After the third unusable set of minutes from the large meeting room, the budget request lands on the desk: a conference microphone, maybe a system with ceiling microphones. The assumption behind it is plausible, namely that a better microphone produces a better transcript.
That assumption is partly true. But it leads to an investment that solves one part of the problem and leaves another part untouched. This article compares both routes honestly: what a conference microphone genuinely improves, what it structurally cannot do, and in which situations several linked phones are the stronger choice.
What a conference microphone genuinely improves
First, so no straw man appears: the quality advantage of a good conference microphone over a phone lying flat on a table is real and measurable. It has three causes.
Better capsules and a suitable directional pattern
A phone microphone is a compromise between size, telephony and video recording. A dedicated conference capsule is built for one job: capturing speech from several directions with as little self-noise as possible. The better signal-to-noise ratio pushes back the distance at which a voice disappears into the noise floor.
Then there is the directional pattern. A cardioid microphone rejects sound from behind, an omnidirectional one captures evenly from all directions. For a group around a table, omnidirectional is the right choice, and it is usually implemented better in conference microphones than in a smartphone that was not designed for it.
Beamforming arrays and their real strength
Higher-end systems use several capsules and compute directional information from them. They can electronically focus the recording on whoever is speaking and attenuate sound from other directions. That noticeably reduces the share of room reverberation in the signal, and reverberation is the enemy that eats consonants in a meeting room.
For speaker attribution an array also supplies a second feature alongside sound, namely the angle of arrival. That genuinely helps, especially with similar voices.
Where this advantage actually holds
In short, a conference microphone plays to its strength in a clearly bounded situation: one fixed room, mid-sized groups of roughly four to six people, a table of manageable length. There it moves quality from patchy to usable, and it does so without anyone having to do anything beyond entering the room.
What it structurally does not solve
And this is the part usually missing from the budget request. Three limits remain regardless of price.
The position stays a single position
A table microphone records from the middle of the table, a ceiling microphone from above. In both cases sound pressure still falls with the square of the distance, roughly 6 dB per doubling of the gap. Someone at the head of a long table is far away, and a better capsule does not change that order of magnitude. An array shifts the limit, it does not remove it.
This is why hardware scales badly against this problem: a microphone that costs twice as much does not double the usable range. Beyond a certain group size, every room microphone meets the same physical wall.
Speaker attribution stays an estimate
Even with beamforming, all voices end up in one signal whose sources the software has to reconstruct. When two people sit next to each other they arrive from the same direction, and the directional information stops helping. When two talk at once, separation on a mixed track remains mathematically underdetermined.
Which means: the question of who said something stays a probability statement, and that attribution is exactly what decides whether minutes are dependable or merely look convincing.
It only solves the room you equipped
The limit with the biggest practical consequences is the most mundane one. Hardware sits in a room. The meeting at the client's office, the discussion on a construction site, the conversation in the next office, the second and third meeting room: all of it stays undocumented or needs another device.
For many teams that is the heart of the problem, because the important conversations often do not happen in the equipped room.
The other approach: one close microphone per voice
The alternative inverts the logic. Instead of looking for a microphone that can handle the whole room, it uses the several microphones that are already there. With transcription of in-person meetings with Sally, several phones in the room link into one shared recording.
How it works in practice
One person starts the recording in the Sally app, other participants link their phone to the same recording. Each device becomes the close microphone for the person in front of it. After the meeting Sally merges all tracks into one clean audio track and produces the transcript, summary, decisions and tasks. The result is speaker-separated and reaches up to 98.8 % accuracy according to Sally.
Why it gets better rather than worse with group size
This is the decisive difference. A room microphone gets worse with every additional person, because the average distance grows. The phone approach gets better with every additional linked person, because one more voice is captured close up and gets its own channel.
The two curves therefore run in opposite directions, and in practice they cross in mid-sized groups. A single phone is incidentally enough for documentation, working under the same conditions as any other single device.
Direct comparison
| Criterion | Conference microphone | Several linked phones |
|---|---|---|
| Signal quality, near people | very good | very good |
| Signal quality, distant people | falls with distance | unchanged, every voice is close |
| Speaker attribution | acoustically estimated, array helps | unambiguous via device channel |
| Simultaneous speech | not cleanly separable | separable, distinct tracks |
| Behaviour in large groups | gets worse | gets better |
| Tied to a location | yes, per room | no, anywhere with phones |
| Purchase | hardware per room | none, app on existing devices |
| Effort during the meeting | none | linking phones, a few seconds |
| Depends on participation | no | yes, quality rises with participation |
When each approach wins
Fair is fair, there are clear winning situations for both routes.
The conference microphone is the better choice when
Meetings almost always happen in the same room, groups run to four to six people, and nobody in the team should have to operate an app or bring in a personal phone. It is also the right investment if you want to equip the room for hybrid calls with good speech intelligibility for the remote side, because a conference system covers that job as well.
Several phones are the better choice when
You work in changing rooms and on site at clients, groups are frequently larger than six people, or speaker attribution genuinely has to be correct because tasks and commitments are derived from it. Likewise when in-person meetings should land in the same system as online meetings from Google Meet, Zoom, Microsoft Teams and Webex, so search and task tracking work across everything.
The cost side over two years
A conference microphone is a one-off investment per room, a software assistant costs recurring per person. So the honest comparison depends on the number of rooms and the number of users. With a single room and a small team, hardware is quickly cheaper. With three meeting rooms plus meetings away from the office the calculation tips, because every further room needs hardware again while the off-site case is still not solved. The 30-day free trial is enough to compare both variants in your own meeting room.
What applies in both cases
Whatever the technology, a recording needs a legal basis. In Germany, Section 201 of the Criminal Code makes secretly recording the spoken word a criminal offence, so recordings require the consent of everyone involved, and comparable two-party consent rules exist in several other jurisdictions. In practice: announce the recording at the start and obtain agreement, whether a microphone sits in the middle of the table or phones lie on it. Visible devices help here, because they make documentation transparent.
With Sally, all data is processed exclusively in Germany, GDPR-compliant and without third-country transfer. What that means in detail is on the page about GDPR and security.
Conclusion
A conference microphone is not a bad purchase. It genuinely improves the wanted signal, damps room reverberation and is the most convenient solution in a fixed room with mid-sized groups, because it works without anyone participating.
But it only solves part of it: recording stays one position, attribution stays an estimate, and the solution stays in one room. Anyone who needs unambiguous attribution, works in changing environments or documents larger groups gets further with one close microphone per voice, and those microphones are already on the table.
Compare it in your own room: test Sally free for 30 days, or read the overview on transcribing in-person meetings first.
