Add to Collection
Tools Used


Design a Roman/Gurmukhi/Punjabi font using ASCII/Unicode that looks as though it is written in Tibetan, Uchen script. Suitable for Tee-shirts and… Read More
Design a Roman/Gurmukhi/Punjabi font using ASCII/Unicode that looks as though it is written in Tibetan, Uchen script. Suitable for Tee-shirts and so on. Read Less
In addition to providing a medium for quantities of information in the form of body text - such as what you are reading now - text also has an artistic role where being able to read the text quickly is not quite so important, so a few words or names or pieces of text that signify enough for the inquisitive mind to make the effort to take a while to read - essentially 'display text' - that is the role of this font to start with but as you get used to it, it is actually quite easy to read and can take on the role of body text in a limited environment, say a page, a few hundred words..
In addition to providing a medium for quantities of information in the form of body text - such as what you are reading now - text also has an artistic role where being able to read the text quickly is not quite so important, so a few words or names or pieces of text that signify enough for the inquisitive mind to make the effort to take a while to read it - essentially 'display text'.
Devanagari has been popular in this role for a while, as has the Tibetan Uchan script - an example of which you can see above. Uchan, like Devanagari and Gurmukhi, is an abugida, with each syllable taking up one consonant block position, with vowel symbols added where the vowel sound is not the default 'a' (like the last 'a' in 'panorama'). In this way, it works very well as a candidate for transliterating text that could otherwise appear in Devanagari, Gurmukhi, Gujarati and so on.
You can see in the image that Uchan has symbols - with a top line that is often semi-continuous - that represent consonants and occasionally, there is a second consonant underneath - this is a conjunct (such as the 'pr' in 'print') and is the same as how paer characters appear in Gurmukhi. In fact, like the paer characters, 'y', 'r' and 'w' change their shape to a simplified version - the second character you can see being a consonant that starts with an 's' sound, then has a 'g' with an 'r', like 'sgr' in 'misgraft'.
Additionally, the vowels are added to the top and to the bottom of the consonant blocks whereas in Gurmukhi (and Devanagari and Gujarati) exceptions are made for ਇ and ਈ and their dependent forms - this is preserved in the Tsheg font so that it is easier to read.
You will also notice that the dande lines are roughly double the height of most of the characters, so two-consonant conjuncts are fairly commonplace. In Gurmukhi, this is also the case when it is used for Punjabi.
Further, like Punjabi, double dande is used for end of sentences and a single dande is sufficiently similar to a comma in use that we can use it as such. In Uchan, the dande is called a 'Shad' and the double dande is called the 'Nyis Shad'.
Propenultimately, instead of spaces, there are little dots - this dot is called a 'Tsheg' (pronounced 'cheg'). The dots appear where a space would be between words but you only get them one at a time.
Penultimately, you will notice is that the overall style is one that slopes down to the right.
Finally, it would be good if you could take a piece of text that has already been typed out in Punjabi and just change the font with the font doing the hard work of making sure there there are no double Tshegs, that lines don't start in the middle of words and so on and only with a minimum of effort, make it conform to the finer details of the font - people seem to be in the habit of never using double dandes at the end of sentences (you will need to add them) and worse, using full stops instead (whilst these can work all right, they can produce the wrong result under certain circumstances - see below).
The image shows, laid out in a consistent way, the parts of the alphabets used by Punjabi in: Devanagari; Tibetan Uchan; Gurmukhi; and then, the Gurmukhi Tsheg font.
Highlighted in Devanagari and Gurmukhi are the glyphs from each that are similar to the Tibetan Uchan (we'll stay with the Gurmukhi names so that we know what we are talking about).
Note that in Tibetan, there are no long vowels. In order to make long vowels, a small symbol is added (you can see this under the equivalent of an aunkard ( as in 'ཨུ' to 'ཨཱུ'), or, the vowel sign is doubled up ('ཨེ' to 'ཨཻ' for example).
Further, there were no retroflex consonants so, in order to accommodate Sanskrit - which uses them - an additional line was added which is the line that corresponds to 'Taenka' ('टठडढण' ie 'ཊཋཌཌྷཎ'). The designs for these are simply the dental line that follows (ie 'ཏཐདདྷན'), written left-right inverted. Remember that the dentals came first and were adapted to accommodate the palletals.
Also note that in the Uchan conjunct 's' - 'w' ('སྭ') the 'w' changes in a way that is similar to Gurmukhi ('ਸ੍ਵ'), ie, the paer-form changes shape as in 'y-r-w-h' in Gurmukhi. In Uchan, similar things happen with the 'r' and 'y' as well, ie 'སྲ' ('sr') and 'སྱ' ('sy').
Also the Uchan 'dh' geminate in 'Adhak' ('དྡྷ') similar to Devanagari ('द्ध'), ie, it is explicit. This is done with Punjabi by locating the Adhaks, creating the conjunct with them, ie 'ਕ' with 'ਕ' as in 'ਹੱਕ' or 'ਕ' with 'ਖ' as in 'ਰੱਖਦੇ'. See below...
So, we want a font that is:
    # able to take unaltered Gurmukhi Unicode text and 'transform it' without any editing, only needing the slightest editing of human-introduced nonconformities such as double-dande and so on;
    # readable as Gurmukhi;
    # has the line running across the top as in Gurmukhi;
    # has a single Tsheg after each word instead of a space;
    # will word-wrap on a word processor at word joins (ie tshegs), looking at them in the same way as spaces in the original Punjabi;
    # has a general sloping-down-to-the-right feel about it as in Uchan;
    # has half-height glyphs so that conjuncts can be built (the paer 'y-r-w-h' versions being used by Punjabi simply accommodated or built in the same way);
    # has explicit geminates (adhaks) so that they can be used if you want to;
    # handles double-dandes intelligently; and,
    # provides start-of-paragraph/ start-of-line decorative character (variants of yig mgo tsheg shad ma) in varying widths (༅། ། ༄། ། ༄༅། ། and so on - there are many designs).
    # has at least one version of Latin text so that it is readable by Latin-only readers;
    # provides useful alternatives so that it can be used by image editing/ text editing programs that do not have the ability to use the 'liga' (ligature) facility in TTF/OTF fonts; and
    # provides ways of working around some of the more curious ways that some word processors work.
One of the more obvious features of Uchen and therefore the Tsheg font, is that the letters are written in the top half of the space on the line. Below that is a space that in just Punjabi, would be used for aunkard, dulaunkard and ਸੁ, ਸੂ the paer characters such as ਪ੍ਰ, ਲ੍ਹ, ਸ੍ਵ and ਧ੍ਯ. In the top example in the image on the right, you can see how, in the word 'book', the space under the 'b' is used for the aunkard.
As you can see, you can just add letters to make increasingly complex conjuncts - so long as the font has them in it.
Below is a table of the conjuncts that are in the font...
There are limits of legibility if cramming the letters between the top line and the bottom of the text - normally, two will fit on nicely and three is pushing it but there are some four-letter conjuncts included in the font. However, it is a display font so artistic value has a higher weighting than legibility because reading speed is not an issue here.
In Uchan, the rough equivalent of a comma is represented by the equivalent of a single dande. Commas are used in Punjabi to represent short pauses in the same way that they are in English - the mark being the same as well.
In Uchan, the single dande/comma replaces the Tsheg/dot that is used as a word-space. In the Tsheg font, the amount that the cursor moves on with the single dande is a space short of its width - the single space being the next character thus taking the cursor to the end of the single dande.
The advantages of doing this are:
    # the word processor sees the space after the comma and uses that as a break so that if a new line is needed, it makes it after the dande - you never start a line with a dande;
    # the cursor takes existing text with correct single spaces after commas and reproduces them correctly on the screen - it looks correct; and,
    # the font uses contextual substitutions so that it does not draw a Tsheg/dot if the preceding character is a dande so you don't end up with a lot of dots around for no apparent reason.
Full stops are a different matter however.
Some people will use a full stop instead of a dande and double dandes are hardly ever used - the normal keyboard layout doesn't have one - in Punjabi text. This is wrong and causes some editing - remember though, that this is a display(/short length body) font so the length of text you are going to have to do this with is only going to be small any way.
Whilst you can leave full stops and double-dandes as they are - they will appear the same - they will not be processed by the font in the same way if there is an exception to there use. For correct processing, you need to have two single-dandes at the end of a sentence - doing this gives you an artistic choice.
This is the rule and where it comes from. In Uchan, you used a double dande at the end of a sentence (ie, what looks like dande, space, dande) unless you have a 'k' ('ཀ')or a 'g' ('ག'), in such cases, you use a single dande (ie, what looks like space, dande because, in effect, the final letter provides the first dande as part of the letter itself). You can add vowels to this and it is all right but if you make a conjunct out of it, you go back to using a double dande. For example: 'ཀ །' 'ཀུ །' 'ཀྦ། །'.
From this, we can take it that if the last thing at the end of a sentence is a full-height straight line like a dande, we miss the first one out.
Above, you can see what happens to the letters 'k' ('ਕ' - no such line so double-dande default) and 'g' ('ਗ' line exists so single-dande default) when we add various vowels and so on to them.
Note that in Uchan, the vowels and nasals appear above and below whereas in Gurmukhi, some things appear to the sides as well thus a bihari can modify the 'k' like so 'ਕੀ'.
Again, this is done automatically for you by the font.
An advantage of this is that the font just lets the computer program see letters and punctuation followed by a space so it knows where to make a line break.
On the screen, we see a double dande as a dande followed by a space followed by a dande.
As a result, the space that seems to be between the two dandes will never be broken up across a line boundary and, you will not end up with a dande at the beginning of a new line where it is joined to the end of a word on the previous line.
There is a great deal more to this font - it has only been touched upon lightly here. You can see more about the font and download it for yourself to use for free - like all of my fonts -  from the link here.
Of course, you can always use the font to write your name - either in Roman text...
... or in Gurmukhi ...
... (click here for a page with lots of Gurmukhi names on it using this font) and then make a tee-shirt out of it, using the on-line printing facilities that are dotted around the planet ...
...  or go to a Print-On-Demand site for tee-shirts, mugs etc, made with this font, of names, UDHR, phrases, and so on by clicking here.
This was a very interesting font to design for a number of reasons (in no particular order of importance): the technicalities of programming it so that it would behave correctly in word processors and image editors; designing the glyphs so that they were legible Punjabi that could be ready by pretty much anybody (who can read Punjabi); designing two sets of ASCII Roman/Latin glyphs in the same style for those who can't; and, making it look right.