PPI is used as a method of defining image resolution in relation to document size. You have your pixels per inch, height in inches, and width in inches. When you multiply your PPI across the document width, you get your resolution in inches.  So what we have here is an equation.

PPI x Width in inches = Width in pixels

And the same goes for height. Scott said he wanted images that were 640px wide by 72ppi. We have now two defined elements of our equation, which will now look as such

72ppi x Width in inches= 640px

or

Width in inches= 640px  / 72ppi

If we change 72ppi, the only thing that will change is the document size, which does not impact filesize or image quality when views on the web.

Lets see this inaction though.

We can see here that our previous math works out fine(if you did it ) and our document size ends up being 8.88. Now how about changing the PPI to 240.

It looks like our math still works, as the document size becomes 2.66. I guess the universe doesnt end today after all :/ So now the big question. How do these files save out. Is one bigger then the other, or does one look any different?

Nope. Same file size. Identical display. So why are Scotts images larger at one dpi than the other? He cheated. He didnt follow his original rules of having the pixel dimensions at 640px. He simply resized the images and said PPI did something magical, when it didnt. I guess he didnt check what the image resolution of his images were. PPI does not matter when viewed on the web. He ends by saying:

“Sometimes being pedantic gets in the way of actually knowing stuff that matters”

Well allow me to retort. Some times pretending like you know what you are talking about gets in the way of learning something new.

Recently I have grown increasingly annoyed at some issues with image quality and today I am sorely disappointed in apertures RAW decoding.

The problem seems to occur when an over exposed image has a smooth fall off into a normally exposed area causing posterizing around the highlight area. It is most noticable in certain tonal ranges as we see. The worst part is Adobe Camera Raw has non of these problems, neither does DPP. It turns out that the problem is also noticable in OSX’s Finder previews, but to a much lesser degree. This seems to imply to me that the problem goes beyond just Aperture, and is applicable to everything that uses Apples Camera Raw decoding. It also means Aperture is doing something special in addition, which is exacerbating the problem.

Higher res jpeg_examples available.