Opinions about a font combo making website.

[jcavill]

This topic was imported from the Typophile platform

Hello everyone!
My name is Jerry and I am a current student at Kutztown University in Pennsylvania taking the communication design program.
A group of us are taking part in a startup weekend and decided to begin the creation of a website (type-wich.com) to help with the creation of font combos for individuals looking to create anything in the design world. Or anywhere else really.
So, the way it will work, an individual will pick a font for the header and a list of fonts that will be possible good looking matches for the body. This way it can be quicker for designer to find another font that will work well with their first pick easier.

This site will also work as a filtration system to bring in most of the fonts from freelancers/type founders and smoothly dispense it to an individual. Because of this, there is a possibility for someone to see a font which they might like and buy it through us from the designer.

So, I am asking for ideas and opinions on what you guys/girls think of this idea!

[bemerx25]

So a website that helps designers pair fonts...for any medium (print/web)? And they could then purchase a license to said fonts through this site?

Rhetorically:
So why are they called "designers" then if they need to rely on an outside service to select better font pairings? Isn't that why clients hire them? To make these design decisions?

Back on topic:
I personally think this won't succeed - you're going to be competing in a fairly compressed market filled with fairly large vendors who offer similar services (not necessarily the font pairing aspect, but being able to view fonts, see how they typeset, and then making them easily purchasable.) But the decision is up to you and who knows?

[hrant]

I think this is a good idea (it's just a tool for designers to use - it shouldn't replace their own judgement) but what might make it really special is some kind of formula that recommends pairings (as opposed to simply brute-forcing canned lists of pairs).

hhp

[Thomas Phinney]

http://fontfuse.webink.com/

[dberlowgone]

"...some kind of formula that recommends pairings (as opposed to simply brute-forcing canned lists of pairs)."

That is funny.

[hrant]

Of course this sort of thing would only work very well if we had a robust feature-based font classification system, which we don't. Even then somebody would have to manually classify each font - I wouldn't expect it to do that on its own.

But even without a good font classification system such an algorithm might still serve as a fallible guide that gives ideas, many of which a user would reject.

hhp

[dberlowgone]

What's a "feature-based font classification system"?

[Nick Shinn]

Anything goes.

**

I wouldn’t worry about how your site assesses which fonts combine well.
That’s curation and market research, not design.
As a design student, you will get the most out of your project by designing the site to show users how two fonts selected will look in various settings.
Many type foundries already have this kind of functionality, which you can use as a benchmark.

As for choosing your short-list of faces and matchmaking their hook-ups, that's something you should do through your own research of typography, rather than looking for some automated process, which will not develop your knowledge and taste in typography.

[hrant]

David, what I mean is a sort of anti-Bringhurst classification; not based on cultural phenomena (like calling Didone "Romantic") but based on actual features, like serif shape. Nick Sherman -who has teamed up with Indra on classification work- could do a much better job explaining it. Assuming I correctly understand where he's coming from...

Anything goes.

So you like chocolate covered olives?

will not develop your knowledge and taste in typography.

Actually being forced to think of something algorithmically typical results in much deeper insight.

hhp

[Nick Shinn]

… being forced to think of something algorithmically typical results in much deeper insight.

Being forced to think algorithmically is not why people go to design school, no matter how good it might be for them.

[hrant]

I don't think you should assume that:
- All design schools are flawed in that respect.
- Jerry cannot "bypass" such a flaw.

hhp

[This Machine Floats]

http://ifontyou.com/ does this to some extent, it shows examples of font pairings at work.

[hrant]

This one's for you Nick!
http://www.bbc.com/travel/slideshow/20121107-americas-most-outrageous-bu...

hhp

[Nick Shinn]

I don't think you should assume that:
- All design schools are flawed in that respect.
- Jerry cannot "bypass" such a flaw.

I’ve taught type design at a number of design schools, as part of a B.Des or similar program. It’s one of the more dry, technical subjects in their curriculum, and many students hate it, no matter what. Being forced to learn about sidebearings and alignment zones in Fontlab (a difficult application that none of them are likely to use professionally) is too “algorithmic” for many of them. I’ve tried a number of different approaches, and those where there is less emphasis on technical rigor and more on open-ended invention and personal expression are preferred by students—although I don’t believe, as you say Hrant, they offer quite such deep insight into typography. So one strives for balance.

As this project has been described, there is plenty of opportunity for creative analysis and categorization of type forms in designing a “combo” site. My advice was simply not to get sucked (or forced!) into spending a lot of time figuring out a complex seemingly objective formula for type categorizating and matching, but rather to take it as a given, becuase the point of the exercise is to design a web site, not to develop a system of typeface classification.

But perhaps I am wrong. Perhaps that’s where graphic design is headed.