First, let’s tackle dpi vs. PPI (note lower case and upper case!).
The term ‘dpi’ (dots per inch) refers to scanner input resolution. That’s how many dots (or “cells”) the scanner creates horizontally as it scans down the film or paper. For an 8x10" original to be reproduced at 1:1, a scanner set to scan at 300 dpi will save a file with 2400 by 3000 pixels. The EXIF header of that file may say “300 dpi” in it, but you can reproduce those 2400 by 3000 pixels at any PPI.
At the monitor, ‘dots’ are the individual red, green, or blue LEDs or phosphors that create the image. Monitor dot pitch (distance across each dot) varies with monitor size. Monitor PPI is a measure of RGB pixel output across an inch of screen. Monitor dot pitch and PPI are different.
At the printer, ‘dots’ are used to reproduce pixels. In high resolution Epson inkjet printing, there may be as many as 5760 ink dots per inch, although 2880 or 1440 is plenty for most uses. From the file pixels, the printer driver will create an array of 300 or 600 pixels per inch to be printed, and then create the dot pattern for each pixel in the array. MANY dots may be used to represent one single pixel!
The term ‘PPI’ (pixels per inch) is used to refer to output resolution. In a scanner, dots are used to create pixels. In a printer, pixels are used to create dots. ‘PPI’ should be used as a measure of how many pixels in the ORIGINAL file (processed from raw data) are used for each inch of printed output.
In a digital camera, sensels are used to gather light, and post-processing by the camera or external software creates the pixels using lots of higher math.
Pixels are just numbers in files… THEY HAVE NO SIZE unless you specify PPI for a specific array.
When scanning, consider the ultimate use of the file. Scanning film at 300 dpi works for an 8x10 inch sheet of film reproduced at 8x10 inches on a printer. But scanning full frame 35mm film at 300 PPI yields very bad results when enlarged to 8x12"! Instead, you need a file scanned at about 8.5 times 300 dpi. That’s roughly the diagonal difference between 43.266mm (35mm film frame) and 14.42 inches (8x12 print). So round it to 2400 dpi, which is a setting on most flatbed scanners.
When I print, I calculate the EFFECTIVE PPI from the original full size image, with no interpolation or downsizing. My camera files process to 4608 by 3456 pixels. I want to feed my lab or printer a MINIMUM of 240 of those per inch of paper, at 8x10. So I make sure that, before cropping, I know my ultimate effective PPI.
When a print will not be viewed closer than its diagonal dimension, effective PPI can be lower for larger print sizes. A 16x20 at 180 or 200 PPI will look fine at 26 inches, provided the photography was done well.
Both Photoshop and Lightroom Classic have new tools for “Super Resolution.” they work surprisingly well, if you have a fairly large image to start with. So if you’re making posters, try it!