WHY DO SUBTITLES USUALLY NOT MATCH THE AUDIO?

If you are used to watching your films and TV shows in the original version (and you know both languages or can more or less guess one of them), you may have realised that the dialogue that you are hearing does not exactly match the subtitles that you are reading. It even happens when both audio and subtitles are in the same language.

Why does this happen? Is it a mistake on the translator’s part? Well, you may be surprised to hear – it is no mistake at all.

Subtitling is not as simple as reflecting exactly what is being said in the dialogue. There are very specific limitations on subtitles, especially in professional and commercial contexts, which mainly have to do with space constraints and human reading speed.

Television screen with Netflix logo, location unknown, date unknown. More: View public domain image source here

These are the main restrictions in subtitling:

Space: the subtitles should be in either one or two lines. If you see more than that, it is not professional subtitling. The goal is not to cover the image and for subtitles to be fast read and understood.
Position: the most common position for the subtitles is centred at the bottom of the screen. The right or left alignment is usually used when there are text elements on the screen that the subtitle could hide or other elements that could make the subtitle difficult to read.
Segmentation: subtitles should be divided following the flow of natural speech. The same happens with the configuration of the two lines – it must be done naturally, dividing the subtitles into syntagms for better comprehension by the viewers.
Characters per line: translators usually work with 38 characters per line (including spaces), which is a widespread convention for cinema and TV, although it can range from 32 to 40 characters per line.
Time on screen: subtitles must be there long enough for the viewers to read them but they should go soon enough for them not to start reading them again. We usually have the subtitle on screen for at least one second, but for no more than five seconds. This is calculated by knowing that the general audience has an average reading speed of 15 characters per second and 75 characters in five seconds.

These space and time constraints are implemented in professional subtitling so that the viewers can read the subtitles at ease, but they also mean that translators will sometimes not include ALL the information. Of course, the most important details will appear for the audience to understand what is going on in the video they are watching.

Interestingly, TikTok and Instagram Reels are not following these conventions at all. You may have seen videos where every single word appears alone on the screen, one after another. Humans’ reading speed is increasing, especially Gen X’s. Is this the end of subtitling conventions the way we know them?