Hello maco. At the default (lie) of "96 dpi/(72 pt/inch)" there are is a 4:3 relationship between the pixel-count (dots) and the nominal point size (points). In the UbuntuBeta Mono and Inconsolata the height (line-spacing) of the font is exactly 1 em with 2:1 height:width ratio and all spacing is internal. The font rendering is dependent upon the number of pixels available for drawing within that fixed height called the pixels-per-em ("PPEM"). So, at 96 dpi:
8 ppem == 6 point (getting close to the hinting limit; only 4×8 pixels)
12 ppem == 9 point ('font.png' screenshot above; 6×12 pixels)
16 ppem == 12 point (Console/VGA size; 8×16 pixels)
Bi-level 8×16 pixels is the really important one for bitmap console use. However, with sub-pixel rendering you get slightly better granularity for monochrome text in one axis (generally horizontal) at the cost of colour-shifts.
The attached screenshot is at 4×8 pixels unhinted (hinting has not been done yet), which was more usable than I expected and probably would gain from visual recognition in regular use if you really need a terminal with 350×130 character cells. I think Amélie prepared a addition '1' possibility with a slightly wider slab serif but with the monospace this time of thing can be refined afterwards. How does it look to you?
Hello maco. At the default (lie) of "96 dpi/(72 pt/inch)" there are is a 4:3 relationship between the pixel-count (dots) and the nominal point size (points). In the UbuntuBeta Mono and Inconsolata the height (line-spacing) of the font is exactly 1 em with 2:1 height:width ratio and all spacing is internal. The font rendering is dependent upon the number of pixels available for drawing within that fixed height called the pixels-per-em ("PPEM"). So, at 96 dpi:
8 ppem == 6 point (getting close to the hinting limit; only 4×8 pixels)
12 ppem == 9 point ('font.png' screenshot above; 6×12 pixels)
16 ppem == 12 point (Console/VGA size; 8×16 pixels)
Bi-level 8×16 pixels is the really important one for bitmap console use. However, with sub-pixel rendering you get slightly better granularity for monochrome text in one axis (generally horizontal) at the cost of colour-shifts.
The attached screenshot is at 4×8 pixels unhinted (hinting has not been done yet), which was more usable than I expected and probably would gain from visual recognition in regular use if you really need a terminal with 350×130 character cells. I think Amélie prepared a addition '1' possibility with a slightly wider slab serif but with the monospace this time of thing can be refined afterwards. How does it look to you?