The quick experiment.
Hypothesis: Different font faces (namely sans/lighter) will increase contrast & decrease reflection on the Sony PRS 700
Results: Having a lighter font seems to decrease the amount of reflection seen on the device (less of the nighttime window effect) and clarity (less optical color mixing); however, probably not worth the extra effort given current methodology for setting/changing the eBook's font-face.
The Experiment:
Created a PDF document - using Lorem Ipsum to not distract most viewers with the actual content/context of the text.
Used Calibre to convert PDF to LRF - overriding the CSS as per Zelda's post noted above.
Uploaded fonts to PRS 700 (see Zelda's post).
Uploaded eBooks.
Set book in two differing lighting conditions.
Took picture of each eBook - in as similar setting as possible.
Collaged images together in photoshop (no other editing was performed on the images - straight from camera to box - to image file).
The top row of images were taken in my bathroom - consistency of light, and no sunlight. Using a tripod & Canon Rebel XTi.
The bottom row of images were taken outside - overcast at the time - and in as quick succession as possible (to maintain light).
Fonts used: Georgia, Courier New, and Verdana.
Issues/Variables:
The Georgia font which was displayed did not have the signature tail for the capital Q for that font - which makes me wonder if this is really Georgia.
Arial was attempted; however, the line thickness seemed exaggerated/blurred in comparison to the sans-serif (arial/helvetica) font used by the device's menus.
Verdana was used because the device better maintained the line thickness of the font.
The x-height of Georgia is slightly less than that of Verdana or Courier New. Therefore, there are more characters appearing on the Georgia page than the others. (Haven't counted.)
The Serif Baskerville font is decidedly lighter than that of Georgia; therefore, might have similar results as the Verdana. However, the Baskerville I have is not in TTF format (is DFONT) - which seems to be the font-type the device can read (from this initial test anyway).
Cheers,
Josh
ps. Might need to click on the image a few times to get to the large (original download) version.
