Start with a pencil.
Scan your drawings.
Auto trace with Adobe Streamline or Fontscan.
Clean up in Illustrator or Freehand.
Copy and paste into Fontlab.
Work on shapes some more.
Space.
Kern.
Read the manual.
Good luck.

-smc

O.K. lots of people are going to disagree with me now, but anyway....

I think a fundamental understanding of writing is needed before one can begin to design typefaces. Writing, whether with a calligraphic pen or other tools help establishing a kinesthetic understanding, or, a feel for proportions and flow. Writing with a broad-edge pen will help you to appreciate the principles behind thick-thin transitions, which is essential for designing serifed text faces. A good writing manual would certainly help. Edward Johnston's Writing and illuminating, and lettering might be a good place to start, or Gerrit Noordzij's Stoke of the pen, or other good calligraphy books. After getting to grips with the principles of writing, lettering could follow. Lettering could be seen as a liberation from writing, to free oneself from the bounds of the writing instrument, but still retaining a strong tie with the fundamental principles of writing. Experimentations and explorations could then follow.

FontLab is a very sophisticated piece of software for making cutting-edge digital faces, but as a design tool it's a bit counter-intuitive. I prefer working out ideas on paper first, scan it and take it from there.

Of course I haven't talked about spacing and kerning yet! I have no idea... I just do it by eye. But I think I did learn a lot about spacing from doing calligraphy (again, I think it's a kinesthetic thing) and rubbing down Letraset! Oh yeah, look at Walter Tracy's Letters of Credit. Also Twentieth century type designers by Sebastian Carter will give you a lot of insights about how different type designers work and give you lots of inspiration!

Hope this helps. Gook luck!

well i think the best way to start is to go do some researching to see what kind of typefaces are out there.. and then start creating something that fit with u.

My recommendation is to start with a Serif rather than a Sans. To understand better the shapes and counters forms. Its not because they (Sans) lacking of contrast that they are more easy to design. You will learn more with Serif fonts.

Keith, one day I will write a treatise as to exactly why chirography is not only unnecessary, but in fact detrimental. Liberate yourself now - don't wait for lettering - lettering in fact shares chirography's incapacity to elaborate on typographic spacing.

hhp

All of these are great suggestions Henriko. I would have to add that in my case, the lettering and calligraphy classes that I have been taking have added to my creativity and understanding of some letter forms. I am happy I have taken them.

Adding to my first post I would like to state that the more finished your drawings are from the start are the easier your life will be when you import them to a digital format. ( I should note on the contrary however that many designers never touch pencils. )

Another good book to take a look at would be Counterpunch: Making Type in the Sixteenth Century, Designing Typefaces Now.

-smc

xxx

i remember reading somewhere that chester (thirstype)
started some faces based on sketches in his doodle-book,
and some were started and finished all on the computer.
i think (and consider the source here) that it depends
on the face, it's intentions, vibe, etc. i think the best
place to start is trying to replicate what you see in
your head. easier said than done though.

Adding to my first post I would like to state that the more finished your drawings are from the start are the easier your life will be when you import them to a digital format.

This depends on how you intend to 'import' them to a digital format. I gave up on scanners several years ago, when I discovered that even quite expensive ones introduce distortions by compressing or expanding in one direction. These days, if I draw on paper at all, I only make rough drawings to get the shape down: then I prop the drawing beside my screen and use it as a visual guide to create the shape on the computer. It was only after I had been doing this for a couple of years that I read a comment by Hermann Zapf regarding the kind of drawings he prepared for a manual punchcutter, as contrasted with those prepared for industrial production. The drawings for industrial production needed to be very clean and precise, because the individual operating the pantograph, camera, Ikarus tablet or scanner (depending on the era) was neither qualified nor expected to interpret the design, only to reproduce it as exactly as possible. A good punchcutter, however, is an expert interpreter: able to take a designer's quite rough drawing and create a workable typeface from it, glyph by glyph, on the end of little pieces of steel. As a result, Zapf says his drawings for the brilliant punchcutter August Rosenberger, with whom he worked for many years, were much rougher than he needed to produce for mechanical production. I believe the same is true in digital type design, especially when one is making one's own outlines: it is not necessary to produce perfect drawings if one develops the ability to interpret rough drawings directly with the computer tools.

John, I design most of everything directly on screen generally without any sources on paper.

But, my point is sometimes, some style can't be achieved directly on screen, you will lost too much time. I refer to "Script" forms who imitate writing and so on, generally all fancy lettering stuff. Just try to draw directly on screen something like Zapfino, ex Ponto, Sanvito, Caflish Script...

The best, is to know perfectly the limitations of the tools you use (because you know them), and use the most adapted to your whishes rather than trying to do something with a tool who make your life more complicate.

To comeback to non experienced designers. Generally is more easy to separate the learning of B

> it is not necessary to produce perfect drawings

That's a very good point. In fact it's a waste of time. For example, when you're drawing an italic, there's no use normalizing the angles on paper.

On the other hand, certain types of designs (like highly irregular ones) might benefit from "finished" drawings.

I use paper myself, but really only because drawing software isn't very good (yet).

hhp

Sorry also for strange display of the top of the page in filled in red on Windows, when its ok on Mac. If someone can learn me how to correct this!...

i'm on Windows, and changed property "bgColor=red" to "bgColor=white" on both <table> commands. it seems to be working now:

sorry off topic. Thanks Eduardo, but I have already tried that, and I lost the red lines that I like! so...

All of the above seems true. There are good arguments for all these points. Everyone finds their own way that works best for them.

I think, as a beginner, drawing has more freedom and can help to identify clear objectives about what one may want out their design. Experienced designers might be able to draw directly on the computer because they know where points should land after having placed thousands of them. The beginner does not have this in their arsenal yet. So I still see scanning as a good way to start even if it is necessary to interpret these to accommodate for scanner distortion.

Is is necessary to draw every character perfectly before moving to the computer? No. But but defining clear characteristics of your design does seem important to me. After these characteristics, such as a serif for example, are discovered, defined, whatever it is easy to copy and paste these in the computer from glyph to glyph. But do not do this for everything. You don't want your "m" to be two "n"s stuck together.

Henriko, however you start, start with the words "hamburefonts" or "handgloves".

Start with lowercase then move to the uppercase.

-smc