Longshadow wrote:
300 PPI or 300 DPI?
Here we go again...
Pixels are numbers without dimension. The same file of pixels can be reproduced to any size output. You might not LIKE that output if you don't have enough pixels, but you can still reproduce them!
Dots are physical points on a flatbed scanner bed, or ink spots sprayed onto paper, or exposed points of light projected onto light-sensitive paper. PPI refers to how many pixels of original image file data are spread over each inch of output. Original means un-interpolated, as in not artificially re-sized. PPI is usually used to determine whether an image will enlarge well to the desired print or display size.
dpi refers to how many "cells" exist in each inch of a scanner bed. A 300dpi scan divides the flatbed scanner glass into 300 square cells per linear inch, both horizontally and vertically, in a grid pattern. Typically, this is converted to a file by the scanner driver software containing the same number of pixels, assuming a 1:1 or 100% of original size file definition.
> If you tell the scanner driver, "Scan this 8x10" print at 300dpi at 100%," you get 2400x3000 pixels in the output file.
> If you tell the scanner driver, "Scan this 8x10" print at 300dpi at 200%," you get 4800x6000 pixels in the output file.
> If you tell the scanner driver, "Scan this 8x10" print at 600dpi at 100%," you get the same 4800x6000 pixels in the output file.
Once the image is digital/virtual/just pixels, "dpi" is meaningless until the image is printed on a SPECIFIC device. Some will argue that it is still meaningless. However, there are some graphic arts applications where dpi information in a file is used to size an image for page layout when it is imported onto/into a page layout.
[b]Besides scanner INPUT resolution, dpi also refers to printer OUTPUT resolution.[/] How many
dots per inch does your printer use to reproduce the
input pixels you send into its driver? You can send a file for reproduction at 8x10 inches at 300 PPI, but the printer driver converts that to some number of dots native to its design and technology. A desktop printer might spray 5760x2880 dots per inch to reproduce a 300 PPI photo. It might be able to print the same 300 PPI photo at 2880x1440 dots per inch in 1/4 the time, without too much reduction in image quality.
Except for Sigma Foveon-equipped cameras, digital cameras typically use an array of monochrome sensor elements (sensels or sub-pixel sensors), covered with color filters, to convert photons to electrons. The output signal from each sensel is digitized, then processed in camera or externally to convert sub-arrays of adjacent sensel data into discrete "bit maps" of pixel values. From that point forward, the file is just like a file from a scanner. Any "dpi" setting in the file is just arbitrary to many software applications.
If you hang around the graphic arts and photographic industries long enough, you will shake your head and laugh at the confusion over these concepts.
Dots are not pixels. Pixels are not dots. Pixels are digital. Dots are analog. It's bits vs. atoms, folks!