Why one recording device often produces a poor transcript
A single phone or laptop can record an in-person meeting, but the quality of the resulting transcript depends heavily on where that device is placed.
The person sitting beside the microphone may sound clear. Someone at the other end of the table may be difficult to hear. Quiet speakers, side conversations and people talking over one another create further problems.
Common issues include:
- speakers sitting at different distances from the microphone;
- voices being obscured by room noise;
- several people speaking in quick succession;
- overlapping speech;
- people turning away from the recording device;
- one loud speaker overpowering quieter participants;
- the microphone being blocked by papers, laptops or other objects.
A recording can sound understandable to a person listening back while still producing an incomplete or inaccurate automated transcript.
The problem is not only the transcription model. It is the quality and position of the audio source being supplied to it.
What distributed capture changes
Roundtable records the meeting through devices positioned around the room.
Each phone or laptop hears the conversation from a different location. A participant may sound distant on one recording but clear on the device positioned near them.
Roundtable combines these different perspectives to create a more complete record of the meeting.
This can help with:
- capturing quieter speakers;
- reducing dependence on one central microphone;
- separating contributions made by different people;
- preserving more of the discussion when participants are spread around the room;
- producing a transcript that is easier to turn into useful notes.
The aim is not simply to create more audio. It is to capture each part of the conversation from the device best positioned to hear it.
Learn more about how Roundtable identifies speakers in an in-person meeting.
How to transcribe an in-person meeting with Roundtable
01
Create the meeting
The organiser opens a new Roundtable meeting and shares the joining link with the people in the room.
There is no separate recording device to configure and no meeting bot to invite.
02
Join from the devices in the room
Participants open the link on their phones or laptops.
Devices should be distributed around the room so that the discussion is not being captured from only one position.
Not every person necessarily needs a separate device, but placing devices near the people speaking improves coverage.
03
Start the recording
Once participants have been informed that the meeting will be recorded, the organiser starts the session.
Roundtable captures the discussion across the connected devices.
04
End the meeting
When the discussion finishes, the organiser ends the recording.
Roundtable processes the captured audio and creates a transcript and structured meeting record.
05
Review what was discussed
Participants can review:
- the full conversation;
- the main topics discussed;
- decisions made during the meeting;
- actions and responsibilities;
- unresolved issues;
- the people associated with individual contributions.
This gives the team both a detailed transcript and a shorter record of what matters next.
What affects the quality of an in-person transcript?
No transcription system performs equally well in every room.
The quality of the transcript can be affected by:
Room size
A larger room creates more distance between speakers and recording devices. Distributed capture helps, but devices still need to be positioned near the discussion.
Number of participants
As the number of speakers increases, it becomes harder to separate short contributions and identify rapid changes between speakers.
Background noise
Air conditioning, open windows, adjoining conversations and people moving around the room can interfere with speech.
Overlapping conversation
When several people speak at the same time, parts of the discussion may be difficult to separate clearly.
Device placement
A phone placed beside a speaker will usually capture that person more clearly than a laptop positioned several metres away.
Speaking style
Quiet speech, fast speech, unfamiliar names and technical language can all affect the transcript.
Roundtable is designed to improve the source audio available for transcription, but it cannot remove every limitation created by the room or the conversation itself.
Transcript versus meeting notes
A transcript and meeting notes serve different purposes.
Transcript
A transcript is a detailed record of what was said. It is useful when someone needs to:
- return to the exact discussion;
- check how a decision was reached;
- confirm what a participant said;
- review context that was excluded from the summary.
Meeting notes
Meeting notes are a shorter, structured interpretation of the discussion. They highlight:
- key topics;
- decisions;
- actions;
- owners;
- deadlines;
- unresolved questions.
Roundtable produces both, so participants do not have to choose between a complete record and a practical summary.
When is in-person transcription useful?
In-person transcription is particularly useful when the conversation contains information that may be difficult to reconstruct later.
This includes:
- leadership meetings;
- school staff meetings;
- board and committee meetings;
- client meetings;
- project reviews;
- workshops and planning days;
- stakeholder interviews;
- research discussions.
It is most valuable when participants need to contribute actively rather than spending the meeting writing their own notes.
Frequently asked questions
Can I transcribe a meeting with one phone?
Yes, but the result will depend on the size of the room, the number of speakers and the phone’s position.
One phone may be sufficient for a small, quiet discussion. It becomes less reliable when participants are spread around a larger table or several people speak at once.
Does every person need to join from their own device?
Not necessarily. Roundtable works best when devices are distributed near the people speaking.
A small meeting may not require one device per participant. A larger meeting may benefit from broader coverage around the room.
How accurate is in-person meeting transcription?
Accuracy varies according to the recording conditions.
Room noise, speaker distance, overlapping speech, accents, technical terminology and device placement can all affect the transcript. Roundtable improves the available audio by capturing the meeting from several positions, rather than claiming that every room will produce the same result.
Can Roundtable tell who is speaking?
Roundtable uses the recordings captured across the room to distinguish between speakers and attribute contributions within the transcript.
Speaker attribution can become more difficult when people talk simultaneously, sit far from any device or have similar voices.
Related solutions
Explore connected room workflows
In-person meeting notes
Roundtable turns in-person meetings into clear notes using the phones and laptops already in the room.
Explore in-person meeting notesMeeting notes without a bot
Roundtable can create AI meeting notes for a fully in-person meeting without inviting a bot to Zoom, Microsoft Teams or Google Meet.
Explore meeting notes without a botSpeaker attribution
Roundtable creates a clearer record of who said what during an in-person meeting by comparing audio captured across the phones and laptops distributed around the room.
Explore speaker attributionCreate a more useful record of the meeting
Roundtable is designed for meetings that happen in physical rooms, where one shared microphone is rarely enough.
If your organisation wants to test distributed transcription during real staff, leadership or client meetings, you can request access to the pilot.
Request pilot access