Phantom Modulation. Am I mad?

Aaron Sittig's picture

I'm working on a sans serif design, not quite mono-lineal, some humanist
structure but no humanist modulation, and I find myself constantly
questioning if I have drawn my shapes correctly. When setting sample
words, some parts of strokes appear thicker and some thinner. Parts
that would normally be made thinner by the shape of the pen, the upper
left and lower right of the 'o', right stroke of 'v', appear thick. Where I
might expect a thickening of the stroke, lower left and upper right of
the 'o', left stroke of the 'v', I see instead a thinning.

Am I mad or is this a common experience? I imagine that my mind
'expects' the normal pen based modulation that I've been accustomed
to working with in my past few projects and is reacting against the
mono-linearity of this new design. I'm finding it very distracting
and I don't know what to do about it except to keep measuring all
my strokes to confirm that they are in fact the same width.

Ideas? Similar experiences?

Aaron Sittig's picture

Having a hell of a time uploading this image on my slow connection.

Aaron Sittig's picture

I give up

http://www.xtypa.com/ext/typophile/phantommodulation.gif

aquatoad's picture

Hi Aaron. You're definitly mad. So am I :-)

Are you designing on screen? If so, do printed letterforms have the same issue? The trouble I've had designing onscreen is by the time you zoom out to get a sense of the whole letter (or even a word) the anti-aliasing starts messing with my mind as you've described. While I still design on screen, there is an element of trust in your measurements untill you pull a print..

In your sample, the only one that looks off is the v. If anything the thick should be on the left.

Good luck.
Randy

Miss Tiffany's picture

Maybe this will help. :-)

Aaron Sittig's picture

> Are you designing on screen?

Yes

> If so, do printed letterforms have the same issue?

Not that I've noticed, so if so, it's not as strong. I'll give it a
look when I get home.

> In your sample, the only one that looks off is the v.
> If anything the thick should be on the left.

I first noticed the effect with the v; it seems strongest there.
According to my theory, the left stroke should appear thin.
After repeatedly seeing 'v' with conventional modulation,
thick on the left, thin on the right, by comparison a mono-
lineal v will seem too thin on the left and too
thick
on the right compared to what your brain has
become used to.

When I resumed work on this design last night, after
reading around 50 pages of Adobe Garamond, the effect
was very evident. Looking at it today, it's not quite as clear
in the round characters, still somewhat visible in the v.
Maybe that's the lesson, don't try to be original after
extended reading or your gut feeling will tell you to make
your design look like the typeface you've just been
immersed in.

John Hudson's picture

You're not mad and this is quite common. I find that even the slightest 'correction' for this, i.e. the slightest modulation, will compensate to produce the appearance of an unmodulated stroke. The likeliest explanation for the phenomenon (and ghost modulation is quite a good term for it) is that we are so accustomed to the 'normative' forms of our letters being modulated, and in a particular way, that our minds reject the absence of modulation and see, instead, a reverse modulation. The effect is most obvious in letters with 'mirrored' forms, e.g. A O V o v; in Tiffany's illustration above the effect is much less troubling in the d. I suppose if one was exposed only to totally unmodulated letterforms for an extended period of time, one would eventually be 'cured' of this phenomenon, but how likely is that? It is sensible to introduce very minor modulation to counter the perception.

hrant's picture

When I was reading Smeijers's "Couterpunch", in the chapter about adding something to the lc "j" to make it sit on the baseline*, my mind started thinking about "reform", and it immediately hit me that the Latin lc "a" is facing the "wrong" way! I now realize that it's not so simple, but the power of suggestion is indeed powerful.

* BTW, I don't know if he realizes it, but that sort of thing is totally contrary to his otherwise highly conservative approach to type design!

> we are so accustomed to the 'normative' forms of our letters being modulated

I think that's part of it (especially for a type designer), but there are [probably] optical issues as well - just like a perfect circle/square *appears* imperfect.

hhp

Aaron Sittig's picture

> I think that's part of it (especially for a type designer), but
> there are [probably] optical issues as well

I think it's fair to call this an optical issue, not just a matter
of familiarity. It's likely similar to color fatigue, where the
world looks blue after you've spent a couple minutes
wearing yellow sunglasses. But I also think that since we
spend so much more time reading modulated text than we
do wearing colored sunglasses, the effect is more
permanent.

> I find that even the slightest 'correction' for this, i.e. the
> slightest modulation, will compensate to produce the
> appearance of an unmodulated stroke.

And it seems you're not alone. Myriad compensates for this
effect in both the o and v. Adobe's Frutiger compensates
only in the v, but does so quite dramatically. Agfa/
Monotype Gill Sans compensates in o and v. Linotype
Helvetica Neue compensates only in v. Neufville Futura has
no compensation at all.

So it appears that I should add some optical compensation
to my design. Seems to be standard best practice today for
the most common digital sans fonts.

aquatoad's picture

Aaron, care to show more? How far along are
you? The attention to detail above shows good
promise. Keep us posted.

Randy

Aaron Sittig's picture

If I can find the time this weekend to finish up the upper
case, I'll post a pdf over in Critique.

Syndicate content Syndicate content