… email on dots per inch, pixels per inch and the process of sending a digital photo to a printer…

“When the digital photo (bitmap) is in the digital camera, there is no "per inch”. This bitmap is merely a “mosaic” made up of a finite number of horizontal and vertical pixels. The “per inch” is an artifice that is introduced when we start to think about the size of the image when going to print, usually in an image editor.

I’m assuming our goal is to print the image so that it appears as sharp as possible. The key word here is "appears". At this juncture, all digital camera images will need to be up-sampled for large prints (resampled to a larger horizontal and vertical pixel count.) How much resampling depends on the size of the print, the original size of the image in pixels and what the optimal resolution that is needed for that particular printer. Whether you resample or not, we should use unsharp mask to maximize the appearance of sharpness (acutance). That’s a voodoo art form whose discussion is best left for another day. It could easily be argued that if you are NOT going to use unsharp mask you should let the printing software do the resampling. More later.

The previous posts are correct in suggesting that the printer resolution is merely a statement of “dot-addressable per-inch density” on the final printed page. An Epson 2200 printer can spray any one of 7 colored droplets every 1/2800 inch horizontally. These printer ink dots (droplets) are used to dither the varying sized individual ink drops. Dithering is when a printer places many microscopic drops in close proximity to produce these varieties of colors; much like a Seurat painting. This dithering produces any of our 16 million colors in the final 24bit digital photo printing. Thus, while your output resolution might be 300 pixels per inch (300h300v), the printer could be using 2880X1440 dots per inch to create this. This very high PRINTER resolution allows for very smooth dithering, but from a practical standpoint this printer res has little relationship to the output resolution you choose in your photo editor. That last statement could easily be challenged but I’ll let it fly for purposes of simplification. How the printer driver actually does the dithering and the accuracy of the dot placement also determines acutance.

You can read in other posts that all printers have a “native” resolution and your images are always upsampled to that native res whenever you submit your image to the printer driver. That’s why so many people have success at using a lower output resolution. They are letting the printer driver handle the entire resampling. I personally have a problem with this strategy, as I believe I can control the sharpening better if I’m in the ballpark to the native printer resolution. As an alternative, the printing program Qimage has you work in the original un-resampled bitmap and resamples at the time of printing using their very sophisticated resampling algorithms. I believe they supply the printer driver with a bitmap that does not need to be resampled thus preserving the integrity of their processing.

There could be another layer of complexity if you are turning your images over to a service bureau to be printed. The pragmatic approach is to talk to them and ask THEM how they want the images supplied. They should tell you what they want for PPI and you should accommodate them. I believe you should always crop your own images because anyone else will ALWAYS screw it up. PhotoShop (and I assume PSElements) has a wonderful cropping tool. It allows you to type in the output resolution (say 300PPI) and the desired picture size and any crop you make will be automatically resampled appropriately. Unsharp mask and ship it out.”