Configure the Mismatch detection rule

The Mismatch detection rule compares each Session item’s Script to its transcription and auto-applies a Property when they differ. This page covers the configuration panel — every toggle, what it does, when to flip it. For the first-time walk-through, do the Catch script-transcript mismatches tutorial.

Open the panel

Open the Session you want to configure.
Open the Auto-rules panel from the Session toolbar.
Switch to the Mismatch detection tab.

The panel header carries a single description — “Detects mismatches between script and transcript text” — and a master Enabled / Disabled toggle. Flipping it off pauses the rule; flipping it on activates it.

The five sections

The panel splits configuration into five collapsible sections. Each shows a summary on its header so you can check the current state at a glance.

1. When to Apply

Add filters here to scope the rule to a subset of Sessions. Each filter is a Session-level Property requirement — for example, “only apply when the Session has the Property Workflow / Final pass.” Leave empty to apply the rule to every Session this user account owns.

2. Assigned Property

This is the Property VoiceQC applies to a Session item when its transcription doesn’t match its script.

Default: Status / Mismatch.
Change it via the Change Property button. The Property picker lets you pick from your Property library or define a new entry on the spot.
The same Property is applied to every flagged item.

If you want different tags for different conditions, you’d need separate rules — multi-property rules aren’t supported today.

3. Comparison Pipeline

Five toggles that control how strict the comparison is. They apply to both the script side and the transcript side before they’re compared character-by-character.

Toggle	What it does	When to leave it on
Convert numbers to words	`25` becomes `twenty-five` on both sides.	Helpful if scripts spell numbers but the ASR transcribes digits (or vice versa).
Convert ordinals to words	`1st` becomes `first` on both sides.	Same as above for ordinals.
Ignore case	`Hello` matches `hello`.	Almost always on.
Ignore punctuation	`Hello, there!` matches `Hello there`.	Almost always on — punctuation rarely survives transcription cleanly.
Normalize whitespace	Collapse multiple spaces, tabs, and newlines into single spaces.	Almost always on.

Defaults: Ignore case, Ignore punctuation, and Normalize whitespace are on. The two number-conversion toggles are off (turn them on if your scripts mix digit and word forms).

4. Synonyms

Tell VoiceQC that two phrases mean the same thing. Each synonym group has a canonical form and one or more variants; before comparison, every variant is replaced with the canonical form on both sides.

A typical example:

Canonical	Variants
`okay`	`ok`, `OK`, `o.k.`
`cannot`	`can't`, `cant`

Add or edit groups in the Synonyms section. Useful when your script style guide differs from how your ASR transcribes filler words and contractions.

5. Test Your Configuration

A scratchpad that lets you paste a script line and a transcription, and runs them through the current pipeline so you can see whether they’d match — without re-evaluating the whole Session.

Use this when you’re unsure whether a normalization toggle is doing what you want. It’s faster than running Re-evaluate Session repeatedly.

Save and re-evaluate

After any change, the Save Changes button in the footer activates. Reset discards your edits and reverts to the saved version.
Click Save Changes.
To apply the updated rule to items already in this Session, click Re-evaluate Session at the bottom left. A toast tracks progress — Evaluating session items… in blue, then Evaluation complete! in green with a count of Properties added and removed.

The rule also applies automatically to new items as their transcriptions arrive — re-evaluation is only needed when you change the rule itself or change the scripts after items were already evaluated.