As an interesting aside, what happens when you send a file to an outside lab for printing, especially if you do not know which printer they are using and what its native resolution is? Inferred from the above, it means that the lab computer's OS would do the resolution changes? How would you then optimize the file before sending (if at all)?
Hi Slobodan,
Excellent point, assuming one subscribes to 'the OS level print pipeline does resampling' idea, which I don't. The sometimes printer specific print application, other times just the printer driver, will do the resampling, and can be set or disabled. Applications like Qimage will do all resampling, add a user selectable amount of sharpening at the pixel level
after resampling, add optional dithering for profile conversions
after all that, and then sends the data stream to the printer driver which only dithers the ink-colors. Maybe the Mac OS adds some complications by enforcing some additional color management.
Some printer drivers even allow to adjust the upsampling quality, with edge enhancement and such. I don't see how the OS could do that in different ways for different printer driver algorithms.
When outsourcing files for printing, I prepare them with the colorprofile and native resolution of the printer that will be used for the actual print. It's not always easy to get that info, but serious print operations are happy to assist by making profiles available and will share info about the equipment they use (and which settings they use, sometimes a compromise for production speed!).
Cheers,
Bart