I spent three hours last week watching a beautifully lit documentary—the kind where the shadows actually mean something—only to have the entire experience ruined by text crawling across the bottom of the frame like a distracted insect. People seem to think that learning how to write good subtitles is about finding the flashiest font or the most expensive plugin to automate the process, but that’s just buying your way out of a craft problem. If your text is fighting your composition or obscuring the very light you spent all day setting up, you haven’t improved the video; you’ve just added clutter to a clean frame.
I’m not here to sell you on a software subscription or a list of trendy typeface pairings. I want to talk about the rhythm of reading and how to place text so it respects the image rather than colonizing it. I’ll show you how to manage the viewer’s attention, how to time your breaks so they don’t interrupt a meaningful silence, and how to ensure your subtitles serve the story instead of distracting from the shot. This isn’t about technical perfection; it’s about making sure the words disappear into the work.
Understanding How to Write Good Subtitles
If you’re looking for a way to fix a poorly composed shot, a better sensor won’t help you. The same goes for subtitles. You can’t just dump a transcript onto the screen and call it a day. It’s about managing the viewer’s cognitive load. If the text is moving too fast or cluttering the bottom third of your frame, you aren’t helping them; you’re fighting the image you worked so hard to light. You have to respect the subtitle character per second limits—if the eye can’t track the text and the performance at the same time, the rhythm of the scene is dead.
It’s also a matter of logic, not just transcription. I’ve seen editors try to cram three different speakers into one block of text, and it’s a mess. You need to follow basic dialogue segmentation rules so the viewer knows exactly who is talking without having to guess. It’s about breaking the text at natural breath points or logical pauses. If you ignore the way people actually speak, you’ll end up with a visual stutter that pulls the audience right out of the story.
Key Things to Know
If you’re looking for a way to make text look “cinematic,” stop. You aren’t designing a title card; you’re managing a viewer’s cognitive load. The biggest mistake I see is people trying to cram too much information into a single frame. You have to respect the subtitle character per second limits. If the text is flying by faster than a human can parse the words, you haven’t provided information—you’ve just created visual noise that fights with the lighting and the performance you spent hours setting up.
You also need to decide if you’re just translating dialogue or if you’re actually following closed captioning accessibility standards. There is a massive difference between a casual social media clip and a piece of work that needs to be inclusive. This means paying attention to dialogue segmentation rules—breaking lines where a natural breath occurs, not just where the line hits a certain width. If you break a sentence in a way that feels jagged, you’re forcing the audience to work harder than they should. Keep the pacing rhythmic, keep the placement predictable, and for heaven’s sake, don’t let the text obscure the eyes of your actors.
Practical Tips and Steps
Stop treating your subtitles like an afterthought or a transcription dump. If you want them to work, you have to respect the viewer’s eye. Start with your pacing; if you’re flying through dialogue, you’re going to lose people. I always look at subtitle character per second limits to ensure the brain actually has time to process the text before the next line hits. If the text stays on screen for a fraction of a second too short, you haven’t written a subtitle; you’ve written a visual headache.
Next, think about the frame itself. You spent hours lighting that shot and positioning the talent, so don’t let a block of white text ruin the composition. Use optimal subtitle font styles—something clean, sans-serif, and perhaps with a subtle drop shadow—so it sits on the image without fighting it. And for heaven’s sake, follow basic dialogue segmentation rules. Don’t split a single thought across two different lines in a way that breaks the rhythm of the speech. It’s about flow, not just filling space.
Common Mistakes to Avoid
The biggest mistake I see is people treating text like a secondary layer that can just sit on top of the image. They pack too many words into a single frame, forcing the viewer to choose between reading the dialogue or actually watching the performance. You’ll see people ignore subtitle character per second limits, resulting in a blur of white text that flashes on screen for half a heartbeat before vanishing. If the audience has to work that hard just to keep up, you’ve already lost the emotional beat of the scene.
Another pitfall is a complete disregard for the visual composition. I’ve seen editors slap text right over a performer’s eyes or directly into the middle of a beautifully lit close-up. It’s lazy. You need to respect the frame. This also means being mindful of subtitle readability for deaf viewers; if you’re placing text over a busy, high-contrast background without a subtle drop shadow or a clean anchor, it’s unreadable. Don’t let your text fight your lighting. If the text is competing with the subject for attention, you haven’t solved the problem—you’ve just added noise to a clean shot.
Final Thoughts
At the end of the day, whether you’re working on a high-end documentary or a quick social clip, remember that text is just another element in your frame. It shouldn’t feel like an afterthought or a technical chore. If you treat your text with the same respect you give your lighting or your blocking, the viewer won’t even notice it’s there—and in this business, not being noticed is the highest compliment you can receive.
Don’t get so caught up in the bells and whistles of new software that you forget the human on the other side of the screen. Stick to the fundamentals: respect your subtitle character per second limits so people actually have time to breathe, and keep your placement consistent so they aren’t hunting for the words. It’s not about flashy graphics; it’s about clarity. If you manage the timing and the legibility correctly, you aren’t just adding text—you’re ensuring your story actually reaches the person intended to hear it. Now, stop tweaking the font and go back to looking at your composition.
Five ways to keep your text from ruining your shot
- Respect the frame. If you’ve spent twenty minutes perfecting a low-light composition or a specific color grade, don’t bury the bottom third of it under a massive, blocky subtitle. Keep the text slim and positioned where it informs the viewer without obscuring the very thing you worked so hard to light.
- Timing is everything, and I don’t mean the edit. A subtitle needs to stay on screen long enough for a human brain to actually process the words, not just glance at them. If the text flashes and vanishes like a strobe light, you aren’t helping the viewer; you’re just giving them a headache.
- Don’t transcribe every single breath. In the real world, people stutter, they trail off, and they repeat themselves. If you’re writing subtitles, you’re editing for clarity. Strip out the verbal clutter so the viewer gets the meaning of the dialogue without having to read a transcript of every “um” and “uh.”
- Break the lines where the thought breaks, not where the margin ends. There is nothing more jarring than a line of text that cuts a sentence in half awkwardly. Group your words into logical, digestible chunks that follow the natural rhythm of how people actually speak.
- Contrast is your only friend. I’ve seen beautiful footage ruined by white text on a bright sky. You don’t need fancy animations; just use a subtle drop shadow or a very slight, semi-transparent black box behind the text. It’s about legibility, not decoration.
The Bottom Line
Subtitles aren’t an afterthought or a technical checkbox; they are part of the visual composition. If they’re fighting your lighting or obscuring the subject’s eyes, you haven’t finished the job.
Respect the viewer’s cognitive load. Don’t dump a paragraph on the screen when a single, well-timed line will do. Give them enough time to read it without stalling the rhythm of the edit.
Clarity beats style every single time. A fancy font might look good in a portfolio, but if the audience is squinting to decipher it, you’ve failed the most basic rule of storytelling.
The invisible layer
Subtitles shouldn’t be a distraction; they’re just another element in the frame. If the viewer is reading the text instead of feeling the light or watching the actor’s eyes, you haven’t written subtitles—you’ve built a barrier between them and the story.
Gethin Moreau
The Final Cut
At the end of the day, good subtitling isn’t about adding a layer of text; it’s about managing the viewer’s cognitive load so they don’t miss the performance you spent all day lighting. We’ve talked about timing, the necessity of clean placement, and the cardinal sin of letting text fight your composition for dominance. If you follow the basics—keeping your lines concise, respecting the rhythm of the speech, and ensuring the contrast is high enough to read without a squint—you’ve already done more than most. Remember, the goal is to make the text invisible. If the audience is thinking about the font rather than the actor’s eyes, you’ve failed the frame.
I’ve sat in edit suites where a brilliant shot was completely undermined by sloppy, cluttered captions that felt like an afterthought. It’s easy to treat subtitles as a checkbox for accessibility, but treat them as a part of your storytelling instead. Don’t let the technicalities rob your work of its soul. Whether you’re working on a high-end feature or a quick documentary short, approach your text with the same intentionality you bring to your lighting setup. Focus on the pacing and the clarity, and I promise you’ll find that the work carries much more weight when the words simply get out of the way.
