Playback & Karaoke
Play recordings in Kalima Studio with word-by-word highlighting, click transcript text to jump to a moment, navigate audio, and review screen recordings.
Once a session has recorded audio, the Studio turns into a full playback workspace. You can listen back to verify the transcript, follow along word by word with karaoke highlighting, click any word to jump straight to that moment, and, for desktop screen recordings, watch the video in sync with everything that was said.
The audio player
When you open a session that has a recording, the audio player appears alongside the transcript with everything you need to navigate the recording precisely.
- Transport controls: Play and pause, rewind 5 seconds, jump forward 10 seconds, and skip straight to the start or end of the recording.
- Seek bar with waveform: A waveform seek bar shows the shape of the audio. Click anywhere on it to jump to that point, with the current time and total duration shown alongside. Hover it with the mouse to see the time under the pointer.
- Chapters: Once the session has an AI summary, the seek bar is split into titled chapters and a chapter button sits next to the speed control. See Chapters below.
- Playback speed: Choose from 8 speeds, 0.25x, 0.5x, 0.75x, 1x, 1.25x, 1.5x, 1.75x, and 2x. The active non-1x speed stays highlighted.
- Volume and mute: Adjust the playback level anywhere from 0 to 100% and toggle mute with the speaker icon.
Navigating the recording
Slowing playback to 0.5x or 0.75x makes fast or unclear speech far easier to verify against the transcript, while 1.5x or 2x is ideal for skimming a long recording. On very narrow screens, the skip-to-start and skip-to-end buttons are hidden to save space, the rewind and forward buttons remain.
Lock screen, media keys, and headphones
While a recording plays, you can control it without looking at Kalima: from your phone's lock screen, the media controls of your browser or computer, and the play/pause keys on headphones and keyboards. They offer play, pause, back 5 seconds, forward 10 seconds, and scrubbing, with the session name as the title.
Chapters
Chapters split a recording into its parts, like the chapters of a long video, so you can see how a session is structured and jump straight to the part you need. They come from the AI summary: when Kalima writes a summary, it also divides the session into chapters in speaking order and gives each one a short title.
- Gaps in the seek bar: A small gap in the waveform marks where one chapter ends and the next begins. When you hover a chapter, its part of the waveform gets brighter. On phones, the progress bar is split into one piece per chapter.
- Title and time on hover: Move the mouse over the seek bar to see the title of the chapter under the pointer, together with the time at that point.
- Chapter button: Below the waveform, next to the speed control, a button shows the title of the chapter that is playing and updates as playback moves on. On very narrow screens, only its icon is shown.
- Chapter list: Click the chapter button to open a list of all chapters with their start times. The chapter that is playing is highlighted. Choose one and playback jumps to where it begins.
Jumping to a chapter
Summaries created before chapters were introduced have none. To add them, click Regenerate in the Summary tab, which uses one summary from your weekly allowance. Very short sessions get no chapters, since a few minutes of conversation rarely divide into parts.
Word-by-word karaoke highlighting
As the audio plays, Kalima highlights the word being spoken right now and dims the words still to come, like karaoke. It is the easiest way to follow exactly where you are in a recording and read along at the same time.
- The current word is boxed: The word being spoken is wrapped in a green highlight box so it stands out clearly as playback moves through the transcript.
- Past stays bright, future dims: Words already spoken remain at full brightness while upcoming words are dimmed, giving you an at-a-glance sense of progress through each segment.
When playback is paused, every word returns to full brightness, there is no dimming while the audio is stopped, so the transcript stays easy to read.
Karaoke highlighting relies on the word-level timing captured with your recording. It works on transcripts that have this timing data and also drives the synced subtitle overlay in fullscreen and picture-in-picture views. For more on the caption overlay and styling, see Subtitles & Display Options.
Click a word to jump there
Instead of scrubbing blindly, you can navigate the recording by reading. Click any word in the transcript while the audio is loaded and playback seeks straight to that word's start time.
This pairs naturally with search highlighting in the transcript editor: search for a phrase, then click the matching word to hear that part of the recording instantly.
Video playback for screen recordings
If you captured a screen recording with the Kalima desktop app, the Studio plays that video back in sync with the audio timeline. You see what was on screen alongside what was said, all on one shared timeline.
- Automatic sync. Press play and the video follows the audio automatically wherever a recording exists for that stretch of the timeline.
- Graceful gaps. If part of the timeline has audio but no screen video, the player pauses the video for that stretch and resumes automatically when video is available again.
- Self-healing. If a video segment ever fails to load, the player retries once on its own before showing anything.
Video only appears for the parts of a session where a screen recording was actually captured, so it is normal for the video to pause while the audio keeps playing through a gap. If a particular video segment still fails after the automatic retry, a permanent error is shown for that one segment only; your audio and transcript stay fully available, and other video segments are unaffected.
Common questions
Eight: 0.25x, 0.5x, 0.75x, 1x, 1.25x, 1.5x, 1.75x, and 2x. Whichever non-1x speed you pick stays highlighted so you always know what you have selected.
Yes. Click any word in the transcript while the audio is loaded and playback seeks to that word. As the audio plays, the current word is highlighted karaoke-style so you can follow along.
This happens when word-level timing is not available for the session, or the recorded audio is not loaded yet. Make sure the recording is loaded and playing; word-click seek and karaoke highlighting both depend on the recording's word timing being present.
Chapters come from the AI summary, so the session needs a summary first. A summary created before chapters were introduced has none until you regenerate it, and very short sessions get no chapters at all. Kalima also only places a chapter where it can find that chapter's opening words in the transcript, so after heavy edits to the transcript some chapters can drop out until you regenerate. On a shared session, viewers see chapters only if the summary is shared with them.
Screen-recording video only exists for the parts of the session where it was captured. Wherever there is no video for a stretch of audio, the player pauses the video and resumes it automatically once video is available again. This is expected behavior, not an error.
When playback is paused, every word shows at full brightness, the dimming of upcoming words only happens while the audio is actively playing. Press play and the karaoke highlighting resumes.
On very narrow screens those two buttons are hidden to make room for the core controls. The rewind (5 seconds) and forward (10 seconds) buttons remain available, and you can still click the waveform to jump to the start or end.