> ## Knowledge Base Index > Fetch the complete knowledge base index at: https://help.comicink.ai/sitemap.xml > Use this file to discover available pages before exploring further. > Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash). # Background music and voices A Story Video carries two kinds of sound: the **spoken audio** inside each scene — the narrator and your characters — and an optional **background music track** that plays underneath the whole video. --- ## Adding background music Story Videos include a built-in **music library**, so you do not need to source or upload a track yourself. Search the library by **title**, **artist** or **tag** — tags let you search by mood or genre rather than by name, which is usually the faster way in. Preview what you find, then add the one you want. A video carries **one** background track at a time. Adding a different track replaces the current one. ### Volume, looping and length Once a track is attached you can: - **Adjust its volume**, so the music sits under the narration instead of competing with it. - Leave the length alone. The track **loops** if it is shorter than your video, and is **automatically trimmed to the video's content length** if it is longer — so it always ends when the video ends, with no manual trimming and no silence at the tail. If you later change a scene's **speed**, the video's total length changes with it. Music and captions are **re-timed automatically** to match, so you do not need to re-trim or re-sync anything after a speed change. --- ## The voices in a scene The spoken audio in a scene comes from the script: the **narration** line is read by the narrator voice, and each **dialogue** line is spoken in the voice of the character who says it. Both are produced when the scene is rendered. Because the spoken audio has to fit inside the scene's duration, there is a per-scene budget of **19 spoken words** across narration and all dialogue combined. Going over shows an amber warning — nothing is silently trimmed, but you should shorten the lines or the delivery will be rushed. --- ## Fixing a voice with Fix Audio If a rendered scene's audio is wrong — a mispronunciation, a garbled word, a line you want changed after the fact — you do not have to re-render the whole scene. Each scene has a **Fix Audio** action that repairs the spoken track in place. What it does: it isolates the original audio from the clip, separates the individual speakers, clones their voices, and lays a corrected track over the muted clip. The picture is untouched — only the sound is rebuilt. Inside the Fix Audio window the script is editable **line by line**, with each line labelled either **Narrator** or by the **character name** who speaks it. Edit a line and the audio regenerates from your edit, in that speaker's cloned voice. For the full walkthrough, including when to reach for it and what it costs, see [Extending scenes and fixing audio](https://help.comicink.ai/en/article/ai-extend-and-fix-audio-127kq64/). --- ## Next steps - [Caption editor for Story Videos](https://help.comicink.ai/en/article/caption-editor-for-story-videos-zph6az/) — putting the spoken words on screen. - [Rendering scenes](https://help.comicink.ai/en/article/rendering-scenes-per-scene-vs-generate-all-z30txn/) — how scene audio is produced in the first place.