Some months ago or maybe even longer, YouTube released a feature for the community to contribute subtitling videos or translating previous subtitles to different languages. As used as I am to deal with the frustration of learning a new language, I welcomed this idea with open arms. This feature solves many problems like allowing more videos to be subtitled for deaf people, translating subtitles for people that don’t know a language (making the video more available for a complete new audience), helping people to get used to the sonority of languages that they are learning while understanding them completely or letting people to be more involved with their favourite YouTubers.
As it can be seen in the video, it’s pretty easy to use this new feature. The tool for adding subtitles is just a timeline with boxes and the video. Pretty simple, right? and if you want to translate some subtitles, you won’t probably need anything else, specially since you can add automatic translation as a base.
If you are like me, you’ll find a little bit annoying that they decided to use shift+arrows to go backward or forward in the video, which are the same controls if you want to select text with the keyboard. Also, with the English pronoun “I” in uppercase, it’s very easy to play the video accidentally if you aren’t quick enough to release the shift before pressing the space bar.
Those are minor complains, it would be great if they were solved but you can get used to them. But what takes really longer is the transcription of an audio because every block of text has to be adjusted manually to the beginning and the ending of each part of the audio, making it a very tedious process.
And here is where there can be a lot of improvement. YouTube currently offers an automatic transcription of the audio that, leaving aside how good or bad it is, it recognises pretty well when someone starts and finishes to talk. In the worse case, it can save you some time placing the boxes with text. In the best scenario, it would be as easy as fixing some automatic transcriptions.
With this idea in mind, and knowing that the tool lets you upload a file with the subtitles, I tried to find a way to export the generated subtitles into a file, to see how much time I could save. A quick search pointed me to ccsubs.com, a page that fulfils this purpose.
After a few seconds I got the file. It did its job extracting the autogenerated subtitles but sadly the timing were wrong in most of the texts. I tried to make some adjustment to those times using regex and after that, I uploaded the file.
It was pretty exciting to see how all the boxes were almost placed in their right position. I started to fix some of them manually using the tool provided by YouTube, and it was here where its simplicity failed me: When some boxes needed to be bigger, I had to push many of the next boxes one by one. Or on the contrary, if the box was smaller, I had to pull all the following boxes. It would have been great to have the possibility to push/pull all the boxes from one specific time a certain amount of time as a unique action.
Another problem was that some subtitles were very short sentences that could be mixed with the previous or the next box. YouTube doesn’t offer you any facility to do this, you have to copy the content of a box, paste it on another and then remove the spare one. An option for joining boxes would be great.
It definitely saved me some time, but there are a lot of room for improvement. And, since people are contributing using this feature for free, wouldn’t it be smart to provide them the best tool?
For those of you in a hurry, here’s the TL;DR
YouTube feature for subtitling can be improved with:
- possibility to add automatic transcriptions as subtitle base
- possibility to join blocks of texts
- possibility to push/pull all the boxes of text from a certain point in the time line
