Optimize transcription
Sally allows you to optimize and regenerate an existing transcription - for example by choosing a different AI model, adjusting the number of speakers, or selecting the correct meeting environment.
This helps you significantly improve your transcription quality without having to re-record the meeting.
Quick navigation:
1. Optimize & regenerate transcription
- Open the meeting you want to optimize.
If you want to optimize the transcription of an imported audio file, open it via the "Recordings" tab.
- Open the transcription on the right-hand side.
- Click “Optimize transcription”. This opens the settings menu.
- In the settings window, you can configure how the new transcription should be generated (see the table below for detailed explanations of each setting):
| Setting | Explanation | Options |
|---|---|---|
| AI model | The AI model determines how precise and robust Sally's transcription will be. Depending on the language, environment, and audio quality, different models can produce significantly better results. Choose the model that matches your situation best. | - Automatic selection: Sally chooses the best model based on audio quality and spoken language. - Sally v1 Common: Versatile multi-language model with low latency — ideal for general meetings. - Sally v1 Common Office Set-Up: Optimized for office environments with enhanced noise and echo suppression. - Sally v1 German: High-precision transcription model specifically for German language meetings. - Sally v1 German Office Set-Up: German-optimized model with additional tuning for typical office background noise. |
| Spoken language | Select the language used in the meeting. Choosing the correct language greatly improves accuracy. | Any of our 35 supported languages |
| Number of speakers | Determines how well Sally can separate and attribute different voices. Providing the correct number improves speaker diarization noticeably. | - Detect automatically - Specify exact number - 1–5 participants - 6–10 participants - 11–15 participants - 16–20 participants - 21–25 participants - More than 25 participants |
| Meeting environment | Helps Sally choose the correct acoustic profile. A precise environment selection improves transcription quality — especially in noisy or echo-heavy surroundings. | - Detect automatically - Field recording / on-site - Office - Dictation - Interview (face-to-face) - Conference / panel discussion - Media production / podcast / studio - Meeting room / boardroom - Online meeting - Phone call / call center / hotline - Unknown environment - Presentation / lecture / seminar / classroom |
| Regenerate summary | Recreates the meeting summary based on the new transcription — useful if the original transcript contained inaccuracies or missing context. | Checkbox (on/off) |
- Click "Regenerate".
Sally will now optimize the transcription based on your selected settings.
This process may take a few moments. You can return at any time to run the optimization again.
2. Recommendations for best transcription results
Sally creates transcripts based on the recorded audio.
The more structured your meeting is and the better the audio quality, the more accurate the following will be:
- Transcript
- Speaker identification
- Summary
- To-dos and decisions
The following recommendations help you achieve optimal results from the start and avoid common issues.
2.1. Introduce all participants by name at the beginning
A short round of introductions at the beginning significantly improves speaker identification - especially in longer meetings or meetings with many participants.
Ask all participants to briefly introduce themselves at the beginning.
Example:
“Hi, I’m Anna Müller from Marketing.”
“My name is Tobias Schneider, Sales.”
The clearer the initial voice association, the more precise the speaker separation will be later in the transcript.
2.2. Clearly announce the agenda and topic changes
Sally processes meetings most reliably when they follow a clear structure and topics are distinctly separated. Ideally, one person should clearly present the agenda at the beginning, and topic changes should be explicitly announced.
Examples:
“Today we’ll cover three points: budget, timeline, and responsibilities for the upcoming CRM project.”
“Now let’s move on to the budget.”
“Let’s continue with the next topic: implementation.”
Clear verbal structure improves transcript segmentation and leads to more precise summaries.
2.3. Avoid speaking at the same time
When multiple people speak simultaneously, content cannot be clearly assigned.
- Only one person should speak at a time.
- If interruptions happen, pause briefly and continue one after another.
- Decisions should be clearly summarized by one person at the end.
Example:
“To summarize: The deadline is March 15, and Max is responsible.”
Clearly formulated closing statements significantly improve the quality of the summary.
2.4. Speak clearly and at a consistent pace
Very fast, unclear, or mumbled speech increases the likelihood of transcription errors.
- Articulate clearly.
- Speak at a natural pace.
- Insert short pauses between sentences.
- Pronounce technical terms clearly and completely.
For important decisions or action items, it helps to state them consciously and clearly.
2.5. Use a high-quality microphone
Audio quality is one of the most important factors for accurate transcripts.
- Use a headset or external microphone instead of your laptop’s built-in mic.
- Keep the microphone approximately 15–25 cm (6–10 inches) from your mouth.
- Ensure a stable connection for Bluetooth devices.
- Avoid changing your distance from the microphone while speaking.
The clearer and closer the audio signal, the more accurate the speech recognition.
2.6. Reduce background noise
Background noise can significantly reduce recognition accuracy.
- Choose a quiet room.
- Close windows and doors.
- Mute yourself when not speaking.
- Avoid typing or paper noise while others are speaking.
- Use noise-cancelling headphones if necessary.
Continuous noises such as fans, construction sounds, or parallel conversations in the background can negatively affect transcription quality.
2.7. Avoid echo and reverberation
Strong room echo makes speech separation more difficult.
- Use rooms with carpets, curtains, or soft furniture.
- Avoid large, empty conference rooms.
- Make sure multiple devices in the same room are not unmuted at the same time (feedback loops).
Echo can cause statements to appear duplicated or distorted in the transcript.
2.8. Do not switch languages during the meeting
Sally processes each transcript most reliably when one primary language is used.
- Stick to one language throughout the meeting.
- If multiple languages are required, create separate meetings or recordings.
Frequent language switching may reduce word recognition accuracy.
2.9. Test your setup before important meetings
Many audio issues can be prevented with a short test beforehand.
- Start a brief test call.
- Check microphone quality and volume levels.
- Confirm that the correct microphone is selected.
- Make sure Sally is properly added to the meeting.
A short test can save significant follow-up work and ensures clean results.