Subtitles with different colors for each speaker: the practical guide
By the Clipzi team · Updated
Subtitles with different colors for each speaker let a viewer know who is talking without looking for lips or labels. In an interview or a podcast clip watched on mute, that is the difference between following the argument and losing it after the second line.
The idea is simple. Doing it is less simple, because the most common subtitle format, SRT, has no official way to store color. This guide covers why it helps, which colors to pick, what each format can and cannot do, how to do it by hand in Premiere Pro, and the workflow that saves the most time: one SRT file per speaker.
Why give each speaker a color
- Interviews and podcasts: two or three people trade short lines. Without a visual cue, a quick back and forth reads as one person contradicting themselves.
- Muted viewing: on Reels, TikTok and Shorts many people watch without sound, so the subtitles carry the whole conversation. If you are cutting clips from a podcast, our guide to turning a podcast into short clips covers the rest of that workflow.
- Accessibility: deaf and hard of hearing viewers cannot use voices to tell people apart. In the UK, the BBC subtitle guidelines name color as the preferred way to identify speakers. US practice leans more on placement and name labels. Color plus position works for both.
Which colors to use for each speaker
Start from the broadcast palette: white (#FFFFFF), yellow (#FFFF00), cyan (#00FFFF) and green (#00FF00). These are the colors the BBC guidelines list for speaker identification, chosen because they stay readable on dark backgrounds. Give white to the person who talks most (usually the host), so most of the screen stays neutral.
Rules that hold up in practice:
- Keep each person's color fixed for the whole video and across every clip you cut from it.
- Avoid putting red and green on two speakers next to each other: it is the pair color blind viewers confuse most.
- Use a dark stroke or a semi-transparent box behind the text. Yellow on a bright sky disappears otherwise.
- Do not go past four colors. With more people, combine color with a name label the first time each person speaks.
Why SRT has no color (and what VTT and ASS can do)
SRT is just a number, a timecode range and text. Some tools write an unofficial font color tag inside the text, and many desktop players such as VLC respect it. Many editors and upload platforms ignore it or show the raw tag. Treat it as a bonus, never as the plan.
WebVTT (.vtt) has real speaker support. You mark lines with a voice tag (the letter v followed by the name, for example v Ana) and style them with CSS in a STYLE block, so every line from Ana shows in yellow. It works well for video on your own website. Support on social platforms and in editors is inconsistent.
ASS/SSA (.ass) is the most flexible: you define one style per speaker with its own color, font and position, and assign each line to a style. Aegisub edits it and FFmpeg can burn it into the video. Premiere Pro does not import it as captions.
So for an editor working in Premiere or DaVinci Resolve, the reliable route is not color inside one file. It is one track per speaker, each track with its own style.
How to do it by hand in Premiere Pro
Premiere gives each caption track one track style. You can select a single caption and change its fill color in the Properties panel, which is fine for 10 captions and painful for 400. Careful with tutorials that tell you to right-click a caption and pick Label: that only changes the clip color on the timeline, not the text the viewer sees.
The workable manual method is to split the dialogue by speaker:
- 1
Split the lines
Get the captions of each person into a separate caption track. Doing this inside Premiere means moving lines one by one, which is the slow part.
- 2
Style each track
Select a caption on track 1, set color and position in the Properties panel, and save it as a Track Style. Repeat on track 2 with a different color, maybe a different height.
- 3
Show them together
Premiere shows only one caption track at a time. Select all captions and run Graphics and Titles > Upgrade Caption to Graphic. They become graphics on the top video tracks, keep each track's style, and all render at once.
Subtitles with different colors for each speaker from separate SRT files
A speaker-separated SRT is a normal SRT that only contains one person's lines. The timecodes are the same as in the full transcript, and where someone else talks, the file simply has a gap. Two speakers, two files; three speakers, three files. Every editor that imports SRT can read them, which is why this beats any color tag.
In Premiere you import each file and drop it on the sequence. Each one lands on its own caption track, ready to get its own color and position as described above. The full import process, plus the encoding and sync errors that show up, is in our guide on how to import SRT into Premiere Pro.
The hard part is getting the split right. Automatic speaker detection is good but not perfect: it may split one person into two when they move away from the mic, or give a short interruption to the wrong person. Whatever tool you use, review that before exporting.
This is the job of the Subtitles screen in Clipzi's caption generator. You upload the long video, the transcript comes back with speakers detected, and you review it against the video: rename Speaker 1 to the real name, merge two speakers that are the same person, fix words or whole lines, and reassign a line to whoever said it. Then you export one SRT for the full video or one SRT per speaker (Pro and Business plans), and each speaker becomes its own caption track in Premiere Pro. The plan for podcasters explains what else comes from the same upload.
Other options
- DaVinci Resolve: subtitle styling is also set per track. Import each speaker's SRT on its own subtitle track and style each one, or convert the lines to Text+ titles if you need several visible at once.
- Dedicated subtitle editors (Aegisub for ASS, or online subtitle editors) let you assign colors line by line and burn them in. Good for a finished video, less handy when you still have to edit in Premiere.
- Short clips with one speaker on screen: one caption color per clip is enough. In the Clipzi editor you pick the color of the word-by-word captions for each clip.
Export one SRT per speaker
Upload your interview or podcast, review who said what on the Subtitles screen, and get a separate SRT for each person. Start free, no credit card.
Try Clipzi FreeNo credit card required