View Full Version : Benchmarking PDF speeds on ebook readers


Jürgen Hubert
01-05-2010, 12:02 AM
I'm curious how well the same PDF works on different PDF-compatible ebook readers. For testing purposes, I recommend uploading this PDF (http://rapidshare.com/files/277090513/CAT21000_EclipsePhase.pdf) (a tabletop role-playing game book) on your ebook reader and testing how smoothly your pages turn with it - it's one of the most graphics- and layer-intensive PDFs I know and my iRex DR1000S choked on the PDF fairly frequently.

(Note: Yes, the above download link is located at Rapidshare. However, the entire PDF is under a Creative Commons Share-Alike license - as outlined on page 7 of the PDF - and the authors did in fact upload the PDF to P2P sites themselves, so downloading it is entirely legal.)

frabjous
01-07-2010, 03:56 PM
I had trouble getting that file from rapidshare. I ended up getting it from a torrent. I've split it into a two-part .rar file, so it could be uploaded here. I think dealing with rapidshare and similar things is too much trouble.

It takes about 5-10 seconds to go from one page to another on my Sony 505. The font size is too small for a comfortable read anyway. I didn't spend too long with it.

Were there particular benchmark tests you wanted us to try?

AJ Starr
01-07-2010, 07:40 PM
Just as a learning experience with the rar files, I did as you suggested.

Took me a second try to realize you had to combine the two rar files when you unzipped them, but all went well.

My Astak 5" Pocket Pro performed just fine.

After the extensive detailed TOC (which my PP wanted to go to each link, since we can go to links in pdfs) I got to normal pages. It averaged my counting 5 to 6 seconds for most pages. I had to make sure the page completed the loading before I pushed the next page button again, or the PP went ahead and changed pages again.

However, since the file is made up of the individual pdf pages, not reflowed text, even the 9 levels of zoom for the PP would not do anything. Not even the rotate screen allowed the text to get big enough.

Hope this contributes to you benchmark.

AJ