Eduardo,

I 'll comment asap.

-smc

It's nice, Eduardo. I suppose that you're aware of
Peter Verheul's Newberlin, that also uses straight lines
only. But I find your Maest more interesting. I agree
that there's a crispness in the lc that the UC is
lacking somewhat. Some of its characters could be a bit
narrower, the bowls of P and R a little smaller.

I'm not sure whether the sort of variation in the spacing
rhythm that you have here was intentional or not. For
instance: how letters seem to coalesce in 'qua', only to
disperse again in 'scele' (and especially in 'ss' ;). This
"elastic" behaviour may bother some people (not me).

One possible problem that I see is that in smaller settings
the 'n' could be mistaken for an 'u'. Also, there may be
a problem in certain cases with the gap created by
the 'c' in some combinations (like ci, ct). But generally
it will work, I think.

abra

Eduardo

i like it a loti agree with rodolfo about the n/u problem, but my main critic is the bottom of your lc "c" and "e", i think it should be more extended, at small sizes it looks to weak to me.

for me maest its much better than newberlin, here the beautifull contradiction betwen straight lines and the cursive type is more evident.

CG

Eduardo,

You should really check out Preissig
- much more inspired than Newberlin.
When I first saw Maest I was reminded of this.

I took a look at the PDF on your site (very nice
by the way) and I rather like Maest at smaller
sizes. Better in fact.

The "f" seems a little spindly to me and I think
maybe the dots on the "i" and the "j" are a little
to high.

While I see what everyone has said about the
"n" I don't think it is that big a deal. It does not
bother me.

But I totally agree about extending characters.

Otherwise it is looking good to me. Maest is
more expressive than Newberlin.

-smc

Nice. I enjoy the irregularity and think it too many of the characters conformed to the same rules it wouldn't have the same flavor at all. If anything I would think it would be good to have alternates for the characters so that the user could keep this irregularity more consistant. For instance in your paragraph it seems that some of the words have the first few character aligned on the baseline and then the last few fall off. It might be nicer if there really was a consistant up and down or at least irregualar action/flow of line.

thanks for your comments!

>Peter Verheul's Newberlin

i think i had seen this one long time ago, but i didn't
actually know it until you mention it.


>there's a crispness in the lc that the UC is
lacking somewhat.

yes, what do others think about it? i think this is my
major problem, since an all-caps settings looks just
ugly. check the pdf sean mentioned here.


>Some of its characters could be a bit narrower,
the bowls of P and R a little smaller.

i've already tried some variations in UC, and i think
they should be wide. i'll search the old caps in my
files to show you.


>variation in the spacing rhythm

this is just bad spacing. it wasn't supposed to be so
irregular, so i'm going to fix it. ;)


>n/u problem

i'm with sean: this doesn't bother me either.


>bottom of your lc "c" and "e"

maybe it can be extended just a little. this can be
improved with a good spacing, too, i think.


>Preissig

this one i knew. great typeface.


>rather like Maest at smaller sizes. Better in fact.

i totally agree, but it seems quite contradictory,
doesn't it?


>The "f" seems a little spindly

this is very intentional. :-)


>the dots on the "i" and the "j" are a little to high.

i'll fix this. the dots are a little light too.


>if there really was a consistant up and down or
at least irregualar action/flow of line.

good point tiffany. i think i'm just going to make the
characters' baselines more regular.

>n/u problem
i'm with sean: this doesn't bother me either.

I'm not sure whether it is a matter of you being bothered
or a matter of testing actual readers to see if there
may be a problem here.(*) (I'm using the conditional
since the first time around.) Type, like any kind of
design, demands testing. Again, depending on what one
expects from one's typeface (no need to test anything
when it's only a vehicle for self-expression ;). Also:
one may well choose sometimes to enhance a stylistic
aspect in [slight] detraction from a functional one. In
a face like Maest - largely driven by a formal principle
- this would be a perfectly valid decision, btw.


>variation in the spacing rhythm

this is just bad spacing. it wasn't supposed to be so
irregular, so i'm going to fix it. ;)

>if there really was a consistant up and down or
at least irregualar action/flow of line.

good point tiffany. i think i'm just going to make the
characters' baselines more regular

In the s there is certainly a spacing problem. But in
both the aspects above I'd be careful not to regularize
it too much. It could lose some of its charm. ;)

abra

> Type, like any kind of design, demands testing.

Right on. It's a craft.

> Niufa doesn't exist

But if it overpowered the "Ninfa" reality in your mind, there must be something going on (especially if "Ninfa" is a word/name). Maybe it contains a familiar letter-string that made it into a powerful [what's called] pseudo-word.

hhp

> Type, like any kind of design, demands testing.
Right on. It's a craft.

Well, let me disagree. Let's say, as an exercise, that
there are three kinds of people in this thing: artists,
craftsmen and designers. I prefer to think that I'm in
the third group. You, I know, like to see yourself in
the second one.

Maybe it contains a familiar letter-string that made
it into a powerful [what's called] pseudo-word.

I guess it's simpler than that: the n may be too much
like an u. I'd call it (sorry, Eduardo! ;) a matter of

> especially if "Ninfa" is a word/name

It means "Nymph" in Portuguese.

Also:
>...testing actual readers...

Agreed. The key word here is readers.
Not designers or craftsmen.

>one may well choose sometimes to
enhance a stylistic aspect in [slight] detraction
from a functional one. In a face like Maest -
largely driven by a formal principle - this would
be a perfectly valid decision,

Again, in this case, "n" I think passes.

>no need to test anything when it's only a
vehicle for self-expression

Maest is more than this.

>I guess it's simpler than that: the n may
be too much like an u.

I had no problem reading the sample.

-smc

>...testing actual readers...

Agreed. The key word here is readers.
Not designers or craftsmen.

Yes! Peer criticism and "opinions" can never
pass for testing! (Call this equation 1.)

>one may well choose sometimes to
enhance a stylistic aspect in [slight] detraction
from a functional one. In a face like Maest -
largely driven by a formal principle - this would
be a perfectly valid decision,

Again, in this case, "n" I think passes.

That's what I said: "perfectly valid".

>no need to test anything when it's only a
vehicle for self-expression

Maest is more than this.

Certainly! And I hope that nothing in what I
wrote would imply that I think it isn't.

>I guess it's simpler than that: the n may
be too much like an u.

I had no problem reading the sample.

Well, now I have to ask you to refer back to
equation 1 above. Would you accept that
the opinion of one reader, who, not by chance,
is also a type designer, can pass for testing?
I don't think so. One fundamental rule of testing,
btw (and not that it matters here), is that you
don't ask the subject's opinion. You observe
his or her behaviour in a number of controlled
situations. But this is idle talk, really. Maest
is very nice as it is, irregular spacing and
baselines and all.

> The key word here is readers.

Well, agreed. But readers (which I prefer to call "users", since there's more to communication than reading) don't design their own stuff. Designers do that for them. And the point is that a real designer is not an artist because an artist's main motivation is self-expression. That doesn't mean there's no art in design - not at all. In fact there is unavoidably art in anything a human does - which is a good thing. The main difference between Art and Craft is motivation.

Testing is exactly something that a designer needs to do but an artists will not.

hhp

Re: arts, crafts, design...

i think type design is a craft too. i
know this typeface is a little too
"artsy", but it is far away from pure
self-expression exercise. btw, it's
more like self-limitation exercise. ;)


Re: testing actual readers...

does anyone actually do that?
i really want to know if you have
any experience doing it.

if anyone is still interested, here are the old caps (01=oldest/04=today).

the pdf version:

old caps post2.pdf (18.1 k)

Eduardo:

About your caps: haven't looked at the PDF
yet, but from your gif I think I rather prefer
the narrower version. That doesn't mean it
couldn't be a tad less narrow than in 01.

Re: testing actual readers...
does anyone actually do that?
i really want to know if you have
any experience doing it.

I did it when I was developing Houaiss but
not systematically and not in the scale I would
have liked - there was no budget for that.

Considering how type design is conducted
nowadays (generally small firms, mostly
one-person firms) I don't think formal testing
is really something that is done. I'm not sure
about larger and more organized firms like
FontBureau. Going back to the typology above,
there is a craftsman's spirit in the thing -
and craftsmen don't conduct tests, they rely
on their ingrained experience (what passes for
intuition) for decisions.

But one could argue that testing should be
as important as it is, for instance, in
automobile design (it was Otl Aicher, I think,
who said that designing a typeface is like
designing a car).

I'd like to go back for a moment to the n/u
question and the small montage I made with
your type, that was not commented:



What's interesting here is that the same glyph
can stand for two different characters and
remain perfectly legible, without confusion,
depending only on context and lexicon (this is
the principle that was taken to its painful limit
in Tobias Frere-Jones's Sum of the Parts).

This means that, by itself, it's not unambiguously
decipherable. Is this a problem? In most typefaces
it would be. You could, just for fun, conduct a
test with this glyph. For instance: are your
Japanese grandparents still alive? Do they read
Portuguese? (badly, I suppose?) You could prepare
some sample texts, including "trap" words where
this glyph could be read either way and see what
they make of them. Do the same thing with young,
competent readers (aren't these two adjectives
contradictory? ;) and see what you get. Have fun.

This is exactly the difference between legibility and readability in a font.

hhp

thanks rodolfo and hrant. sorry for not replying it sooner.


>About your caps: haven't looked at the PDF
>yet, but from your gif I think I rather prefer
>the narrower version. That doesn't mean it
>couldn't be a tad less narrow than in 01.

i'll try new variations...


>I did it when I was developing Houaiss but
>not systematically and not in the scale I would
>have liked - there was no budget for that.

what do you mean "not systematically"?
if you had no budget in that huge project, i
wonder how i was supposed to do it for my small
project. ;)


>Considering how type design is conducted
>nowadays (generally small firms, mostly
>one-person firms) I don't think formal testing
>is really something that is done. I'm not sure
>about larger and more organized firms like
>FontBureau.

this is just what i expected.


>Going back to the typology above,
>there is a craftsman's spirit in the thing -
>and craftsmen don't conduct tests, they rely
>on their ingrained experience (what passes for
>intuition) for decisions.

there are intuitive and logical solutions in maest,
like any other typeface.


>I'd like to go back for a moment to the n/u
>question and the small montage I made with
>your type, that was not commented:

i didn't comment that because i needed to
think more about it.
but i liked the montage. ;)


>What's interesting here is that the same glyph
>can stand for two different characters and
>remain perfectly legible, without confusion,
>depending only on context and lexicon (this is
>the principle that was taken to its painful limit
>in Tobias Frere-Jones's Sum of the Parts).
>This means that, by itself, it's not unambiguously
>decipherable. Is this a problem? In most typefaces
>it would be.
>
>This is exactly the difference between legibility and readability in a font.

i think hrant's words are better: instead of "confusion x
function", let's call it "readability x legibility". in
rodolfo's montage, both words are readable. but
"function", where [n=flipped u], is more legible:
anyone can tell which one is n and which one is u. in
the word "confusion", where [u=n], we have
ambiguity, but we are still able to read and
understand it.

in the actual font, n differs from u, but i guess for
some people it is too close to their own "self-
conscious" image of an u. that's why rodolfo pointed
it in "Ninfa". in the other hand, as we know, reading
is not a fully "conscious" task, and that's why i think
the original n/u works. it will work even better in
small (text) sizes, as sean commented before,
probably because reading it requires less visual
"consciousness" than reading in bigger (display)
sizes.
(is this making any sense? ;)


>You could, just for fun, conduct a
>test with this glyph. For instance: are your
>Japanese grandparents still alive? Do they read
>Portuguese? (badly, I suppose?). You could prepare
>some sample texts, including "trap" words where
>this glyph could be read either way and see what
>they make of them. Do the same thing with young,
>competent readers (aren't these two adjectives
>contradictory? ;) and see what you get. Have fun.

i think the "young competent readers" test was already
done. i showed it (roughly the same sample i posted
here) to a few non-designer colleagues, and nobody
complained about anything. actually, they seemed
like don't care about it. i'll try showing it to older
people (not my grandparents; they barely know
portuguese).

> reading is not a fully "conscious" task, and that's why i think the original n/u works.

Reading is indeed largely subconscious, but in some contexts you will have problems. Like "gnu" would usually be read as "gun" - especially since the latter is much more common, and both are in the same class (noun).

> nobody complained about anything

But that's not enough, exactly because reading is largely subconscious!

hhp

reading is not a fully "conscious" task, and that's
why i think the original n/u works. it will work even
better in small (text) sizes, as sean commented before,
probably because reading it requires less visual
"consciousness" than reading in bigger (display)
sizes. (is this making any sense? ;)

Yes - if we disregard the fact that you're almost sounding
like Hrant... ;)


i think the "young competent readers" test was already
done. i showed it (roughly the same sample i posted here)

Sorry, but then you have tested nothing. A Latin sample text
or a series of isolated words aren't testing anything. (And
"opinions" or the lack of "complaints" are useless.) You have
to have a true text sample in the subjects's language,
something that she or he would actually read. Running the
risk of sounding like Hrant myself, something that could be
read "immersively" (though that will hardly be the case in a
test).

As I think I have already proven, the n glyph can also
stand for an u, so you'd have to have words that explore
this ambiguity, for the test to have any interest and be
fun. That's were the catch is. BTW, don't think of Houaiss
as an example. Real-world projects, with tight deadlines,
are very often less open to this sort of experiment. Think
of it as testing yourself. ;)

Oops, Hrant, I hadn't seen that you were there! ;)

Perhaps try typesetting various words with
the "n/u" comparison to see if ambiguity
will truly be a problem.

Try words such as:

annual
annular
gnu / gun
nuclear / unclear
nonnuclear
nonunified
noun
nuisance
nunu
unnumerical
unannounceable

If it still proves to be a problem, I imagine
the problem can be easily corrected
by slightly openiing the "curves" in the "u"
and "n", and adjusting how the broken "curve"
joins to the stem.

I really like this font, and I'd love to know
when it's finished and released.
This is one I'd definitely use.

David

> I hadn't seen that you were there!

It's OK, Hrant. I think your replies to Hrant were right on.

> Try words such as:

And don't forget proper names (especially "foreign" ones). When you don't know how to pronounce a name, reading comes to a complete and long stop. Like a botched Formula-1 pitstop (I hope one happens to Ferrari tomorrow :-).

hhp

Try words such as:

In fact, I think these words aren't so useful to
what I had in mind. What would be more interesting,
in my opinion, would be a test where you had real words
that could be read either way (as, in Portuguese,
"canto" and "cauto" - only, this isn't a good example
for canto is so much more common than cauto that it
will always take precedence). This kind of thing.
Of course, using the n-glyph.

> this isn't a good example for canto is so much more common than cauto

But that's exactly what would make setting "cauto" in a phrase a good test (even if it's semantically different than "canto").

hhp

OK, but then the n-word, not the u-word, should be
the rarer one.

>>>Reading is indeed largely subconscious, but in some
>>>contexts you will have problems. Like "gnu" would usually
>>>be read as "gun" - especially since the latter is much more
>>>common, and both are in the same class (noun).
>>
>>Try words such as:
>
>In fact, I think these words aren't so useful to
>what I had in mind. What would be more interesting,
>in my opinion, would be a test where you had real words
>that could be read either way.

as rodolfo pointed, mis-reading may occur in words where "n"
could stand for "u". when we have both letters, in the same
word, there's no ambiguity. "n" can be mis-read as an "u",
but "u" can't stand for "n". look:




>>this isn't a good example for canto is so much more
>>common than cauto
>
>But that's exactly what would make setting "cauto" in a
>phrase a good test (even if it's semantically different than
>"canto").

the word "cauto" is very rare, so many people don't even
know it is an actual word (i didn't know it).


***

>I really like this font, and I'd love to know
>when it's finished and released.
>This is one I'd definitely use.

thanks, david! :-)
i let you know i finish this.

Predictably, I could read all the running text in your sample just fine, but the problem remains (like in "gnu"). Why not just bring up the arch of the "n" like you're lowered the one in the "u"?

--

BTW, was that a great race or what? Tifosi, eat my finial.

hhp

>Why not just bring up the arch of the "n"
like you're lowered the one in the "u"?

i know. rodolfo

I don't know - it used to be a lot less commercial. Plus when I started watching Ferrari wasn't polluting* it. Hilfiger as a sponsor indeed... Anyway, I'm attracted to dangerous sports. Like alpine skiing really turns me on.

* Actually, they were polluting the tracks: they would consistently explode mid-race and leave Agip all over...

But actually the moment I was hooked was during the Mexico race of '86. Senna put harder compound tires on the left than on the right, because the track was mostly right turns... And he won (of course). Senna was different. That Donnington wet race in '93 was just alien. Anyway, the extreme technical aspect gets me. Which is why I never watch Indy racing (or whatever it's called now): it's 19th century tug-boat racing.

Plus, once you go to an actual race (my first was in '90) it becomes just out of this world. A friend had brought a camcorder, and it stopped working from the noise vibrations.

I have autographs of Prost, Hakkinen and Wendlinger (don't ask), and I've come within 3 yards of Senna, after he won in Monaco in '93 - that was a miraculous race. Aaah, the days.

hhp

i see you really enjoy it.
well, next race is here in s

Eduardo,

With all this discussiou oops... discussion over the "n".
No one commented on your second PDF - the one
with other versions or stages that your design
has seen.

I think the caps in 2 are great! For me, they
really add to the rhythm. Look at the "M" in
"Midas" for instance. In 2 it flows really nice.
While in 4, I sort of pause when I reach it. I am
not saying 4 is is not pretty I just think 2 might
be a bit more consistent in expressing the
personality of Maest. But it really depends on what
your goals are. I think you mentioned that you were
really trying to be disciplined in your design.
I say loosen up a bit and let the type face design
itself just a little bit more.

You could take 2 even further and create
swash characters as an addition. That would
be really cool! In addition this could help to
separate it even further from the designs that
where mentioned earlier.

But then again 2 will raise even more legibility issues.
Do the "I", "J", and "T" look too much alike? And then you
have the whole "n" story all over again. What a story it told.

I think that UC "U" would look cool and be more calligraphic
with a stem coming down the on the right. Like a LC.

Thanks for posting your second PDF. I really
like it.

-smc

thank you, sean.

I think the "M" in 02 is very nice too, but the hardest thing is
to make the other Uc characters coherent with it. Frankly, I
don't have any idea of how to do it. Btw, that's the same
problem in 04.
For example, "A" and "B" just don't get along. It seems like "A" is
one font, and "B" is another. I think the same thing happens with
other "round" characters, as "B", "D", "P", "R" and "S". I just need
to spend more time on this.

But I liked your idea of making swash characters. I'll see what
I can do. ;)