Why changing an image's DPI tag alters no pixels and causes zero quality loss

A print shop recently rejected a high-resolution 4000×2667 pixel photo solely because its metadata listed 72 DPI, even though the image contained more than enough pixels for an 8×10 inch print at 300 DPI. DPI is stored as a metadata tag in the file header — not embedded in the pixel data itself — and exists in different locations depending on the format: JFIF and EXIF fields in JPEGs, and the pHYs chunk in PNGs. Because rewriting this tag only updates a small integer in the header without decoding or re-encoding pixel data, the operation is completely lossless. By contrast, using an image editor's 'resize to 300 DPI' function typically resamples and re-encodes the file, introducing unnecessary quality degradation in formats like JPEG. The correct workflow is to verify that pixel dimensions divided by target DPI meet the required print size, and if so, simply rewrite the metadata tag rather than altering the image itself.