Getting captions onto your video is a five-minute job. Getting them right takes a little more — here is how to do both.
Captions are no longer optional. Whether you are publishing to TikTok, YouTube, Instagram, or LinkedIn, a significant portion of your audience is watching without sound, and every viewer benefits from text that reinforces what is being said on screen. Adding captions is straightforward once you understand your options — auto-generated vs. manually written, burned-in vs. exported as a subtitle file, styled for readability vs. styled for engagement. This guide walks through each decision so you can caption your videos accurately, efficiently, and in a way that actually works on the platform you are targeting.
Use an auto-caption tool appropriate to your editing software — Premiere Pro's Speech to Text, CapCut's auto-caption, or a browser tool like Kapwing or Descript. Upload the video or audio file and let the tool generate a rough transcript synced to your timeline. For poor-quality audio, consider running the audio through a noise reduction plugin before transcribing to improve accuracy.
Go through the auto-generated text from start to finish and fix every error. Pay close attention to proper nouns, brand names, technical terms, and any word the transcription tool consistently misspells. Even one wrong word per minute adds up and looks unprofessional. In tools like Descript and Premiere, you can edit the text directly and the timing adjusts automatically.
For short-form vertical video going to TikTok, Reels, or YouTube Shorts, burn the captions into the video during export for full style control and guaranteed visibility. For long-form YouTube uploads or LinkedIn content, export an SRT file and upload it separately. If you are publishing the same content in multiple formats, you can do both from the same transcript.
Select a bold or black-weight sans-serif font — Montserrat, DM Sans, Inter, and Impact are common choices. Set the size large enough to read on a mobile screen without squinting. Add a drop shadow, outline, or background fill to ensure contrast against any background. For short-form, use word-by-word or phrase-based captions depending on your speaker's pace. Keep each caption block to one to two lines maximum.
Place captions in the middle of the frame on vertical video, clear of the bottom UI overlay and top account info areas. On horizontal video, use the lower-third area but keep at least a five-percent margin from the bottom edge. Preview the video on a mobile device before finalizing to confirm nothing is being cut off or obscured by platform chrome.
If the content and pace suit it, apply word-by-word highlighting or a color-pop emphasis to key terms. This adds visual energy and reinforces the most important ideas. Keep the animation style consistent throughout the video — mixing multiple highlight styles or colors without intention looks chaotic. Stick to one or two colors maximum, typically your brand palette.
Export the video and, before publishing, play it back on a mobile device to check that captions are visible, readable, and correctly timed end to end. If uploading an SRT to YouTube or LinkedIn, publish the video and then check the caption display in the platform's viewer. Edit any remaining timing errors directly in the platform's caption editor rather than rerendering the entire video.
Auto-captioning has improved dramatically in recent years. Tools like Adobe Premiere Pro's Speech to Text, CapCut's auto-captions, Descript, Kapwing, and the native auto-caption features inside TikTok and YouTube can generate a rough transcript in seconds and sync it to your audio with reasonable accuracy. For clean, close-mic audio with a single speaker talking at a normal pace, accuracy rates are high enough that auto-captions give you a strong starting point rather than a mess to fix.
Manual captioning — typing every line yourself, then syncing it to the video — is slower but gives you full control. It is the right choice when audio quality is poor, when multiple speakers overlap, when there is heavy background noise, or when the spoken content is technical and filled with industry terms that auto-transcription routinely misspells. Many editors use a hybrid approach: generate auto-captions to get the timing and rough text in place, then go through and correct errors by hand. This is almost always faster than typing from scratch.
Burned-in captions — also called open captions or hardcoded captions — are baked directly into the video frame during export. Every viewer sees them regardless of their device settings. This is the standard for short-form content on TikTok, Reels, and YouTube Shorts because the platforms display native subtitles inconsistently across devices, and you want no chance of captions being invisible to someone who needs them. Burned-in captions also give you full design control: font, size, color, animation, and positioning are all part of your edit, not left to the platform.
SRT files — SubRip text files — are separate text files that contain your caption text and timestamps. You upload them alongside the video rather than embedding them in the frame. Platforms like YouTube, Vimeo, LinkedIn, and Facebook support SRT uploads and display them as closed captions that viewers can toggle on or off. SRT captions are better for long-form content because they do not clutter the visual frame by default, they are accessible to screen readers, and they can be translated into other languages without re-editing the video. For a YouTube video, uploading a corrected SRT is also how you ensure the auto-generated captions are accurate without rerendering the file.
Many workflows use both: burned-in captions for short-form cuts that go straight to TikTok or Reels, and an SRT file for the long-form version uploaded to YouTube. You can generate the SRT from Premiere, Descript, or any auto-caption tool and clean it up in a plain text editor before uploading.
Readable captions require bold or black-weight sans-serif fonts, sufficient size for mobile viewing, and engineered contrast against any background. White or yellow text with a dark outline, drop shadow, or semi-transparent background fill are the most reliable combinations. Avoid thin-weight fonts, decorative scripts, and any color that blends with typical backgrounds. A caption that requires effort to read will be skipped — and once a viewer starts skipping your captions, they stop processing your audio too.
Safe zones are the margins inside the frame where content will not be cropped or obscured by platform UI elements. On vertical video for TikTok and Reels, the bottom of the frame is where the username, caption text, and action buttons sit. The top is where the following status and ads appear. Your captions should live in the middle zone — roughly between 30% and 75% of the frame height — to stay clear of both. On horizontal video for YouTube, keep captions in the lower third but clear of the very bottom edge, where platform playback controls overlap on some devices. When in doubt, keep a visible margin of at least five percent on all sides.
Animated captions — where each word highlights, pops, or fades in as the speaker says it — have become a signature of short-form editing. The karaoke or word-by-word style keeps constant visual movement on screen, which helps sustain attention during scroll-stopping content and reinforces pacing on fast-talking clips. Tools like CapCut, Premiere Pro's Essential Graphics, After Effects, and third-party plugins like AutoCaption or Mango AI can generate this style automatically from a transcript.
Word-by-word animation works best when the speech tempo is fast and the energy is high. For slower, more conversational content, full-phrase captions that display a short group of words and hold until the thought ends are easier to read and feel less chaotic. Choose the animation style based on the pace of your speaker, not just what looks active. A high-energy highlight animation on a slow, reflective voiceover will feel out of place and actually make the content harder to follow.
Color emphasis — making one key word per phrase appear in a different color — is a subtler form of animation that works well across both pacing styles. It draws the eye to the most important term in each sentence without requiring constant motion, and it gives your captions a branded, intentional look.
TikTok has a native auto-caption feature that generates captions directly in the app. They are simple in style but accurate enough for clean audio, and they display differently depending on the viewer's device and caption settings. For full style control, burn captions into the video before uploading rather than relying on TikTok's native tool. If you do use native captions, always review them in the app before publishing — the editing interface lets you correct individual words.
YouTube supports both auto-generated closed captions and uploaded SRT files. Auto-generated captions go live within a few hours of upload and are editable in YouTube Studio. For any video where accuracy matters — tutorials, interviews, educational content — upload a corrected SRT file. YouTube's captions are also indexed by search, so accurate captions improve your SEO by giving the algorithm more text to work with.
Instagram Reels supports auto-captions via a sticker in the app editor, but these are limited in style. For branded, designed captions on Reels, burn them in at the editing stage. LinkedIn auto-generates captions for uploaded videos and allows SRT upload for greater accuracy. Twitter and X support SRT files on uploaded videos but auto-caption accuracy on the platform is lower than YouTube.
Use auto-captions as a starting point and then correct them by hand. For most clean audio recordings, auto-captioning tools are accurate enough that you only need to fix a handful of errors per minute of footage. Typing captions from scratch is only worth the time when audio quality is so poor that auto-transcription produces more errors than correct words.
Burned-in captions are permanently embedded in the video frame — every viewer sees them automatically. An SRT file is a separate text file that platforms like YouTube and LinkedIn display as optional closed captions that viewers can turn on or off. Short-form content generally benefits from burned-in captions. Long-form content on platforms that support SRT upload benefits from the flexibility and SEO value of a separate subtitle file.
On vertical video for TikTok and Reels, keep captions in the center-to-upper-middle portion of the frame — roughly between 30% and 75% of the frame height. The bottom of vertical video frames is consistently covered by platform UI including usernames, follow buttons, and navigation controls. Captions placed too low will be partially or fully hidden on many devices.
Yes. YouTube's search algorithm indexes caption text, which means accurate captions expose your video to more keyword searches. Auto-generated captions also get indexed, but they often contain errors that dilute relevance. Uploading a corrected SRT file ensures that the text YouTube indexes matches what is actually said, which can meaningfully improve discoverability for tutorial and educational content.
Yes, if the video itself is the same cut. If you burn captions into the vertical video during export, that file can be uploaded to all three platforms without modification. The safe zone positioning and font size choices you make should be tested across all three before you set a template, since the UI chrome placement varies slightly between platforms. A center-positioned caption block clears the UI safely on all three.
Studio432 edits concert footage, music videos, gaming content, vlogs and short-form for creators worldwide. Send the footage, get a quote by email.
Start a project →