I am talking about images--I'm converting PDF's to JPEGS because of the extra zoom capability--and matching up DPI/Aspect Ratio, etc is critical to making a "readable" oversized (12.5"x22") image. This is only tangentially related to PDF's -- they're just pictures on the e-reader. I only mentioned PDF because of my use case--keeping detail through the conversion.
It may be easier to just show you what I'm talking about, I'm uploading some examples for you to look at. I rendered them at 100% jpeg quality so that jpeg artifacting wouldn't be an issue.
Here is a sampling of the original PDF's. (4 pages, 2.9MB)
View the following on your ereader in Large image mode:
Here is a sampling of the images at 83 DPI (8 pages, 1037x913 each,7MB)
Here is a sampling of the images at 166 DPI (8 pages, 2075x1826 each, 21MB)
Here is a sampling of the images at 332 DPI (8 pages, 4150x3652 each, 57.3MB)
Notice that with each example, the images double in size, yet you're not having to scroll around twice as far-the same amount of the image is shown, regardless of the original image size.
So what image size is the e-reader downsampling to? Because that would be the resolution that you could get the greatest detail in an image.
So, this question can also be asked in another way--what's the optimal size to use in JAP when exporting to jpeg? 1600x1408? 2075x1826? Or are the images to large, causing to e-reader to do its own downsampling, causing artifacts in the image?