What a request for 300 dpi is actually asking
It is two requests wearing one number. The first is administrative: the file must state 300 in its resolution field, because a preflight script reads that field and refuses anything else. The second is substantive: at the size the piece will be printed, there must be about 300 pixels for every inch of paper. A file can satisfy the first and fail the second, which is why a submission is sometimes rejected twice for what looks like the same reason.
This route opens with the resolution field set to 300 and the operation set to leave the pixels alone, because that is the correct answer to the administrative half. The panel then reports the print size those pixels support, which is the honest answer to the substantive half.
The print size your file already supports
Divide either pixel dimension by 300 and the result is inches of paper. The table below runs that division on the counts that turn up most often, so a file can be matched against a target before anything is exported.
| Pixels | Inches | Centimetres | Nearest common size |
|---|---|---|---|
| 600 x 600 | 2.00 x 2.00 | 5.08 x 5.08 | a 2 inch ID photograph |
| 1050 x 1500 | 3.50 x 5.00 | 8.89 x 12.70 | 3.5 x 5 print |
| 1200 x 1800 | 4.00 x 6.00 | 10.16 x 15.24 | 4 x 6 print |
| 1500 x 2100 | 5.00 x 7.00 | 12.70 x 17.78 | 5 x 7 print |
| 2480 x 3508 | 8.27 x 11.69 | 21.00 x 29.70 | A4 |
| 2550 x 3300 | 8.50 x 11.00 | 21.59 x 27.94 | US Letter |
| 3600 x 5400 | 12.00 x 18.00 | 30.48 x 45.72 | 12 x 18 print |
Read it the other way for a scan: a full A4 page captured at 300 dpi is 2480 by 3508 pixels, and a scanner set to that density on that paper produces exactly that grid.
When 300 needs more pixels than you have
The arithmetic is unforgiving in this direction and it is better to see it early. A 1024 pixel edge supports 3.41 inches at 300 dpi and nothing will change that except more pixels. Three responses are reasonable, and only the first two are free.
The two honest constraints:
- Print it at the size the pixels support, or accept a lower density for a piece read from further away. A poster at 150 dpi viewed from a metre is completely ordinary practice and looks correct; a postcard at 150 dpi does not.
- Resampling upward to reach the pixel count will let the file satisfy an automated check, and it will not add anything a printer can resolve. The panel will do it if you ask, and it will keep saying that the extra pixels are interpolated for as long as they are on screen.
Questions about the 300 figure
- Where does the 300 figure come from?
- From the acuity of the eye at reading distance. A person with ordinary vision resolves roughly one minute of arc, which at about 30 cm works out near 300 samples to the inch. Above that the print has structure no unaided eye can find; below it, at that distance, the grid starts to become visible. It is a convention built on a measurement, which is why it has survived unchanged through several decades of printing technology.
- A form asks for a 300 dpi photograph. Is the pixel count implied?
- Yes, once the physical size is stated. A passport-style photograph specified at 35 x 45 mm and 300 dpi is 413 x 531 pixels, and no other pair of numbers satisfies it. If a form gives only a density and no size, it is asking about a document scan, where the paper size is already known.
- My file is 1000 x 1000 and the panel says 8.47 cm. Can I make it print larger at 300?
- Not without adding pixels, and pixels added by a kernel are computed rather than recorded. 1000 pixels at 300 dpi is 3.33 inches and that is arithmetic, not a limitation of this tool. The realistic options are printing it at 8.47 cm, accepting a lower density for a piece that will be viewed from further away, or going back to a higher-resolution original.
- The vendor rejected the file even though it now says 300. What else could be wrong?
- Most often the pixel count against the placed size, which is a different check from the density field. A preflight computes the effective resolution at the size the image sits on the page, so a file marked 300 dpi that has been scaled up to fill a frame reports its effective figure, not its stored one. The panel here shows the achievable resolution so the two can be compared before submission.
- Should the field say 300 or 300.0? Does the precision matter?
- The field is written as a whole number over one — 300/1 in the Exif rational, 300 in the JFIF density, and 11811 pixels per metre in a PNG's pHYs chunk, which reads back as 299.9995 dpi because a whole number of pixels per metre cannot land exactly on 300 — a discrepancy of under two parts in a million. Anything that reads the file gets a clean integer rather than a repeating fraction to interpret.