Insight
More Frightening Than Decay: The Fear of Never Decaying — The Story of 「Sandoll Jeongche」.
Fonts are broadly divided into display types and body types.
Body types are difficult to produce, require a long development period, and tend to be used continuously once adopted. For this reason, a large proportion of newly released fonts are display types. That said, foundries with a long history almost always possess at least one highly refined body type. A well-made body type is proof of a foundry’s history and tradition—it is a flagship model designed to be developed over a long period and sold over an equally long time.
A Long-Postponed Task Finally Surfaces
Through a user survey conducted on SandollCloud, Sandoll’s font platform, we received clear feedback indicating a lack of body text fonts. It was a task long acknowledged but repeatedly postponed, as it required a substantial investment of both budget and manpower. In the spring of 2017, accumulated guilt finally bore fruit, and a dedicated task force was formed.
Notably, the project director was not a font designer, but a typographer with extensive experience in printing, publishing, and typesetting. This was a deliberate decision. Rather than approaching the project from the perspective of someone who had made many fonts, we chose the perspective of someone who had used many fonts. This allowed for a meticulous examination of what a body type truly needs to fulfill. In effect, a long-time heavy user, reflecting on long-held dissatisfactions with Hangul typography, stepped directly into production.
Building a Team with Strangers
As the scale of a project grows, so does the burden on those responsible. To be honest, the dominant thought was not “I must do this well,” but rather “What if this fails?” Everyone involved was kind and supportive, yet the pressure was immense. I read Korean and Japanese calligraphy theory books indiscriminately, practiced brush writing whenever possible, and before going to sleep, watched calligraphy videos from Korea, China, and Japan. Once a conceptual direction was established, I created prototype fonts myself to test their validity. This process enabled smooth communication with veteran font designers. Failure felt like something from which it would be difficult to recover.
Next came team formation. When an external director works with designers they are meeting for the first time, what should come first? We decided to extract what each person held in their mind and establish a shared goal and vocabulary. The fastest and most effective method was a hands-on workshop—avoiding digital tools altogether. We talked, listened, wrote, and drew together, articulating our ideal body type structures, identifying differences in thinking, and gradually narrowing them. One workshop involved extracting the skeleton of the respected SM SeMyeongjo and rebuilding flesh upon it.
Concepts and formats designed to concretely visualize what each person considers an ideal body text font.
When drawn at a slightly larger scale, elements that were previously buried or hidden are revealed with stark clarity.
To draw at this scale requires a great deal of consideration. Completing a single sheet of paper took several days.
What matters here is not the result, but the process of refining one’s thinking itself.
Researching Glyph Density and Frequency
To create a font compliant with KS standards, 2,350 Hangul syllables must be designed. Understanding the identity of this character set felt essential before proceeding. We therefore analyzed glyph density and frequency. Density was calculated by assigning values to stroke lengths and summing them, while frequency was derived by decomposing the 2,350 syllables into jaso and tallying their positional occurrences (initial, medial, final). Initial consonants such as ㅇ, ㄱ, ㅅ, ㅎ, ㄴ appeared most frequently, while final consonants such as ㄴ, ㄹ, ㅇ, ㅁ, ㅂ, ㄱ dominated. This research allowed us to establish a clear priority for which glyphs to draw first—an indispensable step in creating a new process.
The analysis covered 2,350 characters, with each character given equal weight. However, it was necessary to identify which characters are used most frequently. Unfortunately, there was no way to determine which characters people use most often today.
A New Process: The Power of Short Iterations
To create something new, a new process is required. One cannot draw a forest after seeing only a single tree. Designing characters one by one makes it difficult to predict how they will behave in full sentences. The quality of a body type depends on how early the designer can begin reading continuous text. In conventional processes, typesetting tests occur only after many glyphs are completed, making meaningful revision difficult. Since revision effort scales with glyph count, we instead selected approximately 460 key characters—core glyphs plus particles and endings—and wrote text using only those. Because literature is a common use case for body types, we wrote a short, fictional narrative. With so few words available, the result resembled a four-dimensional fantasy novel.
First, Second, and Third Refinement
By drawing a few hundred characters, typesetting them, reading the result, and revising it repeatedly, the font gradually takes shape. This process could be described as test-driven development, agile methodology, or iteration. Since the text is composed of only a few hundred characters, the task at this stage is simply to design. However, it is crucial to remember that the characters will inevitably look crude at this point, and the moment one gives in to the urge to refine them, the entire development plan begins to unravel.
The results of the first typesetting test were, as expected, unsightly—and fortunately, I resisted the urge to intervene. Fixing ‘ㄱ’ inevitably leads to fixing ‘ㅋ’ and then ‘ㄲ’; once you try to correct one character properly, the scope of the work quickly balloons. It may seem easier to perfect one character before moving on to the next, but without drawing many characters and seeing them typeset together, it is difficult to know what the best choice truly is. In other words, unless there is a clear and reliable criterion for judgment, I chose not to make changes even when something felt unsatisfactory. More precisely, I could not afford to change it.
Instead, I initially focused on determining the positions of the longest strokes, such as those in ‘ㅣ’ and ‘ㅡ’. Consonant components like ‘ㄱㄴㄷㄹ’ were merely placed at roughly appropriate sizes and positions. The process was divided into six rounds, each with clearly defined checkpoints, gradually raising the level of completeness step by step. Even when there were countless areas I wanted to revise, if they did not fall within the checklist for that round, I deliberately ignored them and moved on. For a designer whose instinct is to immediately reach out and fix any awkwardness, this was far from an easy process—something like living under a damp blanket for six months.
The second round was much the same. There was still a long way to go. Rather than being distracted by how crude it looked, the focus was kept strictly on the combinations of consonants and vowels. At this stage, the positions and lengths of the strokes were what mattered most; the outlines and surface details could be refined later.
Fourth, Fifth, and Sixth Refinement
By the fourth round, the typeface finally began to take on a convincing shape. With the addition of punctuation in the fifth round, it became genuinely readable, which was gratifying after such a long wait. Once the skeleton had settled into place, it was finally possible to examine the outlines in earnest. To do so, we had to move away from evaluating the type solely through running text and the overall flow of lines. Instead, we created proofing texts focused on individual words and single characters.
In detailed work, purposefulness matters more than a designer’s personal sensibility. There is no true end in this field, and without entering it hand in hand with others, it is easy to lose one’s way. Like scuba divers plunging into the open sea, teamwork that safeguards one another is essential. Well-designed proofing texts fulfilled precisely that role.
As numerals and Latin characters were added, ambition inevitably grew. The Latin set was expanded to cover the full pan-European range, italics were included, and even decorative swash capitals were added. Punctuation was also designed separately for Hangul and for Latin characters. It was demanding work, but once the end began to come into view, the process became exhilarating. The worry of “What if this fails?” gradually transformed into a determination to “Make it good.”
Back to the Basics—The Launch Event for the New Text Typeface, 「Jeongche」
The work of a font foundry can be broadly divided into two categories:
(1) creating fonts on commission for corporate clients (custom font development), and
(2) creating its own fonts (in-house font development).
Developing one’s own fonts is rewarding, but it is not easy, as revenue is generated slowly and over a long period of time. Hosting a launch event on top of that is an audacious and highly unconventional challenge—one that remains extremely rare even internationally. Dozens of people spent a full year preparing for it. Invitations were sent with care to those we hoped would attend, venues were reserved to ensure that no guest would be turned away, sandwiches were prepared in case attendees missed dinner after work, and time was allotted for those who needed to step out and eat. We did not want anyone to leave empty-handed, so gifts and prizes were prepared, and a talk show was organized with designers who had created books using Jeongche. We resolved not to let the event run too late, mindful that traveling home late at night could be unsafe.
(That said, both the presentation and the talk show ultimately ran over time.)
『Sandoll Jeongche』 is not the name of a family, but of an entire lineage. That alone suggests its scale. Development is still ongoing. This is why its name is expressed using three-digit numbers. The first digit indicates modernity, the second indicates weight, and the third represents character width. 「Jeongche 930」 is more modern than 「Jeongche 530」, while the weight and width remain the same. At times, a single Latin letter is added as a suffix. 'i' denotes italic. 「Jeongche 530」 is the roman style, and 「Jeongche 530i」 is the italic style.
Talk Show Panel (From left to right)
Janghyun Moon-CEO of General Graphics, Kyungsoo Lee-Creative Director at Workroom, Ilsun Hwang-Senior Manager at Minumsa.
More Frightening Than Decay Is Not Decaying at All
No matter how much time passes, a digital font remains exactly as it was at the moment of its creation. Something that does not decay is called immortal (不朽). The world contains both immortal masterpieces and immortal failures. If something I created were to wander the world forever as an undecaying failure, what a sorrowful fate that would be. For this reason, rather than dreaming of a masterpiece, I am more concerned with avoiding failure. The disgrace of a failure outweighs the honor of a masterpiece.
At Sandoll, there exists—often unspoken—an implicit lower bound: “it must be at least this good.” (There is no need to define an upper bound.) A realistic and flexible mindset—that the world can be made more beautiful simply by not being ugly—has naturally become a philosophy. This is because, over the past 38 years, we have experienced little by little that fonts truly can make the world more beautiful. Having done this collectively and consistently, we may have sensed it slightly earlier than others. I would like to share this quietly sweet feeling with you as well.
Insight
More Frightening Than Decay: The Fear of Never Decaying — The Story of 「Sandoll Jeongche」.
Fonts are broadly divided into display types and body types.
Body types are difficult to produce, require a long development period, and tend to be used continuously once adopted. For this reason, a large proportion of newly released fonts are display types. That said, foundries with a long history almost always possess at least one highly refined body type. A well-made body type is proof of a foundry’s history and tradition—it is a flagship model designed to be developed over a long period and sold over an equally long time.
A Long-Postponed Task Finally Surfaces
Through a user survey conducted on SandollCloud, Sandoll’s font platform, we received clear feedback indicating a lack of body text fonts. It was a task long acknowledged but repeatedly postponed, as it required a substantial investment of both budget and manpower. In the spring of 2017, accumulated guilt finally bore fruit, and a dedicated task force was formed.
Notably, the project director was not a font designer, but a typographer with extensive experience in printing, publishing, and typesetting. This was a deliberate decision. Rather than approaching the project from the perspective of someone who had made many fonts, we chose the perspective of someone who had used many fonts. This allowed for a meticulous examination of what a body type truly needs to fulfill. In effect, a long-time heavy user, reflecting on long-held dissatisfactions with Hangul typography, stepped directly into production.
Building a Team with Strangers
As the scale of a project grows, so does the burden on those responsible. To be honest, the dominant thought was not “I must do this well,” but rather “What if this fails?” Everyone involved was kind and supportive, yet the pressure was immense. I read Korean and Japanese calligraphy theory books indiscriminately, practiced brush writing whenever possible, and before going to sleep, watched calligraphy videos from Korea, China, and Japan. Once a conceptual direction was established, I created prototype fonts myself to test their validity. This process enabled smooth communication with veteran font designers. Failure felt like something from which it would be difficult to recover.
Next came team formation. When an external director works with designers they are meeting for the first time, what should come first? We decided to extract what each person held in their mind and establish a shared goal and vocabulary. The fastest and most effective method was a hands-on workshop—avoiding digital tools altogether. We talked, listened, wrote, and drew together, articulating our ideal body type structures, identifying differences in thinking, and gradually narrowing them. One workshop involved extracting the skeleton of the respected SM SeMyeongjo and rebuilding flesh upon it.
Concepts and formats designed to concretely visualize what each person considers an ideal body text font.
When drawn at a slightly larger scale, elements that were previously buried or hidden are revealed with stark clarity.
To draw at this scale requires a great deal of consideration. Completing a single sheet of paper took several days.
What matters here is not the result, but the process of refining one’s thinking itself.
Researching Glyph Density and Frequency
To create a font compliant with KS standards, 2,350 Hangul syllables must be designed. Understanding the identity of this character set felt essential before proceeding. We therefore analyzed glyph density and frequency. Density was calculated by assigning values to stroke lengths and summing them, while frequency was derived by decomposing the 2,350 syllables into jaso and tallying their positional occurrences (initial, medial, final). Initial consonants such as ㅇ, ㄱ, ㅅ, ㅎ, ㄴ appeared most frequently, while final consonants such as ㄴ, ㄹ, ㅇ, ㅁ, ㅂ, ㄱ dominated. This research allowed us to establish a clear priority for which glyphs to draw first—an indispensable step in creating a new process.
The analysis covered 2,350 characters, with each character given equal weight. However, it was necessary to identify which characters are used most frequently. Unfortunately, there was no way to determine which characters people use most often today.
A New Process: The Power of Short Iterations
To create something new, a new process is required. One cannot draw a forest after seeing only a single tree. Designing characters one by one makes it difficult to predict how they will behave in full sentences. The quality of a body type depends on how early the designer can begin reading continuous text. In conventional processes, typesetting tests occur only after many glyphs are completed, making meaningful revision difficult. Since revision effort scales with glyph count, we instead selected approximately 460 key characters—core glyphs plus particles and endings—and wrote text using only those. Because literature is a common use case for body types, we wrote a short, fictional narrative. With so few words available, the result resembled a four-dimensional fantasy novel.
First, Second, and Third Refinement
By drawing a few hundred characters, typesetting them, reading the result, and revising it repeatedly, the font gradually takes shape. This process could be described as test-driven development, agile methodology, or iteration. Since the text is composed of only a few hundred characters, the task at this stage is simply to design. However, it is crucial to remember that the characters will inevitably look crude at this point, and the moment one gives in to the urge to refine them, the entire development plan begins to unravel.
The results of the first typesetting test were, as expected, unsightly—and fortunately, I resisted the urge to intervene. Fixing ‘ㄱ’ inevitably leads to fixing ‘ㅋ’ and then ‘ㄲ’; once you try to correct one character properly, the scope of the work quickly balloons. It may seem easier to perfect one character before moving on to the next, but without drawing many characters and seeing them typeset together, it is difficult to know what the best choice truly is. In other words, unless there is a clear and reliable criterion for judgment, I chose not to make changes even when something felt unsatisfactory. More precisely, I could not afford to change it.
Instead, I initially focused on determining the positions of the longest strokes, such as those in ‘ㅣ’ and ‘ㅡ’. Consonant components like ‘ㄱㄴㄷㄹ’ were merely placed at roughly appropriate sizes and positions. The process was divided into six rounds, each with clearly defined checkpoints, gradually raising the level of completeness step by step. Even when there were countless areas I wanted to revise, if they did not fall within the checklist for that round, I deliberately ignored them and moved on. For a designer whose instinct is to immediately reach out and fix any awkwardness, this was far from an easy process—something like living under a damp blanket for six months.
The second round was much the same. There was still a long way to go. Rather than being distracted by how crude it looked, the focus was kept strictly on the combinations of consonants and vowels. At this stage, the positions and lengths of the strokes were what mattered most; the outlines and surface details could be refined later.
Fourth, Fifth, and Sixth Refinement
By the fourth round, the typeface finally began to take on a convincing shape. With the addition of punctuation in the fifth round, it became genuinely readable, which was gratifying after such a long wait. Once the skeleton had settled into place, it was finally possible to examine the outlines in earnest. To do so, we had to move away from evaluating the type solely through running text and the overall flow of lines. Instead, we created proofing texts focused on individual words and single characters.
In detailed work, purposefulness matters more than a designer’s personal sensibility. There is no true end in this field, and without entering it hand in hand with others, it is easy to lose one’s way. Like scuba divers plunging into the open sea, teamwork that safeguards one another is essential. Well-designed proofing texts fulfilled precisely that role.
As numerals and Latin characters were added, ambition inevitably grew. The Latin set was expanded to cover the full pan-European range, italics were included, and even decorative swash capitals were added. Punctuation was also designed separately for Hangul and for Latin characters. It was demanding work, but once the end began to come into view, the process became exhilarating. The worry of “What if this fails?” gradually transformed into a determination to “Make it good.”
Back to the Basics—The Launch Event for the New Text Typeface, 「Jeongche」
The work of a font foundry can be broadly divided into two categories:
(1) creating fonts on commission for corporate clients (custom font development), and
(2) creating its own fonts (in-house font development).
Developing one’s own fonts is rewarding, but it is not easy, as revenue is generated slowly and over a long period of time. Hosting a launch event on top of that is an audacious and highly unconventional challenge—one that remains extremely rare even internationally. Dozens of people spent a full year preparing for it. Invitations were sent with care to those we hoped would attend, venues were reserved to ensure that no guest would be turned away, sandwiches were prepared in case attendees missed dinner after work, and time was allotted for those who needed to step out and eat. We did not want anyone to leave empty-handed, so gifts and prizes were prepared, and a talk show was organized with designers who had created books using Jeongche. We resolved not to let the event run too late, mindful that traveling home late at night could be unsafe.
(That said, both the presentation and the talk show ultimately ran over time.)
『Sandoll Jeongche』 is not the name of a family, but of an entire lineage. That alone suggests its scale. Development is still ongoing. This is why its name is expressed using three-digit numbers. The first digit indicates modernity, the second indicates weight, and the third represents character width. 「Jeongche 930」 is more modern than 「Jeongche 530」, while the weight and width remain the same. At times, a single Latin letter is added as a suffix. 'i' denotes italic. 「Jeongche 530」 is the roman style, and 「Jeongche 530i」 is the italic style.
Talk Show Panel (From left to right)
Janghyun Moon-CEO of General Graphics, Kyungsoo Lee-Creative Director at Workroom, Ilsun Hwang-Senior Manager at Minumsa.
More Frightening Than Decay Is Not Decaying at All
No matter how much time passes, a digital font remains exactly as it was at the moment of its creation. Something that does not decay is called immortal (不朽). The world contains both immortal masterpieces and immortal failures. If something I created were to wander the world forever as an undecaying failure, what a sorrowful fate that would be. For this reason, rather than dreaming of a masterpiece, I am more concerned with avoiding failure. The disgrace of a failure outweighs the honor of a masterpiece.
At Sandoll, there exists—often unspoken—an implicit lower bound: “it must be at least this good.” (There is no need to define an upper bound.) A realistic and flexible mindset—that the world can be made more beautiful simply by not being ugly—has naturally become a philosophy. This is because, over the past 38 years, we have experienced little by little that fonts truly can make the world more beautiful. Having done this collectively and consistently, we may have sensed it slightly earlier than others. I would like to share this quietly sweet feeling with you as well.