Skip to main content
  1. Home
  2. Web
  3. News

New open source font could eliminate those weird characters after translation

Hillary K. Grigonis
By

adobe source hans serif font type team copy
Adobe's Typekit team recently partnered with Google to develop Source Han Serif. Adobe
The World Wide Web has long been bringing ideas across cultural borders, but a new font developed by Adobe and Google aims to simplify the development of apps and websites across multiple languages, eliminating errors when translating text. Source Han Serif is a new font containing simplified Chinese, Traditional Chinese, Japanese, Korean, and English characters, each in downloads of less than 25 megabytes.

By containing multiple characters in one font, Source Han Serif aims to give developers working on programs in multiple languages consistency across the type design. While English (and many other Latin character-based languages) only needs to span 26 letters with upper and lower case variations along with symbols, accented variants, and grammatical symbols, East Asian languages contain a much broader number of characters. When those characters aren’t part of the font family, the translated letter then becomes a little box marking missing information. The new open source font contains 65,535 different glyphs or characters.

Related Videos

More: Spector makes discovering fonts and colors in print easy

While fonts containing so many characters tend to be at least 40MB, Source Han Serif is less than 25MB, with downloads available with the set as a whole or in language subsets. The smaller file size doesn’t make any usability sacrifices, Adobe senior manager Dan Rhatigan explained, but instead uses files more efficiently by mapping out each region’s shared glyphs, even with as many as four regional variations of each character within the same language.

The font complements Source Han Sans, which launched in 2014 with similar universal goals. The latest font expands on the original as a serif font — or one that contains those little lines where the large line ends, like at the feet of an “A.” Serif fonts are traditionally used in body text while the cleaner look of a Sans Serif (or no serif) font tends to be favored for headings and large text.

Adobe says the font is designed to be screen-friendly, with each glyph containing a separate horizontal and vertical format. The font family also includes seven different weight and style variations (such as bold and italic). By adding a Serif font to the earlier release, the fonts can be used together in more applications, Adobe says.

Source Han Serif (or Noto CJK Serif for Google) supports the subtle differences between writing conventions in four regions. Google

“Source Han Serif pushes the boundaries of what’s really capable with the font format,” Rhatigan said. “What we are able to do in combining all of these and looking how these glyphs can be shared, is that we were able to push the envelope in the size of the family while occupying a relatively small footprint compared to fonts offerings from other countries.”

While fonts containing the vast number of characters in Chinese Korean and Japanese are expensive to develop, Source Han Serif is open source, and available free through Adobe Typekit in North America and Japan. In China and Korea, the font family is available on GitHub. Google is also releasing the font under the name Noto Serif CJK from the Google Noto page.

Editors' Recommendations

Topics
Google has open-sourced TensorFlow, their new deep-learning system

Some tech companies prefer to keep their data and research private, but Google is not just some tech company. In an effort to spread the wealth and make everyone and everything smarter (especially our apps), Google has decided to open source their machine learning system, TensorFlow, making it possible for anyone and everyone to access the huge potential benefits of Deep Learning. In a blog post announcing the new release, Jeff Dean, a Senior Google Fellow, and technical lead Rajat Monga note that TensorFlow is not only "great for research, but it’s ready for use in real products too."

Deep Learning, a technique recently applied by engineers at Pinterest in order to build their visual search tool, effectively teaches computers to see patterns in large sets of data, such as similarities in appearance. It's what Google uses in Google Photos, and while the technology isn't perfect, it has certainly allowed the Internet giant to come a long way and provide some pretty spectacular tools. And now that it's going open-source, the hope is that it'll get even better.

Read more
Internet guerrillas: Inside the DIY broadband revolution with NYC Mesh
nyc mesh guerrilla internet network screen shot 2022 02 20 at 5 53 39 am

Toby Bloch doesn’t look like your average internet installation technician. Instead of a uniform with a corporate logo embroidered on it, he wears worn-in jeans and a thick canvas jacket. Instead of a van, he drives a Subaru -- the back of which is stuffed to the gills with a disorganized pile of hand tools, cables, and odd electronic devices with antennas sticking out of them. And unlike most technicians, he isn’t going to earn a dime for the appointment he’s headed to in Brooklyn.

But oddly enough, that’s precisely the point. Bloch doesn’t operate like a normal internet install tech because he isn’t one. He doesn’t work for Comcast or Spectrum or Verizon or any other large internet service provider (ISP). He’s a volunteer at NYC Mesh: A guerrilla internet provider that helps residents get online without paying a monthly fee to the aforementioned telecom companies.

Read more
How Big Jet TV won the internet
A plane landing during Storm Eunice in the UK in February 2022.

As millions of people hunkered down at home on Friday during the U.K.'s worst storm in 32 years, aviation enthusiast Jerry Dyer jumped in his van and drove to London’s Heathrow Airport to livestream passenger jets coming in to land in the challenging conditions.

Within a few hours of Dyer launching his Big Jet TV livestream from the top of his vehicle at the end of Heathrow’s runway 27L, social media started to take notice, with shares and retweets pushing his audience to as high as 200,000 people during the eight-hour livestream.

Read more