Arial versus Helvetica(swiss-miss.com)
swiss-miss.com
Arial versus Helvetica
http://www.swiss-miss.com/2009/09/arial-versus-helvetica.html
54 comments
Very nice, it would be interesting if you could do a Microsoft Sans Serif - Helvetica comparison as well as those two fonts are almost identical, with only very very subtle differences.
Why do you compare Consolas with others ? It's a monospaced typeface, it should be compared with Courier New/Monaco/Andale Mono etc ... Or Courier vs Courier New ?
Compare the popular typefaces from different platforms (Win/Mac/Lin) would also be very interesting
Compare the popular typefaces from different platforms (Win/Mac/Lin) would also be very interesting
How about Time New Roman vs. Georgia? It's always puzzled me that they look much different on screen, but fairly similar when printed.
That's a good, subtle one: http://cronus.ws/~mta/fonts/GeorgiaTimesNewRoman.html
OK, so I guess the slight differences are exaggerated by the font hinting. But the printer, being much higher resolution, can show the similar font shapes more clearly. Cool.
I'd be curious to see how you did this. Would you be willing to release the source code?
No problem:
http://cronus.ws/~mta/fonts/FontCompareSource.nb
And for those without Mathematica: http://cronus.ws/~mta/fonts/FontCompareSource.pdf
And for those without Mathematica: http://cronus.ws/~mta/fonts/FontCompareSource.pdf
The easiest way to tell is that Helvetica is obsessive about straight horizontal edges, while Arial tends to use angled edges rather unsystematically.
Except, for some reason, the capital R in Helvetica, which is a very curvy letter.
But the R only curves so that its leg meets the base at a logical angle. Arial's straight leg looks straighter, but if you isolate it as a piece itself then that straight part cuts off at a diagonal angle to maintain the flat bottom.
Oh, wow, interesting; I hadn't thought about it that way.
(The R does still have a little serif though, which always bothers me.)
(The R does still have a little serif though, which always bothers me.)
It's easier, you pay for Helvetica while Arial is free. Also no graphic designer of at least basic skills uses Arial. Arial is only used on the Web.
http://www.helveticafilm.com/
About the film: Helvetica is a feature-length independent film about typography, graphic design and global visual culture. It looks at the proliferation of one typeface (which celebrated its 50th birthday in 2007) as part of a larger conversation about the way type affects our lives. The film is an exploration of urban spaces in major cities and the type that inhabits them, and a fluid discussion with renowned designers about their work, the creative process, and the choices and aesthetics behind their use of type.
About the film: Helvetica is a feature-length independent film about typography, graphic design and global visual culture. It looks at the proliferation of one typeface (which celebrated its 50th birthday in 2007) as part of a larger conversation about the way type affects our lives. The film is an exploration of urban spaces in major cities and the type that inhabits them, and a fluid discussion with renowned designers about their work, the creative process, and the choices and aesthetics behind their use of type.
By total coincidence I watched that movie yesterday. I enjoyed it, though it's mostly just a bunch of people's opinion about Helvetica/modernism/post-modernism. Doesn't good much deeper than that in typeface design.
Can't remember where I heard about it (possibly MetaFilter, or here?).
Can't remember where I heard about it (possibly MetaFilter, or here?).
MeFi, it was posted there a couple of days ago including the film link. My favorite comment from the the resulting thread was 'Impact or go home, font dorks', which I have been unable to get out of my head since.
Does anyone know of reputable research and blind tests that back up the common designer sentiment against Arial?
The reaction of many designers seems to smell of elitism and subjectivism more than anything to me, but I'll be happily proven wrong.
Are designers commonly aware of why Arial exists. I presume many designers must understand the very strong business & licensing reasons for Microsoft to have developed this typeface.
Do many designers also understand the distinctly different approach to pixel alignment in display faces, which some people prefer while others prefer Apple's approach (I subjectively prefer Microsoft's approach).
It's easy to pile onto the I hate Microsoft and everything they touch bandwagon, but how much unbiased and objective research has been applied to the question of qualitative value to the target audience? Is an old typeface only good if it's designed by the Swiss?
Something is stinky here. There are so many subjective factors at play while the language surrounding debate rarely recognizes those factors.
[Update: As noted by DrJokepu below, Microsoft didn't develop the font originally. I think the underlying licensing and business considerations remain a similar discussion.]
The reaction of many designers seems to smell of elitism and subjectivism more than anything to me, but I'll be happily proven wrong.
Are designers commonly aware of why Arial exists. I presume many designers must understand the very strong business & licensing reasons for Microsoft to have developed this typeface.
Do many designers also understand the distinctly different approach to pixel alignment in display faces, which some people prefer while others prefer Apple's approach (I subjectively prefer Microsoft's approach).
It's easy to pile onto the I hate Microsoft and everything they touch bandwagon, but how much unbiased and objective research has been applied to the question of qualitative value to the target audience? Is an old typeface only good if it's designed by the Swiss?
Something is stinky here. There are so many subjective factors at play while the language surrounding debate rarely recognizes those factors.
[Update: As noted by DrJokepu below, Microsoft didn't develop the font originally. I think the underlying licensing and business considerations remain a similar discussion.]
Both fonts fulfill the purpose for which they were designed.
Helvetica was built to live in a gridded world: the horizons and verticals of Swiss modernity (look again at R and see how space-consuming it is).
Arial was built to look like Grotesque 215, which is what swiss designers were using before Helvetica. Arial is supposed to be more human (read: humanist), and is for on-screen and on-paper --- not signage.
Designers prefer Helvetica because it is built to be used by designers; any combination of letters will pack neatly in space, both because of the letterforms themselves and the superior kerning.
Arial is meant to be used for less important, less permanent documents. It is more casual, and setting a whole page of text in it does not seem cold or callous.
Helvetica was built to live in a gridded world: the horizons and verticals of Swiss modernity (look again at R and see how space-consuming it is).
Arial was built to look like Grotesque 215, which is what swiss designers were using before Helvetica. Arial is supposed to be more human (read: humanist), and is for on-screen and on-paper --- not signage.
Designers prefer Helvetica because it is built to be used by designers; any combination of letters will pack neatly in space, both because of the letterforms themselves and the superior kerning.
Arial is meant to be used for less important, less permanent documents. It is more casual, and setting a whole page of text in it does not seem cold or callous.
All of which avoids the original question: "Does anyone know of reputable research and blind tests that back up the common designer sentiment against Arial?"
I don't think it's so much a matter of research or blind tests in this case. Like a regular user may not notice that a website loads .1 second faster because somebody optimized the database queries, a regular viewer will not be able to tell the difference between Arial and Helvetica. Designers, who are trained to be sensitive to those details will have an automatic response and be able to pinpoint what the reason is.
But you can test that shaving that 100ms off your website latency has a real impact on the user, because it can show up in your metrics.
Alternatively, you could do the same thing with Arial and Helvetica. Show one to half of your visitors and the other to the rest then compare conversion results between the two sets.
Mu.
There can be no such thing. Both fonts accomplish a purpose. Many like one because they feel it is easier to design with. Any reputable designer will instantly recognize both. They are two of the most used fonts in the world. There is no way to perform a blind test.
The closest thing to reputable researchers in design are designers. They spend their lives experimenting with fonts, and most prefer to use the one to the other. Some of them are following popular opinion mindlessly. Many aren't. Until some metric can be made for good design (fat chance. Also, note that readability is not the only important piece of this puzzle) no journal-worthy piece of science is going to come forward.
There can be no such thing. Both fonts accomplish a purpose. Many like one because they feel it is easier to design with. Any reputable designer will instantly recognize both. They are two of the most used fonts in the world. There is no way to perform a blind test.
The closest thing to reputable researchers in design are designers. They spend their lives experimenting with fonts, and most prefer to use the one to the other. Some of them are following popular opinion mindlessly. Many aren't. Until some metric can be made for good design (fat chance. Also, note that readability is not the only important piece of this puzzle) no journal-worthy piece of science is going to come forward.
> Any reputable designer will instantly recognize both
No one is asking about "reputable designers". The tiny minority of people who can recognize hundreds of fonts at a glance are not interesting when it comes to this question.
There are certainly some objective metrics you can consider. Some that very quickly come to my mind:
1) Effect on reading speed 2) Effect on reading "endurance" (a little harder to define, but reasonable) 3) Effect on recall of text (prose) 4) Effect on recall in advertisements 5) Distance at which text of a given size becomes readable
There's also softer stuff along the lines of "does it make you feel warm and fuzzy", and you can still design experiments for that type of thing too.
So, I'm sorry, but I just don't believe you went you say you can't measure any of this.
No one is asking about "reputable designers". The tiny minority of people who can recognize hundreds of fonts at a glance are not interesting when it comes to this question.
There are certainly some objective metrics you can consider. Some that very quickly come to my mind:
1) Effect on reading speed 2) Effect on reading "endurance" (a little harder to define, but reasonable) 3) Effect on recall of text (prose) 4) Effect on recall in advertisements 5) Distance at which text of a given size becomes readable
There's also softer stuff along the lines of "does it make you feel warm and fuzzy", and you can still design experiments for that type of thing too.
So, I'm sorry, but I just don't believe you went you say you can't measure any of this.
I agree, readability can be tested. I am attempting to claim that readability is not the key factor. There have been font readability studies; but if designers were to follow them, all work would be black-on-white-large-x-height-minimal-serif-double-space-60-character-lines.
I am advocating seeing font as a tool designers use, not an end-user product. When I choose to develop a program using lisp instead of python (or vice-versa), I pick based on preference, knowing that I can produce a competant program with either --- though those programs will certainly differ somewhat --- not because of end-user useability studies of programs created with each.
I believe that equally readable, memorable, aesthetically pleasing design can be produced with either font; they are, after all, very similar. Designers find one more pleasant to use.
I am advocating seeing font as a tool designers use, not an end-user product. When I choose to develop a program using lisp instead of python (or vice-versa), I pick based on preference, knowing that I can produce a competant program with either --- though those programs will certainly differ somewhat --- not because of end-user useability studies of programs created with each.
I believe that equally readable, memorable, aesthetically pleasing design can be produced with either font; they are, after all, very similar. Designers find one more pleasant to use.
It isn't so much an issue of taste (or sentiment) as much as it is an issue of the design itself. Typography is about math/geometry as much as it is about taste. Helvetica was designed first, meticulously. Arial came later, resembling Helvetica, but breaking many typographic rules.
I'm a designer and I'm not particularly (at least not inordinately) impassioned about things like this, but I do find it important to recognize the difference between _design_ and art. It's often less about subjective preferences and more about purposeful/thoughtful solutions—in this case, what constitutes acceptable typography.
I'm a designer and I'm not particularly (at least not inordinately) impassioned about things like this, but I do find it important to recognize the difference between _design_ and art. It's often less about subjective preferences and more about purposeful/thoughtful solutions—in this case, what constitutes acceptable typography.
Note that Arial is not based on Helvetica, it is more like a cousin of Helvetica. Arial's ancestor is Grotesque 215. There's an excellent article about that here: http://www.ms-studio.com/articlesarialsid.html
I agree with you that this is design, often akin to industrial design more than art, which could also be described as art meets practical considerations or too narrowly, art meets commerce. And that's why I'm casting a wide net in my remarks by bringing up pixel alignment considerations, as well as licensing and other considerations. These are critical considerations to typography in practice.
For example the politics and business constraints surrounding the development, licensing, and use of Helvetica in the 90's are way more complex than simply saying, "Microsoft didn't want to pay for Helvetica", which is how many designers put it.
These complaints about licensing also, by the way, give away many of their visceral emotion-based reactions to Arial, rather than objective design purity. Clearly every party is affected in some way by the commerce of design, which often seems to inform designer's reactions too, especially when the licensing question is over-simplified.
For example the politics and business constraints surrounding the development, licensing, and use of Helvetica in the 90's are way more complex than simply saying, "Microsoft didn't want to pay for Helvetica", which is how many designers put it.
These complaints about licensing also, by the way, give away many of their visceral emotion-based reactions to Arial, rather than objective design purity. Clearly every party is affected in some way by the commerce of design, which often seems to inform designer's reactions too, especially when the licensing question is over-simplified.
Especially since Arial wasn't commissioned by Microsoft, it was designed in 1982 by Patricia Saunders and Robin Nicholas, commissioned by IBM for their laser printers. It was originally called Sonoran San Serif. Microsoft later shipped it with Windows 3.1.
Thanks for catching my error in the history.
This is blogspam, http://ragbag.tumblr.com/post/187708731/arial-helvetica-on-f... is the actual source.
Loved this article. Amazing how such a subtle difference can cause such a big difference in visual impact
This is just how to tell them apart easily. The sum of the visual difference between text set in Arial and Helvetica is bigger than that.
Um, the fact that there is a HOWTO tell the difference tells me that there isn't really much of a difference in real-world visual impact. Sure, Helvetica kerns a lot better than Arial, but the vast majority of people don't see the difference.
It might not be obvious given a few characters, but the overall impact is huge.
People who don't know the difference, and can't articulate it, will /still/ know something is wrong, even though they can't put their finger on it.
People who don't know the difference, and can't articulate it, will /still/ know something is wrong, even though they can't put their finger on it.
I would like to think this is true, but honestly find it difficult to believe. Do you have any evidence or research on the topic?
Which one should we use so that people don't feel bad due to our choice of minutely different font?
Meaning, which is the higher quality font between Arial and Helvetica?
Helvetica is the winner. Geometry is much tighter, plus the kerning and hinting is far better, and you have the possibility of using the many faces in the family from ultra condensed, to wide, from light to black.
Helvetica is the winner. Geometry is much tighter, plus the kerning and hinting is far better, and you have the possibility of using the many faces in the family from ultra condensed, to wide, from light to black.
How about not using any of them?
The biggest strength of Helvetica is its biggest weakness as well; it works with everything. Almost anything in any context can be written with Helvetica. In most cases though, it's not a particularly intersting choice, I could say, it's a bit boring to see Helvetica everywhere. Of course there are places where Helvetica is a very good fit; the New York Subway is one example of that in my opinion.
The biggest strength of Helvetica is its biggest weakness as well; it works with everything. Almost anything in any context can be written with Helvetica. In most cases though, it's not a particularly intersting choice, I could say, it's a bit boring to see Helvetica everywhere. Of course there are places where Helvetica is a very good fit; the New York Subway is one example of that in my opinion.
... but they may feel the difference without noticing.
from my own testing, it seems that people will sometimes say "that does not feel right", and when i switch it over to the other, they go "ah, that is better". just my own experience, which is why i find this so interesting
Strange timing. Yesterday I watched the documentary "Helvetica" (2007) which is all about typefaces:
http://www.imdb.com/title/tt0847817/
http://www.imdb.com/title/tt0847817/
Funny, I was just reading about the Baader-Meinhof Phenomenon recently.
Thanks for giving me the name of that phenomenon. I knew about it - wasn't pretending that it was some supernatural occurance - just that even if you know about it, it still feels a bit strange when it happens.
For people more typographically inclined, there is a much analysis and history of Arial at: http://www.ms-studio.com/articles.html
OK now i know the difference and how to spot Arial, but can anyone show me why is Helvetica superior to Arial ?
I was under the impression that Arial was a cheap knock-off of Helvetica. They took a similar font and reshaped it so it could stand in for Helvetica.
Clearly this is uninspired, much like a KDE theme trying to recreate Mac OS X. There's no actual law that says such a knock-off must be inferior to the original but a) it's off to a bad start, and b) even if it ties with the original it's got the taint of being a rip-off motivated by greed.
The full history of Arial, from a link someone else posted: http://www.ms-studio.com/articles.html
Note that Red Hat has in turn commisioned a font that is "metric compatible" with Arial: http://en.wikipedia.org/wiki/Liberation_fonts
Clearly this is uninspired, much like a KDE theme trying to recreate Mac OS X. There's no actual law that says such a knock-off must be inferior to the original but a) it's off to a bad start, and b) even if it ties with the original it's got the taint of being a rip-off motivated by greed.
The full history of Arial, from a link someone else posted: http://www.ms-studio.com/articles.html
Note that Red Hat has in turn commisioned a font that is "metric compatible" with Arial: http://en.wikipedia.org/wiki/Liberation_fonts
If they wanted an exact knock-off they could do so perfectly legally -- the only defensible IP afforded to typefaces is the trademark on their names -- the set of shapes for use in typesetting cannot be copyrighted or otherwise protected.
The link I provided above covers this and suggests that the foundry that created Arial didn't want to be a blatant pirate, so instead did just enough to pretend they weren't ripping off Helvetica.
"What is really strange about Arial is that it appears that Monotype was uncomfortable about doing a direct copy of Helvetica. They could very easily have done that and gotten away with it. Many type manufacturers in the past have done knock-offs of Helvetica that were indistinguishable or nearly so. [...] It’s quite possible that most of the “Helvetica” seen in the ’70s was actually not Helvetica.
Now, Monotype was a respected type foundry with a glorious past and perhaps the idea of being associated with these “pirates” was unacceptable. So, instead, they found a loophole and devised an “original” design that just happens to share exactly the same proportions and weight as another typeface.
This, to my mind, is almost worse than an outright copy. A copy, it could be said, pays homage (if not license fees) to the original by its very existence. Arial, on the other hand, pretends to be different. It says, in effect “I’m not Helvetica. I don’t even look like Helvetica!”, but gladly steps into the same shoes. In fact, it has no other role."
"What is really strange about Arial is that it appears that Monotype was uncomfortable about doing a direct copy of Helvetica. They could very easily have done that and gotten away with it. Many type manufacturers in the past have done knock-offs of Helvetica that were indistinguishable or nearly so. [...] It’s quite possible that most of the “Helvetica” seen in the ’70s was actually not Helvetica.
Now, Monotype was a respected type foundry with a glorious past and perhaps the idea of being associated with these “pirates” was unacceptable. So, instead, they found a loophole and devised an “original” design that just happens to share exactly the same proportions and weight as another typeface.
This, to my mind, is almost worse than an outright copy. A copy, it could be said, pays homage (if not license fees) to the original by its very existence. Arial, on the other hand, pretends to be different. It says, in effect “I’m not Helvetica. I don’t even look like Helvetica!”, but gladly steps into the same shoes. In fact, it has no other role."
I'm afraid most hackers need a chart showing differences between Arial and Verdana or even worse, Arial and Times!
Any feedback is appreciated.