ink/letter
Jacob Langford
Eric Angelini wrote:
>
> hello,
> I'm trying to reorganize the alphabet, starting with the letter which needs
> LESS ink to the letter which needs MAXIMUM ink.
> Having no method to evaluate the surface of a letter, I bought a
This depends on which font you use. You may get a different answer
for different typefaces and then again within each typeface you may
get a different answer for different fonts (because Roman 10pt is
*not* the same as Roman 5pt doubled!).
I don't know much about MetaFont, but maybe you could use MetaFont
to generate a set of alphabets. Ask Donald Knuth.
Jacob (lang...@uiuc.edu)
Randy DeWoolfson
Eric Angelini wrote:
>
> hello,
> I'm trying to reorganize the alphabet, starting with the letter which
> needs LESS ink to the letter which needs MAXIMUM ink.
[snip]
Well it definitely depends on the font and size of font.
And, since I am particular to computers, (as opposed to typesetting
in some other fashion), I would count the active, or dark, pixels
in the letters of the chosen font to give the order.
This assumes that an even amount of ink is used for each active dot,
so more dots = more ink...
you could write a simple(ish) program (for Windows anyway) that
would create a font, Write one of each letter onto the display
context, then count the number of pixels active for each one,
then sort them...
Randy :)
myles
"Eric Angelini" <keyne...@skynet.be> wrote:
>
>
>
>hello,
>I'm trying to reorganize the alphabet, starting with the letter which needs
>Having no method to evaluate the surface of a letter, I bought a LETTRASET
>alphabet (letters we (?) use in Europe when we want to do precision
>typographical non-computerized work=7F); the font was an uppercase helvetica
>and
>I started weighing every letter; I came with this order:
>IJLVPYFTSCAUZEOGXDKRQBNHWM.
>Is there an easier way to do that? I've tried to use Photoshop, magnifying
>=7Fletters and then counting the pixels by hand, but... it doesn't seem to me
>that this is the cleverset method=7F...
>Regards,
>Eric Angelini
you could write a program to count the black spaces in a
"picture" of a single letter. actually, i kind of liked
weighing them.
user
Jacob Langford <lang...@uiuc.edu> wrote:
>Eric Angelini wrote:
>>
>> hello,
>> I'm trying to reorganize the alphabet, starting with the letter which needs
>> LESS ink to the letter which needs MAXIMUM ink.
Please satisfy my curiosity: what do you need this information for?
>This depends on which font you use. You may get a different answer
>for different typefaces and then again within each typeface you may
>get a different answer for different fonts (because Roman 10pt is
>*not* the same as Roman 5pt doubled!).
What do you mean by 5 pt doubled? Do you mean printing at 5 pt and
then enlarging by factor of 2?
To add to the above answer, the result can depend not only on the
attributes of a font (bold, italic) but on the size (or resolution)
that is used. A case in point is the following (I recommend a
fixed-width font).
The following results were obtained by counting individual pixels for
Times New Roman (Normal) at various point sizes (as indicated),
on-screen at a screen resolution of 1024x768. Coverage by pixels and
percent space is given.
Notice the order changes:
60 pt: M W B R H D Q ...
120 pt: M W B Q R D H ...
156 pt: M W B H Q R D ...
Size used: 60 Size used: 120 Size used: 156
Pixel space: 6660 Pixel space: 27208 Pixel space: 45668
1206 18.11% M 4805 17.66% M 8283 18.14% M
1133 17.01% W 4506 16.56% W 7544 16.52% W
1059 15.90% B 4006 14.72% B 6961 15.24% B
1036 15.56% R 3976 14.61% Q 6833 14.96% H
1030 15.47% H 3897 14.32% R 6747 14.77% Q
1018 15.29% D 3893 14.31% D 6709 14.69% R
1009 15.15% Q 3859 14.18% H 6657 14.58% D
947 14.22% K 3690 13.56% K 6450 14.12% K
901 13.53% O 3501 12.87% O 5925 12.97% O
867 13.02% G 3325 12.22% N 5693 12.47% N
809 12.15% N 3296 12.11% G 5663 12.40% G
800 12.01% X 3257 11.97% X 5595 12.25% X
783 11.76% P 2932 10.78% P 5133 11.24% P
776 11.65% E 2826 10.39% E 5119 11.21% E
752 11.29% Z 2813 10.34% Z 5110 11.19% Z
725 10.89% S 2807 10.32% U 4833 10.58% U
685 10.29% U 2769 10.18% A 4805 10.52% S
684 10.27% A 2762 10.15% S 4789 10.49% A
656 9.85% F 2621 9.63% V 4541 9.94% V
656 9.85% Y 2593 9.53% Y 4510 9.88% Y
652 9.79% V 2540 9.34% C 4370 9.57% F
637 9.56% C 2451 9.01% F 4253 9.31% C
609 9.14% T 2430 8.93% T 4106 8.99% T
592 8.89% L 2312 8.50% L 3902 8.54% L
495 7.43% J 1877 6.90% J 3267 7.15% J
476 7.15% I 1802 6.62% I 3156 6.91% I
Obviously, the results will vary depending on the size and resolution.
If you are talking about actual INK on PAPER, be aware that printer
resolution is typically 600 dpi and the pixel count would be
significantly higher than on screen.
If you are wondering how I got these results, the answer is that I
have a lot of free time on my hands, and I spent all day counting
pixels. Boy am I tired!
For Eric and anyone else that may be interested, after reading his
message I wrote a program that prints all characters to the screen for
a given font and then counts the individual pixels used (is this the
most efficient method?).
I'll email the program to those interested (I'm still not sure what
this information is needed for or who would need it; please enlighten
me). If you want a copy, mail me at auth...@internetmci.com. By
requesting a copy you are accepting the terms of the disclaimer below.
About the program:
1) Sorry! Runs only under Windows 95/NT (tested only under Windows
95).
2) Allows you to choose any font, attributes, and size. If the size
is too big for the space provided, the size will be reduced (the
output will indicate the actual size used). Example: for Times New
Roman, 156 was the maximum size I could use.
3) The program only counts uppercase letters.
4) The output is ordered alphabetically. If you want to sort by
coverage as I did above, the easiest way is to select the output, copy
(ctrl-c), paste into a capable word processor, and sort (in MS Word 6,
select the text to be sorted, choose menu "Table / Sort Text").
5) If you have a slow computer, be patient; all pixels for every
letter are counted! On my P133, the program is very fast.
6) Prog. size is about 180K or so. This is my first program under
Borland Delphi (I give Delphi an A.. I give my old Microsoft Visual
Basic 4.0 a C+; I haven't seen VB5 yet).
Disclaimer:
This software is provided free of charge (public domain). The program
is provided as-is and without any warranty. Any damage caused is not
my responsibility. I also do not claim that the output is accurate
(if you would like to inspect the <undocumented> source code, just ask
me for it).
Eric Angelini
user <us...@pixi.com> a écrit dans l'article <335fc...@news.pixi.com>...
> Jacob Langford <lang...@uiuc.edu> wrote:
> Please satisfy my curiosity: what do you need this information for?
I'm part of a group of OULIPO fans (OULIPO means OUvroir de LIttérature
POtentielle - a kind of word puzzles fans which write texts with
constraints - example: write a text using only 'flat' letters like
azeruosmwxcvn (excluding thus typqfghjkl) etc.-
I've written a small text using only the 13 'less-ink' letters and another
one using the 13 'most-ink' letters:
when you look at them you visually *see* how the first one is 'light' and
the second 'heavy'... though they tell the same story and use the same
number of characters...
I'll download yr pgm in a few hours, my son wants *right now* to go with me
for a walk...
Tx for the great job of classification an computing!
I'm back asap!
Eric Angelini Brussels