the layout. If a photograph were scaled to 50% of its original size, its pixels would be forced into a smaller area, effectively increasing its DPI. Scale that same photograph to 200% and the effective DPI falls; hence, just checking the original image file could be deceiving because the image might technically be large enough but, if stretched to fit, might now be larger than can be printed well.
Also worth noting is how close the final printed image will be viewed. An image placed in a business card, brochure, or flyer is viewed up close, so photos and raster-based graphics typically must have enough pixels in a certain area to maintain detail. A poster image may be viewed from farther away so a lower effective resolution might be acceptable. Ultimately, the print provider’s specifications for the project should provide you with guidance, because paper stock, press method, content, and viewing distance are all considerations that might influence the answer.
To test this, you can place a given photo in several copies of the same small file. Have one version at or near the original size, one at less, and one much larger. Check the effective DPI for each in the file or preflight tools and print proofs at full scale. Take care to examine the details, such as hair, small letters, texture patterns, and diagonal edges; these reveal softness and jagged pixels better than a clear sky or blurred background might.
There’s no point in trying to “fix” a poor-quality image by entering a higher value for DPI into the image’s properties; doing this does not magically make the image crisper, and while you can sometimes resample to get better image sharpness, this can only be done to a certain extent, and you certainly won’t recover information that wasn’t in the original image to begin with. Therefore, prior to PDF export of a print, be sure to inspect every raster image in the final dimensions you’re going to use and either replace, reduce, or carefully crop any image file that doesn’t meet your project’s needs.