Making of Resonay

An ambitious plan

Is it possible to create a serious font that would work as an illustration rather than a text? This idea originated sometime in 2014. I had the idea of fresh typeface, which would emphasize its construction with 2 overlapping layers. It should also be as easy to use as possible. I started sketching a script font (something like Abelina from Sudtipos):

One of the first sketches shows the solution of the narrowing of the middle stroke.


It consisted of 2 equivalent fonts that only make sense if they are exactly matched (too difficult, I know, but I believed in the power of this design). The design was to support various irregularities, such as a good script font – random mixing of several variants of characters, many discretionary ligatures, bouncing baseline, and other lively details. The only thing that made the design a little easier was that the characters didn’t have to be joined. The result was to be an attractive script with great possibilities of customizing graphic programs.

The first digital version of the font - it's very fresh, but I wanted to get more out of this principle.


Script AKA serif typeface

Honestly – such a font was simply beyond my skills with FontLab5. But it couldn’t be resisted – I was obsessed with it. Already the first test in InDesign looked promising, the font came with tremendous energy and sharpness. But there were also doubts. How does this necessary accurate fit work together with irregularities and living imperfections? Will it look natural? Will I finish it in retirement? Such sober thoughts led me to mechanize the font so that some shapes could be used modularly. It would be a bit of a strange form of italic. So no bouncing letters at the baseline, no random inaccuracies, but more classic proportions and repeating patterns. But if I already have italic, why not do the upright version as well? The different design of the upright version will certainly be a big challenge because the principle I had designed fit better into Italic. On the other hand, the font thus gains wider possibilities of use, as it approaches the typical form of a text font. So the script suddenly became a serif with a traditional scheme. The brush turned into metal matrices…

Rationalization, but also a mix of styles

As expected, the transition to italic went smoothly, only the upright was a bit uncertain, there was no ease. It was also not possible to clearly determine the position where the anchor points are to be placed (where exactly is the place where the arc is to overturn?). Such an ambiguity remained in the font until its 4th version (the year 2017).

From that moment on, a great rationalization and cleaning of the shapes began. The first to notice was the imperceptible continuity of layers (Base and Cover), so they seemed to emerge and disappear indefinitely. I later found out that these common “singularity points” can be placed on top of curves, as is generally recommended for font design. This saved a few unnecessary anchor points and justified their placement. Upright changed from an embarrassed boy to a smooth gentleman.


Similarly, the decision was made that the division into Base and Cover will not be based on the random alternation, but will take place according to their orientation – all vertical (thick) strokes will be Base and all horizontal (thin) strokes will be a Cover layer. This logic also meant an increase in the number of strokes. E.g. “o” got 4 strokes, instead of the original 2. Also the font lost some dynamics because e.g. said “o” actually became symmetrical in both directions. Symmetry and dynamics are not friends…

At this stage, the typeface was reminiscent of Baroque aesthetics with twisted terminals (maybe too sweet), so I thought of turning some tired curves into something more surprising, which could strike more. Since I was not bound by any historical model, I finally decided on an unconventional combination of ball and sharp terminals (even both in one letter), arranged only emotionally. The rotation of the pen, known from the gothic fracture, also penetrated italic (note the ending of “f” and “j”). Simply, it was not a matter of putting the font in a historical period, but taking from each period what is best suited to the design.

Evolution of Resonay from 2015 to 2020. We can see a shift from random and unbalanced design to clear and sleek curves with thoughtful construction. From trendy angular shapes, through baroque curls to renaissance sophistication.


Crazy idea

Here I came up with another madness – what if the whole Base layer could be used on its own? A huge complication for me, but interestingly expanded options for the user. After all, the maximum possible contrast looks great, it would be a pity if he could not show himself. The design had a natural tendency to do so, but how to solve the crossbar in “f” and “t”? After all, 2 strokes are crossing in different directions, which when I combine them into one, I deny the whole logic that I am trying to do. And what about the various ligatures, initial, final shapes, uppercase, and everything else I’ve already designed? I haven’t figured it out, but I estimate that this has delayed the development of Resonay by at least half a year.

How to solve the crossbar on “f” and “t”so that the base version can work completely independently? The whole crossbar could not move to the Cover version, but also could not stay in full force because it would look inappropriate. This has proven to be the best solution.


Serifs and return dynamics

With the connection of serifs in the cover version, I was puzzled from the moment I left the script form. How to choose the right one from so many options? Over time, it has narrowed at least that there must be a completely smooth transition to the stem and some asymmetry. And since they seemed too fragmented in version from 2018, I decided to go back to simplicity and just copy the flared shape of the Base font + one dynamic line connecting the farthest points of the serif. This also resulted in the dynamization of circular shapes, which now connect the outer and inner contours of the letter and everything remains beautifully smooth. Elegance à la impossible Penrose triangle:

The last change in design was the modification of spherical terminals to more vivid calligraphic curves. So when we look at that development in terms of character, it went from trendy angular shapes, through romantic baroque curls to sober renaissance sophistication.

Another voluntary complication on the horizon

I have noticed that this “Penrose feature” creates an attractive swirling movement in the letters. That’s exactly what was missing there – movement. But let’s go even further. What if the direction of this vortex could be influenced? Of course, I will make two versions – one will swirl from the inside out and the other from the outside in...

Fortunately, more important than the direction of the swirl is its incorporation into the overall shape. Too much attention to the adjustment of the “direction of rotation” would create other problems in addition to the need for new masters. For example, “o” would destroy the slanted axis that is typical of fonts of this type. I decided to leave it that way and make the dynamics of each letter individually according to its natural appearance.

Idea with swirl direction control. In the end, the variant on the right was not implemented due to the excessive complexity of some features.


From FontLab to Glyphs 

About the first third of font development was in FontLab5. Furthermore, I could no longer look at the jagged letters on the retina screen. But the transition to the Glyphs was not as painful as it might seem. Some characters needed to be redrawn completely, because FontLab is much more benevolent to interpolation errors (as the results correspond), while Glyphs doesn’t really allow you to make a mistake. The first interpolation test shocked me. When I managed to make the first compatible contours, their intermediate stage was so sharp and precise that it needed no further intervention. That gave me hope. To give you an idea of how I proceeded – both layers were drawn simultaneously (Base and Cover) into one file so that they could have 100% fitting, then manually sorted and divided into 2 files. Later I found out how the layers work (thanks to Jakob Runge), so again it was possible to keep them in one file and have better control. Although this acquaintance with the new environment delayed me a bit, overall, this program saved me an incredible amount of time. The sample tutorials on the web, where the beginner gets everything he could need, also helped a lot. This allowed me to stuff the font with features I didn‘t even know about before.

Italic variants of the letter “I” in FontLab.


Per Aspera ad Astra

To create a typeface that has no model in history actually means to put obstacles in the way of oneself and to overcome them oneself. The complexity of this layered design was that it is not possible to rely on a proven model here. Even though I still had the Venetian antiqua before my eyes, the division of strokes is a completely subjective matter, and it can be done in the infinity of ways, so you are constantly comparing them and most importantly – you are constantly discovering new ones! Design space was defined as follows:

it must be as visually attractive as possible

no straight vectors will be used there

it will stylize its calligraphic construction, which is normally hidden from us


It's hard to opt for one and best version of  “k”  when there are so many options. And the more you look at it, the more options appear...


With such a free setting, however, it begins to slide easily to overcombination, which could deprive the font of its power. In version 2018, I judged that this had just happened, and the design needed to be moderated. Too many curves have made the font fade. The solution was to make some outlines common to both font layers. This unified the fragmented design into a harmonious whole. Less is always more.

Whats the name?

The title of this font is a separate chapter. I like the names that indicate what the font is about. What can we expect from it. Thus, not according to the names of close people, or according to the place where it originated, it does not even have to show the most beautiful characters. But it is necessary to work towards a logical name, which also captures its atmosphere. Since this font offers the opportunity to see inside it, the first thing that came to my mind was the combination of the words Infra and Script. Logical, right?

However, as the serif evolved from this, I began to think of the suffix “ant” as “antiqua”, which would capture the historical background of the typeface. Since the sans version is also planned, the suffix “ans” can be inferred from this, which looked like a perfect match. I followed this concept for a very long time. I wrote down more than 150 ideas for names with this suffix. Some completely crazy, some quite usable. Here is a short selection:

Amusant, Animant, Arousant, Bellant, Brillant, Charmant, Cherant, Circulant, Clairant, Clarant, Cohesant, Conexant, Corsant, Courant, Creant, Dreamant, Endorphant, Enthusant, Espritant, Flagrant, Flamant, Flavant, Fleurant, Flirtant, Florant, Floreant, Fluant, Fluxant, Fragrant, Fraisant, Frescant, Glamant, Gradant, Illusant, Inspirant, Integrant, Liberant, Librant, Lustant, Organt, Pleasant, Premiant, Pulsant, Resonant, Revelant, Rosant, Spectant, Spherant, Venesant, Visiant

In the end, I opted for “Revelant”. It seemed to fulfill everything – it sounds like the name of a mythical hero, revel in English = entertainment, fun, celebration, and revelation = exposure, discovering, finding. Ant = antiqua. Everything fits. Except for one detail... When I started communicating with people about font promotion, everyone called it “Relevant”. Simply swapping the letters made the well-fitting name (and long months sought) something quite ordinary and inappropriate. People simply automatically assumed it was an error and “corrected” it themselves. Weak name + error was worse in the result than if the font was called “Use me for a great design”. It also happened to N. Thomsen, from TypeMates. I dealt the release of the font with him. After consulting with him, I decided to rename the font. The suffix “ant” seemed exhausted and restrictive to me, so I was looking for something else. An artistic impression of the font led me to the nice and simple name “Portray” (in general, it is a font intended for viewing rather than reading). It turned out that there was a font with a similar name “Portrait”, so it wouldn’t sound very professional either. In the end, we agreed on “Resonay”, which sounds more original, also contains a call to action, and suggests something from the atmosphere of the font.

Lessons learned from this

Typedesign like any other design is made for people. Only users can set a true mirror for it. If they don’t understand your design, it doesn’t mean they’re stupid, or it’s a stupid design, just maybe you didn’t have enough distance from your creation and you overcomplicated it with good intentions. Today’s type design directly seduces to this – the possibilities are constantly expanding, the fonts are becoming more complex. From a certain point, however, it goes beyond the distinctive ability of the user. He may begin to consider your artistic elements or features as a mistake. And so your overwork becomes not only unnecessary for a good idea but even harmful. This probably happens whenever you work on something for too long. However, working on something for a very long time and not killing the freshness in it is the ultimate challenge for any designer. This is the lesson this project taught me.

Resonay is available for purchase on TypeMates.