Aplogies. Haven’t updated the blog for a long time. A lot of work has happened since January on Chukki. Will write loads of posts today. Will also feature some scans. ūüôā

As I mentioned earlier, the design and research is happening parallely. Just to let you know what to expect.

Here are the things that I have done:*

(i) Freehand explorations of base letters
(ii) Selecting one form, that is consistent in width, proportion and form, from the explorations
(iii) Vector outlines of all the hand drawn base forms
(iv) Place the vector outlines on pixel grid, and construct a pixel form
(v) Use the base form to create the proportion, width, and thrust guidelines
(vi) Create all derived and independent forms

These tasks are yet to be done:

Elective 4 (Research based project)

(vii) Analysis of the Kannada script on the web (Objective)
(viii) Case studies of typefaces available on web (Approach)
(ix) Documenting the analysis and write a paper (Deliverable)

Most crucial of the tasks:

(x) Investigating the forms
(xi) Create design guidelines

* Apart from the tasks that are done, I have also spent a lot of time figuring out font creation softwares such as FontLab and SimpleFont.



Yesterday, I had a discussion with my guide. And it was really helpful in two ways. One, I got a chance to explain him the progress of the project, and two, he provided some valuable insights.

He was quite happy with the geometric construction of the letterforms. But he wasn’t entirely satisfied with it. These were his insights:

(i) Complex to simple
Yes, once i categorized the letters based on the basic form. I have to now choose any three letters which are really complex. When I say complex, it has to be a combination of curves, multidirectional strokes, more than two points of thrusts, and maximum width. So¬†when¬†I set them right I¬†would be able to come¬†up with a¬†better¬†‘design criteria’.

(ii) Structure follows form or the other way around
Balance is a key. Now how I am approaching is limiting and may not be able to innovate. I am creating a structure and placing the letterforms inside the structure. There is another way to it. I can bring in spontaneity into the froms. But this approach may be complicated to because I need to standardize the grid. So I need to use both the approaches and create a design criteria based on the analysis from both approaches.

(iii) Look at letters as pure forms
I have been looking at letters as forms. But I need to really look at them as images. To gather some kind of an insight that would drive the design. I would be able to only derive that with some sketches of the letterforms.

That’s it, folks. Happy New Year.

Thank you, Suresh.

Awakening. Sort of.

The Kannada script can also be categorized according to the thrust and movement of each letterform.

In an Indic letterform, there is a dominant feature and a sub-dominant features. And I had exactly the same criteria to categorize the letters according to the form. But now there is another insight that was provided by a language expert, Dr Vidwans, a professor at National Institute of Design.

There is something called ‘Thrust’. Yes and that is a very important aspect in type design, especially in Indic type design. To explain it plainly, Thrust helps in readability. At the end of the day, although each letter is legible, it will co-exist with other letter. So there has to some kind of a hint for the eye to smoothly accelerate through text.

There are two kinds of thrust: (i) explicit thrust and (ii) implicit thrust. The explict thrust can be seen dominant in the Devanagari script. The Devanagari letterforms hang like fruits from the topbar. The Roman letterforms have strong base. And it seems, Oriya scipt is the most balanced.

In Kannada script few letters have top bars and others don’t. So I need to work on the visual grammar. Each letterform has to be seen with resect to forms of¬†other letters.

Right now, I am working on the skeleton of the Kannada letterform. I am looking at letterforms as lives since we seem them all around. We intearct with them constantly.

The reference to living being helps. Skeleton helps consolidate the design criteria such as proportion, size, movement, and balance. Then I need to add body parts and give it a shape. And then I need to look at the soul of the typeface. And try to see how I very implicitly place the soul in each letterform.

I would like to share more when I reach that stage. As of now, this is my design process:
(i) Construction of base and independent letterforms
(ii) Creating a visual grammar
(iii) Placing soul in each letterform

Thank you Dr Vidwans.




The geometric construction has enhanced the look of the base letterforms. They now look serene. And proportions are maintained.

Another interesting thing (once again the thought borrowed from Dr Bhagvat’s process of categorization) is that every letterform will have a specific unit based on the width.

Take a look at the construction of skeleton of the base letterforms. These are intial sketches.

There are many reasons why skeleton should be build upon first.

Skeleton of the letteform will have a structure that helps gain legibility even before the weights are stuffed into the structure. Once the skeleton is perfect, it is just a matter of choice to place different weights around the defined structure.

With this exercise, I have learnt these things:

1. Place letterforms under specific width. Categorize letterforms based on the width¬† too. And if needed, create one’s own unit for width. I have used Graphical unit. One graphical unit being 1 cm. Most letterforms are place inside 4 graphical unit.
2. The logic behind curves in all letterforms can be repeated to control too many variations. So the circles that help construct one base letterform may help construct another letterform without increasing the radius of any of the circles.
3. If the solution to the problem is right, then it has to be beautiful too. Just solving a problem, in type design, is no more a challenge.
4. Before constructing letteforms, clarity can be achieved by creating thumbnails of the trial constructions. Then based on the analysis of thumbnails, final constructions can be carried out. This will save a lot of time and iterations.
5. Why geometry? Because symmetry comes into play. And symmetry definitely works on the screen. And I am sure I will have tough time dealing with these forms on the screen grid.
6. Using graph sheets as medium to plan the construction is recommended. I am using a basic Camlin compass to create circles of specific radius, a blue leaded pencil to mark the centerpoints of the circles and to draw diagonals that help decide the form of most letters, and a charcoal pencil to extract the form out of the construction.



Another challenge was to take all the base forms and see if I would be able to relate something with each one of them.

Now, how would this help me design a type?¬†Maybe it doesn’t help me. But that was my way of looking at aesthetics. I mean if the form has to look really beautiful,¬†beyond good proportions and legibility, it has to be related to the aesthetic form in our mind.

Although I struggled a lot to figure out the metaphors with some help from Asutosh and Armeen since they don’t know the language or can relate to the script.

For me this exercise was tough, for these reasons:

1. Whenever I used to think of a letter, I used to think of the image that is usually associated with the sound. For instance, in Engish, When I say ‘A’ you will think of apple. Now that is based on the sound. But can we think of an image with the form?
2. I am exposed to script too much. And that is in a way, sometimes works against. You are so familiar to forms you treat them as text and you try to change the structure accordingly. But there needs to be an objective base to build a letter set and not intution.
3. Asutosh provided this insight: “Why don’t you try gerometric forms? I mean, use circles and squares. Maybe you will be able to look at proportions mathematically, rather than going by a few hand written examples.” Suddenly everything fell in place. I knew what direction I should take.
4. Dr Bhagvat suggests that categorization of the letterforms can also be done according to: (i) The size (ii) Simplicity (iii) Motion/stroke/angles (iv) Endings/flourishes (v) Anatomy
5. Dr Bhagvat has also studied characteristics of Devanagari. And his analysis is comprehensive, although debatable. But I also need to study the characteristics of the Kannada letterforms. It will help me when I start constructing the letterforms.

At this moment, the stuggle is to understand proportions. Not much standardization has been brought about in the Kannada script.

Categorization: Form


Two semesters back, Letter Design course was making sense to me, since I was interested in type design. But now it really makes a lot of sense. Hehe.

Tarun had given us a few handouts. One such¬†handout was ‘Analysis of the Devanagari script’. Back then I had read it. Now when I went thought the article, some very important points in the handout has given a direction for my project.

The handout especially talks about categorization of Marathi script by Dr Bhagvat. The study is interesting. The categorization he has done is phenomenal considering there is very less information available on the graphic anlysis of Devanagari script.

Now for my benefit I am fine-tuning¬†Dr Bhagvat’s¬†process of categorization. For now I will start in his footsteps, but I will change the process according to the peculiarities of the Kannada script.

As thought earlier, I am going ahead with categorization of Kannada lettes based on form. This exercise will make my life much easier when I have to design the letter set.

Please refer to the image right at the beginning of the post. I have decided to categorize the Kannada letters into three categories ‚ÄĒ Base (Parent forms), Derivatives (Derived from parent¬†forms), and Independent (Unique forms).

Project Proposal 1.0

Chukki: Download Project Proposal

Please do let me know your thoughts on the proposal. Chukki is a classroom project. Any suggestions at this time will be of great value.