> For the complete documentation index, see [llms.txt](https://learn.firecut.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://learn.firecut.ai/features/script-guidance.md).

# Script guidance

### ❓ What is it?

Script guidance turns a raw, unedited recording into a clean first cut that follows a written script.

If you record talking-head or "self-tape" style videos (reading from a script, doing several takes, stumbling over lines and chatting between takes), you usually end up with a long clip full of retries and off-script moments. Script guidance reads your finished script, finds the parts of your footage that match it, and builds a **new sequence that contains only those parts**.

You give FireCut two things:

* **A video clip or sequence** with your speech in it
* **Your script**, as a plain-text `.txt` file

FireCut then makes a new sequence with the off-script takes, slip-ups and retries removed.

Script guidance is in beta and the transcription is experimental, so always check the result before relying on it.

### 🤔 How does it work?

Script guidance in FireCut:

* Transcribes the clip or sequence you pick when you click **Go**. It can be up to 60 minutes long
* Saves the transcript, so next time you pick the same source, clicking **Go** opens it straight away. Saved transcripts are kept for 30 days after you last use them. To transcribe again (for example, after changing **Audio language**), click the three-dots button on the right of **Go** and choose **Re-transcribe and Go**
* Matches your script against the transcript, and colour-codes each matched section in both panels: **Script** on the left and **Video transcript** on the right. Hovering over a section highlights its partner in the other panel
* Allows for small differences, such as misspoken or mistranscribed words, numbers said as words ("two" matches `2`, "twenty-one" matches `21`) and symbols said out loud ("percent" matches `%`, "and" matches `&`)
* Leaves out anything you said that isn't in the script, such as retakes, "okay, let me do that again" and "um"s, and skips script lines you never said
* Lets you check and change every matched section before anything changes in your project (see [Reviewing and editing the mapping](#reviewing-and-editing-the-mapping) below)
* Builds a new sequence with only the matched sections when you click **Create timeline**, and leaves your original clip or sequence untouched

{% embed url="<https://www.youtube.com/watch?v=1Dq3wa9SqAk>" %}

### ⚙️ Settings

There are 9 main settings for Script guidance, in three groups.

#### Transcription

* **Source clip or sequence:** The clip or sequence to transcribe. Expand bins to find it, or type in the search box to filter the list. You can select one clip or sequence at a time, so to use several clips, put them in a sequence first. A captions icon shows the items FireCut has already transcribed
* **Audio language:** Set this to the language of your audio. This is an **experimental** setting and supports **left-to-right languages only** (right-to-left languages such as Arabic and Hebrew aren't available yet)

#### Mapping controls

Open these with the sliders icon next to **Mapping**. Changing any of them maps the script again a moment after you stop, and each slider has a reset button that takes it back to its default.

* **Minimum match length:** How many words in a row must agree before a section counts as a match. The default is 5, and you can choose 1 to 20. Higher gives fewer, more confident matches. Lower gives more (and shorter) matches, but more false positives
* **Fuzzy matching word factor:** How different two single words can be while still counting as the same word. The default is 50%, and you can choose 0% to 75%. Higher is more forgiving of transcription mistakes. Lower is stricter
* **Fuzzy matching phrase factor:** When a single word doesn't match, FireCut looks ahead and compares short phrases instead. This sets how different those phrases can be. The default is 20%, and you can choose 0% to 75%. Higher bridges bigger gaps across errors
* **Fuzzy matching phrase length:** How many characters ahead the phrase comparison looks. The default is 50, and you can choose 10 to 250
* **Align to sentences:** On by default. Moves section edges to sentence endings where possible, which helps most around lines you repeated

#### Timeline controls

Open these with the sliders icon next to **Timeline**. These control the cut, and are used when you click **Create timeline** (they don't change the mapping on screen).

* **Section edge padding:** Extra time kept before and after each matched section, so cuts have breathing room. The default is 0.6 sec, and you can choose 0.1 to 5 sec. Higher gives looser cuts, and lower gives tighter ones
* **Remove pauses longer than:** Trims long silent gaps inside a kept section. The default is 1.0 sec, and you can choose 0.1 to 5 sec. Lower it to remove shorter pauses too, or raise it to keep longer pauses

Together, **Section edge padding** and **Remove pauses longer than** decide how tight the cut feels.

### ⚒️ Reviewing and editing the mapping

After you click **Go**, the results screen opens with your transcript on the right. To add your script, click **Click here to open a script** in the **Script** panel, or drag a `.txt` file onto it. Nothing changes in your project until you click **Create timeline**.

* **Edit the script:** The **Script** panel is editable, and FireCut maps the script again a moment after you stop typing. Use **Clear** to empty it and start over.
* **Adjust a selection:** Hover over a coloured section in the transcript and drag the handles on its edges to grow or shrink what's included.
* **Add a section:** Drag across transcript text that isn't highlighted to keep it anyway (useful for an ad-lib you want to keep).
* **Delete a section:** Right-click a section and choose **Delete section**.
* **Re-run mapping:** Maps the script again from scratch (see the note below).
* **Create timeline:** Builds a new sequence called "FireCut Script - " followed by your clip's name, in the Script bin inside the FireCut bin. Everything that didn't match is cut out and the gaps are closed (not just muted or disabled). If the matched sections already cover the whole clip, you'll see "Script timeline created (no cuts required)" and the sequence is still created.

{% hint style="info" %}
**Manual edits stop the automatic mapping.** As soon as you adjust, add or delete a transcript section, FireCut stops mapping the script again by itself, so it won't overwrite your work. The status changes from **Mapped** to **Edited**, and a **Re-run mapping** button appears. While it says **Edited**, editing the script text saves it but doesn't map it again. Clicking **Re-run mapping** maps the script again and **discards your manual transcript edits** (FireCut asks you to confirm first).
{% endhint %}

### 💡 Tips for best outcomes

* **Use clean audio.** Mute background music and sound effects, and keep speech clear — the cleaner the transcript, the better the matches.
* **Match your script to what you actually said.** The closer the script wording is to your spoken words, the more reliably it maps. If a take drifted from the script, either loosen the fuzzy settings or tweak the script/transcript text.
* **Keep your script in spoken order.** FireCut scans top-to-bottom and matches the first good occurrence within range, so large re-orderings can confuse it.
* **Iterate freely.** The transcript is saved after the first run, so adjusting your script and the mapping sliders never re-transcribes. Tweak, watch the highlights update, then create the timeline.
* **Fine-tune by hand when needed.** Drag the section handles or force-keep a selection rather than fighting the sliders for one stubborn section.
* **Treat the output as a first cut.** It's a fast starting point; do your final trims in Premiere.

### 🙋‍♂️ Common questions, issues, watch-outs

* **"A good take or line is missing from the cut"** --> Loosen the matching: lower **Minimum match length**, or raise **Fuzzy matching word factor** or **Fuzzy matching phrase factor**. Also check that your script says what you actually said. If it's still missing, drag across that part of the transcript to keep it anyway
* **"It kept the wrong take of a line I repeated"** --> Drag the section's handles onto the take you want, or right-click it, choose **Delete section** and drag across the take you want to keep. Keeping **Align to sentences** on usually helps repeated lines snap to clean sentence edges
* **"Less than half of the script matched the transcript"** or **"No script matches found in the transcript"** --> FireCut won't create the timeline when less than half of your script's words match. Check that you picked the right clip and the right script, and that **Audio language** is right. If they are, loosen the matching settings, or edit the script so it's closer to what you said
* **"The cuts feel too tight" / "Words get clipped"** --> Raise **Section edge padding**
* **"Long pauses are left in the cut"** --> Lower **Remove pauses longer than**
* **"I edited the transcript and now editing the script does nothing"** --> Your transcript edits stop FireCut from matching again, so they aren't overwritten. Click **Re-run mapping** to go back to automatic matching. Note that this discards your transcript edits
* **"The transcript is in the wrong language"** --> Set **Audio language**, then click the three-dots button on the right of **Go** and choose **Re-transcribe and Go**. Clicking **Go** on its own opens the saved transcript again
* **"My language isn't in the list"** --> Script guidance only supports left-to-right languages for now. Right-to-left languages (Arabic, Hebrew, etc.) aren't supported yet
* **"My clip is too long"** --> Clips and sequences can be up to 60 minutes long. For a longer recording, put each part in its own sequence and run Script guidance on each one
* **"I have multiple source clips"** --> Put them in a sequence and pick that sequence as the source
* **"Will this change my original clip or sequence?"** --> No. **Create timeline** builds a separate new sequence ("FireCut Script - " followed by your clip's name). In the new sequence, the parts that didn't match are removed and the gaps closed, not just disabled
* **"The transcript has mistakes"** --> Because audio is transcribed automatically (and non-English support is experimental), some errors are expected, and the fuzzy matching settings are there to cope with them. If the transcript is badly off, make sure the audio is clear and free of long silences. If it keeps happening, please share your audio with us at <support@firecut.ai>
