Over the years of work, I have noticed an alarming trend — many otherwise capable subtitlers will often follow their client’s guidelines too strictly, almost dogmatically, without a real understanding of why those guidelines are the way they are and not knowing when to deviate from them to ensure the audience’s viewing comfort. This leads to countless subtitling errors, because no matter how robust and well-thought-out a style guide is, there will always be gaps in it, and so there’ll be moments when you need to make a judgement call based on your expertise rather than a written prescription.
In this new article series, I’d like to stress the importance of a thoughtful, intelligent approach to subtitling and to highlight some of those gaps, starting with arguably the biggest one.
At the same time, when given creative freedom, the most skilled and experienced subtitlers don't obsess over that number as much everyone else seems to believe, because they know just how unreliable it can be, for multiple reasons.
First of all, as I wrote in one of my previous articles, CPS and WPM consider only the volume of subtitle text but not its other properties, such as complexity or format. Unfamiliar words, tricky syntax, puzzling dialogue, italics and some other things will slow down your reading, and these two metrics simply do not reflect that.
Max Deryagin’s Subtitling Studio
Dealing With High Reading Speed
Every so often, people contact me to ask a subtitling-related question. One of the things they ask the most is, “What can I do to keep my subtitles below the reading speed limit?”. This question might seem rather trivial to a professional, but for an AVT student, a fledgling subtitler, or a translator specialized in another field and looking to go audiovisual, it can be quite the opposite. So, to answer it, I’ve decided to write this article.
Without further ado, let’s dive right into it.
Imagine you have a subtitle and its reading speed is way out there. For example, 20 characters per second, which would be considered too much in interlingual subtitling for almost any language.
What can you do to reduce that speed to an acceptable level? Well, three things: change your subtitle’s timing, segmentation or text. Let’s discuss them one by one.
This is what you do first when encountering a fast sub — you try to extend its in-time or out-time (or both) to trade synchronicity for reading speed.
How far you’re allowed to extend them depends on the guidelines you’re following, which can be national, corporate or other. Usually, the out-time extension limit will be between a half-second and two seconds, while for in-time it’s only a few frames, since a subtitle appearing too early harms the viewing experience much more than a subtitle lingering on the screen. One exception to this is SDH, as you need extra time for reading name labels and other non-speech information.
If you decide to extend your subtitle’s in-time, make sure you don’t pick a wrong moment for it to pop on. As you might know, when a sub appears, the viewer immediately shifts their gaze to it, and if something important happens on the screen at that exact instant, they might miss that.
Now, obviously, not all subtitles are isolated like the one above — more often you’ll have a chain of subs covering continuous dialogue. For subtitles like that, extending their timing will require adjusting that of the ones that go before and after. So, you’ll need to mind their reading speed as well.
Another thing to pay attention to is shot changes — they restrict how far you can extend.
In some situations, you will want to cross a shot change to reduce reading speed. The exact way to do that will depend on your guidelines — the Netflix ones, for instance, require that you cross by at least 12 frames.
One interesting scenario here is subtitles like the one below, where dialogue stops right before a shot change. Is it okay to cross it to help your sub’s reading speed? Generally, no, but if nothing else works and if the shot change doesn’t introduce a new scene or setting, it can be an acceptable solution in some cases.
If the dialogue segment in the above scenario is very short, there’s an even bigger dilemma: what do you do to reach the minimum subtitle duration? Do you extend the in-time far back or do you cross? Neither feels right, so use your best judgement to determine which one is the lesser of two evils.
When it comes to subtitling plot-pertinent on-screen text, there’s another restriction: in general, you want your subs’ in- and out-times to coincide with the moments when that text enters and leaves the image.
Not always, though. The following three scenarios call for another approach:
If the text stays on the screen for too long — maybe even longer than the maximum subtitle duration — you’ll have to cut your subtitle short.
If the text overlaps with dialogue and both are very important to the plot, you’ll have to try to squeeze both of them in, so, again, the sub will be shorter.
If the subtitle’s reading speed remains too high after you’ve exhausted every other trick in your book, you might have to extend its out-time past the point of the text’s leaving the image.
Also, on-screen texts can appear and disappear in special ways, e.g. word by word, via fading in and out, rolling on/off, going in/out of camera lens focus, etc. These can be tricky to time — you will have to trust your sense of visual design and timing to make the right choice.
If you can’t use the first method of reducing reading speed, or if it isn’t enough, you can try resegmenting your subtitles. The main technique here is to merge your fast sub with a neighboring one.
This does four things:
Adds the time in-between the subs (even if it’s only a minimum gap);
Averages out their reading speeds.
Makes it so that the viewer needs to move their eyes to the subtitle area only once rather than twice.
Gives you more flexibility as far as text condensation (more on that later).
This tactic works particularly well for languages like Danish, Swedish and Norwegian, because you’ll often get long two-liners which Scandinavians prefer.
The second segmentation technique is similar to the first one, but it’s for when you can’t merge your subs because they have too much text — you’d exceed the maximum number of lines. What you can do instead is try to split the neighboring sub in two and then merge one of its fragments with the fast subtitle.
Now, in some cases merging might be not the best idea. Here are a few of them:
When it creates a two-liner with an awkward line break which you can’t fix due to the line length limit.
When it turns your sub into a spoiler, revealing what’s going to happen next or a punchline too early.
When the gap between the subs is too large, e.g. two seconds or more — it’ll break the flow.
When your two subtitles have different (and incompatible) formatting or positioning.
When there’s a shot change between the subs and you don’t want to cross it.
When one of the subs translates dialogue and the other one — on-screen text.
Also, when they translate two very different on-screen texts.
When merging creates a huge two-liner that affects the image too much — disrupts the shot composition or obscures an important visual element like a face or a plot-pertinent object.
These can be a deal-breaker — or not! — so use your discretion to decide if merging makes sense in each particular case. And remember to split only at the most logical points — linguistically and otherwise.
The first two methods can get you only so far. When they’re not enough, you will have to use the third one — changing the text. More specifically, what you want to do is make it more concise, to condense it. The idea is rather simple here: the less text a subtitle has, the lower its reading speed.
So, how do you condense subtitle text? Here’s what you can do:
His life spanned 82 years → He lived 82 years
Amanda has got lots of money → Amanda is rich
Use shorter synonyms:
enormous → huge
to purchase → to buy
approximately → roughly
they would → they’d
television → TV
Doctor → Dr.
Use acronyms and initialisms:
as soon as possible → ASAP
United States → U.S.
Spell out big round numbers:
1,000,000,000 → a billion
8,500,000,000,000 → 8.5 trillion
Don’t spell out other numbers:
twenty-five → 25
point zero three → 0.03
one-twentieth → 1/20
Use signs and symbols:
98 percent → 98%
200 dollars → $200
Platform number 15 → Platform #15
Make up a word or give an existing one a new use:
I’ll kill all the wasps in here → I’ll dewasp this place
We drove to school in a Prius → We Prius’d to school
Replace longer nouns with pronouns:
Jonathan went home → He went home
The police didn’t notice → They didn’t notice
Use active sentences instead of passive ones:
That was where birds nested → Birds nested there
It was decided by Jane → Jane decided that
Split a sentence into two:
My sister is tall, whereas I am short.
My sister is tall. I am short.
Omit filler words:
You’re kinda cute → You’re cute
It was, like, really cool → It was really cool
No no no, don’t go there → No, don’t go there
Come on, Charlie! Come on! → Come on, Charlie!
Omit hesitations, false starts, tag questions:
But... can we... Can we do it? → But can we do it?
These guys... they never listen → These guys never listen
This is great, isn’t it? → This is great!
Omit dialogue parts that the viewer will understand from the soundtrack or the visuals:
Who drew this gigantic arse?! → Who drew this?!
(teacher pointing at a chalk drawing on the blackboard)
And so on. There are actually more techniques, but I can’t list all of them, because there are too many, some are language-specific, and, frankly, text condensation is too creative and complex a process for me to be able to shoehorn it into a rigid set of rules. But if you’d like to learn more, my advice would be to pick a book titled Subtitling: Concepts and Practices by Jorge Díaz-Cintas and Aline Remael, which will come out in 2020 and have lots of techniques, examples and exercises.
A couple of things to note about this third method of reducing reading speed:
It doesn’t work well for texts that you can’t change much — famous quotes, well-known verses, dialogue in recaps and flashbacks, terms from your client’s glossary, etc.
If the speakers talk way too fast or simultaneously, you’ll have to omit a lot and reformulate heavily. You might feel bad editing out so much, but remember: drastic times call for drastic measures.
As you become more experienced in subtitling, you’ll find it easier to write concisely. It’s an integral part of subtitling where your skill and creativity really shine — and one that every pro has mastered.
Combining the Methods
The methods I described above don’t have to be used separately — you can mix-and-match them however you want. For instance, you can merge two longer subs and then condense the text down to two lines:
Or you can include a nearby interjection (which you’d omit otherwise) and extend your sub’s timing to it:
Or you can merge three subtitles into one, or delete your fast sub altogether if it consists of omittable stuff, or do a million other things — there are many possibilities, so if you get creative, you’ll always find a way to make your subs easier to read.
So this is how you deal with high reading speed — you extend the timing of your subs, condense their text and resegment them. And when you’ve hit the speed required by your guidelines, you can go even further and take into account some extra things I mentioned in one of my previous articles — to make sure that your viewers will be able to read your subs comfortably while having enough time for watching the action on the screen and for listening to the audio.
All right, this is it. If you have any questions or remarks, or if you’d like to share an interesting technique, feel free to leave a comment below.