Troubleshooting

Auto captions getting words wrong? Fix the cause, then the words.

Most wrong words in automatic subtitles come from four places: a speech model too small for the audio, the wrong spoken language set, names and jargon the model has never heard, and silence or music it fills with invented text. Fix those before you start correcting, in that order — a bigger model, the right language, a list of the names, clean audio — and the corrections left to do by hand shrink to a few lines. In suubbs the model, the language, the names list and voice detection are all in ⋯ → Transcription settings, and they only help if you set them before you press Generate.

Get the Mac app — $9.99 → Try it free in your browser
What is going on

Speech recognition writes the likeliest sentence, not the one you said.

Every common kind of mistake follows from that.

suubbs uses Whisper, OpenAI's open-source speech model. Whisper does not look words up; it predicts the text most likely to go with the sound, one token at a time. When the sound is clear and the words are ordinary, the likeliest text is what was said. When it is not, the model still produces a fluent sentence — which is why wrong captions so often read as plausible English that nobody said, rather than as obvious garbage.

That gives four distinct failures, and they need different fixes:

The fixes, in order

Five settings to check before you press Generate.

All of them are in ⋯ → Transcription settings.

  1. Use the largest model your machine can run. In the Mac app the choice is Base (English), Small (English) and Large v3 Turbo (quantised), which is the default and the one the app describes as "best accuracy for the size". It is also the only multilingual model in the Mac app, so for anything not in English it is the one to use. In the browser studio the default is Base (about 78 MB) for a quick first load; Small (about 250 MB) is the step up when Base misses words, and Large v3 Turbo (about 800 MB) is the most accurate a browser can run, if the machine has the GPU memory for it.
  2. Set the spoken language. The Mac app defaults to Detect automatically, which is right when you are unsure. The browser studio defaults to English, so if your video is in another language, change it there first — this one setting is behind a lot of "the captions are nonsense" reports. Either way there are sixteen languages to choose from.
  3. List the names and jargon (Mac app). Type them into Names and jargon, separated by commas — the field's own example is "Dr. Chen, Suubbs, Anthropic". It biases the transcript toward your spellings. It is a nudge rather than a dictionary: it makes the right spelling likelier, and it cannot make the model hear a word that is not audible.
  4. Leave voice detection on (Mac app). Silero voice detection is on by default once its small model has downloaded, and it is what stops silent stretches being transcribed as speech. The browser studio has a simpler stand-in: it skips blocks of audio that are close to silent before the model sees them.
  5. On unclear speech, try beam search (browser). The browser studio's Decoding setting offers Accurate — beam search, which weighs several candidate transcripts instead of committing to the first. It is several times slower, and the gain is largest on mumbled or distant speech.

Set these first, then correct. Pressing Regenerate transcribes the file again from scratch and replaces the whole transcript — your hand corrections and the undo history with it. Get the settings right on a first pass, and only then start fixing individual words.

The large model, on your own Mac.

[nothing uploaded]

The recording

What no setting can fix.

A model can only transcribe what reached the microphone.

If you can still re-record or re-export, these change accuracy more than any setting does:

Timing is a separate question from accuracy. On the Mac, word boundaries are aligned and then snapped onto real pauses in the waveform, measured at about 3 ms from true speech onsets; in the browser they land within roughly 60–140 ms. Either is fine for ordinary subtitles, where whole lines come and go. It matters for word-by-word captions, and the word-level timestamps page explains how it is measured.

Correcting what is left

How to fix a word without breaking the timing.

What happens to the timing depends on how you edit.

Click a line in the cue list and retype it. When you finish, suubbs lines your new words up with the old ones position by position:

For ordinary subtitles this does not show: a line still starts and ends where it did. It shows in a word-by-word highlight, where an estimated word can light up slightly early or late. The ⋯ menu tells you how many words have interpolated timing, so you know whether there is anything to check, and none of it blocks an export.

Read it through before you export

suubbs does not mark the words it was unsure of, so the reliable check is to play the video and read along. Look hardest at the places the four failures above concentrate: names and numbers, the first line after a long pause, anything over music, and the passages where people talk over each other. A line flagged as reading too fast usually needs splitting into two, or rewording — the subtitle reading speed page explains the limit. If a line was invented outright over silence, delete it.

If your captions are right but appear early or late, that is a sync problem rather than an accuracy one; fixing subtitle sync covers it.

Browser or Mac

Where the two versions differ on accuracy.

Same studio, two different speech engines underneath.

  Browser studio (free) Mac app
Default modelBaseLarge v3 Turbo (quantised)
Default languageEnglishDetect automatically
Names and jargon listNoYes
Silence handlingSkips near-silent blocksSilero voice detection
Beam search decodingOptionalNo
Word timingRoughly 60–140 msAbout 3 ms
How much it transcribesThe first 60 secondsThe whole file

The browser studio is the honest way to judge accuracy on your own footage before paying: the first 60 seconds are transcribed with the same studio, and every editing and export feature works on them. Both run on your own computer; neither uploads the video.

Questions

What people ask next.

Mostly about names.

Why do auto captions keep spelling a name wrong?

Because the model writes the likeliest text for the sound, and a rare name is less likely than common words that sound like it. In the suubbs Mac app, type the name into Names and jargon in Transcription settings before generating; it biases the transcript toward your spelling.

Can I train it on my voice or my vocabulary?

No. suubbs does not fine-tune the model. The Names and jargon list is the supported way to steer spellings, and it applies to the run you set it for.

Why are my subtitles in English when the video is not?

The spoken language is set to English — the browser studio's default — or, in the browser studio, Translate to English is switched on. Set Spoken language to the language actually spoken, or to Detect automatically in the Mac app, and generate again.

Why is there a caption where nobody is talking?

Whisper sometimes writes text into silence or music. The Mac app runs Silero voice detection to prevent it, and the browser studio skips near-silent audio; anything that still gets through can be deleted from the cue list.

Will correcting a word mess up the timing?

Only for the word you changed. Words you leave alone keep their measured timing; a replaced word gets an estimated time within its line, and adding or removing a word re-estimates the words after it in that line.

Does Regenerate keep my corrections?

No. Regenerate transcribes the file again and replaces the whole transcript, including your edits and the undo history. Change the settings first, then correct.

Is the Mac app more accurate than the free browser version?

By default, yes: it runs Large v3 Turbo where the browser starts on the much smaller Base model, and it adds the names list and voice detection. The browser can load Large v3 Turbo too, on a machine with enough GPU memory.

Does it flag the words it was unsure about?

Not currently. Read the transcript through while it plays, concentrating on names, numbers, music and crosstalk.

Keep reading

Related

Fewer words to fix.

[the large model, nothing uploaded]