← Back to Blog

How to Add Subtitles to Long Videos Without Crashes

By Picute Team···4 min read

Where long videos go wrong

If you've tried to subtitle a two-hour podcast or a three-hour lecture, you've probably hit one of these:

  • Length caps. Many caption tools are designed around short social clips and set a maximum video length per file. Check the limit on the plan you're considering, not just the free tier.
  • Monthly minute allowances. One three-hour recording is 180 minutes, which can be most of a month on a small plan.
  • File size limits. A long HD recording is a large file, and upload limits can block it even when the minutes would fit.
  • Sluggish editors. An editor that loads the whole video and transcript at once can slow down on a very long timeline.

You can also do it by hand: write or get an SRT, then burn it in with ffmpeg:

ffmpeg -i input.mp4 -vf subtitles=subs.srt -c:a copy output.mp4

That works on any length, but you still need the SRT, and making one for three hours of speech is the slow part.

How Picute handles long recordings

  • No length cap and no file size limit. A 30-second clip and a three-hour podcast go through the same steps.
  • One speech engine. Every job is transcribed by Soniox, in any of 60 languages.
  • Billing by the minute. 1 credit is 1 minute of transcription, rounded up per job, so a three-hour recording uses 180 credits. Starter includes 300 credits a month ($15), Pro 900 ($29) and Premium 1,500 ($49).
  • Rendering costs no credits. Burning the subtitles into an MP4 doesn't use credits on any plan, so re-rendering after an edit costs nothing extra. On the free plan, video exports carry a Picute watermark.

Step by step: subtitle a 3-hour podcast

  1. Sign in at picute.net. New accounts get 10 credits (10 minutes) once, enough to try a short excerpt first.
  2. Upload the file, pick it from Google Drive, or paste the YouTube link (up to 10 links at once).
  3. Choose the spoken language. Turn on speaker labels if more than one person talks.
  4. Review the transcript in the editor. Each word highlights as it plays, which makes a wrong word easy to spot. Start with names, technical terms and any stretch where the audio changes.
  5. Optionally remove silences and filler words.
  6. Export an SRT or VTT file, or choose a caption style preset (Karaoke, Popline, Simple and others) and export an MP4 with the subtitles burned in. Finish your review before rendering, so you only wait for one render of a three-hour video.

When to use Picute and when to use something else

Picute fits when:

  • Your videos run longer than a typical social clip.
  • You want subtitles burned into the video, not just an SRT file.
  • You need subtitles in several languages (translation is built in).
  • You want caption style presets instead of styling every line by hand.

Look elsewhere when:

  • You need full video editing (cuts, transitions, effects): try CapCut or Premiere.
  • You want to edit video by editing the transcript: try Descript.
  • You only transcribe something short once in a while: a pay-per-minute service may cost less.
  • You need a human-checked transcript without doing the review yourself: use a human transcription service.

Related reading

Transcribe a long recording with PicuteNo length cap · unlimited file size · 1 credit per minute of transcription

Frequently asked questions

Is a 3-hour file transcribed less accurately than a 10-minute clip?

Length itself usually isn't the problem. Changes within the recording are: a microphone that gets bumped at 1:47:00, a guest joining remotely on a worse connection, a stretch with noise in the background. Those sections are where to look first in review. A bad ten minutes is sometimes quicker to re-record than to correct by hand.

Should I split a 3-hour file into smaller chunks before uploading?

Not if your tool accepts the full length; splitting only adds stitching work. If you're stuck with a length cap, split at a pause between sentences, never mid-sentence. When you rejoin the subtitle files, shift each later part by the exact time where it was cut using the subtitle shifter, then combine them in Merge SRT with "Second file continues after the first" unticked. That option is on by default and starts the second file right after the first file's last subtitle, which is only correct if the cut fell exactly there.

What about recordings with several speakers, like interviews or panels?

Turn on the speaker labels option. The labels are generic, so rename them in review, and expect to fix a few where people talk over each other. The more voices on a panel, the more labels you'll check. If you can, record each speaker on their own microphone so voices stay distinct.