Confused by U98g2 Font Descent values

[gentlesartori]

I am able to parse all of the U8g2 font data without apparent problems. There seems to be a problem with my parsing of the first 23 bytes header, or my understanding of it anyway. As an example, I am using font data from u8g2_font_5x7 from u8g2_fonts.c. The header (first 23 bytes) in Octal looks like this:

\277\0\2\2\3\3\3\4\4\5\7\0**\377**\6**\377**\6**\0**\1\22\2/\6p

For calculating the Y position of glyphs, I am interested in the descent values. The highlighted values above (in order) correspond to "y offset (descent)", "descent (lower g)", and "descent ')'". In decimal, these values are 255, 255, and 0 respectively. A descent of 255 seems wrong to me, so I am wondering if these bytes should be interpreted differently somehow - perhaps with a sign?

Edit: I realized that each glyph has a bitmap Y Offset as well. Perhaps this is the value that I should be using (makes sense). But - looking at ASCII 36, the bitmap Y Offset is again large at 254. There must be some other calculation that I am missing. Perhaps y offset (from the header) - bitmap y offset?

Please help with my confusion. I am very close to full comprehension of u8g2 fonts!

Many thanks,

Gentle

[olikraus]

Yes 255 means -1. Value 255 is the 8bit 2-complement. Mathematics: if v > 127 then v= 255-v-1 where v is the byte value

[gentlesartori]

Ah! I had a feeling that it was something like that. I hadn't picked up on the +/-127 range. Can I assume that all offset bytes in the font header are bipolar as well?

Thanks or all your help!

Gentle!

[olikraus]

Usually yes, except memory offsets

[gentlesartori]

Right, got it.

Just to check my math;
Since 255-255-1 = -1

then for v(U8)=128, v(I8)= 255-128-1 = -126

Is that correct?

G

[olikraus]

hmm maybe my own formula had been wrong. Correct description should be here:
http://www.cs.emory.edu/~cheung/Courses/255/Syllabus/5-repr/2s-comp.html

[gentlesartori]

Thank you for that information. I also found that my dev environment has a primitive function to convert U8 to I8 that takes care of this for me.