Monday, April 19, 2021

Ϯⲥⲁϫⲓ : The Coptic Language Text-To-Speech Seed

http://coptic-tts.web.appBeen a long while since I’ve written anything here. In my defense, I've been changing jobs, getting married and relocating. All of these definitely took some toll on my spare time 😬.

Enough excuses, the latest project that I'm undertaking for fun is to make some sort of a TTS for the coptic language. I will definitely keep updating this article about its progress. 

For those who don't know, Coptic Language was the old language of Egypt before the Arab Settlement in the 7th century. Coptic’s bohairic dialect is still being used in our beloved Egyptian orthodox ( Coptic ) churches as the main liturgical language. The word Coptic is sometimes even used to name its members, so sometimes we're referred to as "Copts", though we just prefer to be called Christians.  

At the very start of 2021, I got contacted by Fr. Arsany —who's an expert on the subject matter — to add a TTS functionality to his team's digital Coptic-Arabic dictionary : Naqlun which is named after the Naqlun monastery in fayium. With all that history , I really thought this would've been implemented fully by now. To my shock and surprise, it wasn't !! . 

If we’re to create synthesis for a language, we’ll need to understand its phonology and have some high quality recordings for it. We don’t have the later unfortunately for the time being but we have most rules for the bohairic dialect that were passed generation to generation in our church. 

As we had to start somewhere, we thought to use the phonology we know to create a coptic-IPA map that will enable us to then pass the resulting IPA to an online provider to read. The idea was also confirmed when we contacted Prof. Amir Zeldes — who heavily studied Coptic and created the online Coptic dictionary. Prof. Zeldes also believed the idea can work.

As of the 19th of April 2021 , we’ve prototyped this Coptic-IPA mapping and published it on . It’s currently in review by linguists, and we’re on the way to open-source the project and publish as an npm package. We can also port this part to different platforms as needed.

All of our team is really excited and we really hope we’ll reach bigger and more elaborate milestones.