Skip to main content
Choosing the right microphone is critical for achieving high‑quality speech recognition results. Microphone type, placement, and configuration directly affect transcription accuracy, speaker separation, and noise suppression. Corti supports two primary audio capture workflows:
  • Dictation (near‑field) - a single speaker intentionally dictating content into the system.
  • Ambient scribing (far‑field) - natural conversations between multiple participants captured from the room.
Each workflow requires different microphone characteristics to produce the best results. Once you’ve ensured the right device is being used, review audio specifications documented here.

Dictation Microphones (Near‑Field)

Dictation workflows typically involve a single speaker speaking directly into a microphone. Directional microphones are recommended because they focus on the primary speaker while suppressing background noise. These microphones are designed to capture clear speech from close range and are commonly used for clinical documentation, report dictation, or structured note entry.

Key characteristics

Recommended properties for dictation microphones:
  • Directional pickup pattern - focuses on a single speaker
  • Near‑field recording distance - typically 10—20 cm from the mouth
  • Stable wired connection
  • Consistent gain levels with minimal automatic processing
Directional microphones intentionally reduce sounds outside the pickup zone, which improves transcription accuracy for single‑speaker dictation.
Wired devices are recommended over wireless for reliability: Prevent dropped audio packets due to low battery or being moved too far from the receiver.
The following devices have been commonly used and are designed specifically for dictation workflows:
  • Philips SpeechMike Premium
  • Philips SpeechMike Ambient
  • OM SYSTEM RecMic II RM‑4010
In some environments, the following can also work well:
  • Wired USB headset microphones
  • Mobile device microphones (especially iPhones)
  • External USB desktop microphones
Built‑in laptop microphones can work in quiet environments but may capture keyboard noise or fan noise, so external microphones are generally preferred.

Ambient Scribing Microphones (Far‑Field)

Ambient scribing captures natural conversations between multiple speakers in a room. This workflow is common in clinical consultations, meetings, or call‑center environments where speech should be captured passively without interrupting the interaction. Unlike dictation workflows, ambient audio capture requires microphone arrays with beamforming to detect multiple speakers and suppress environmental noise.

Key characteristics

Recommended properties for ambient microphones:
  • Microphone array with beamforming
  • Far‑field pickup range (1—3 m)
  • Automatic gain adjustment for varying speech levels
  • Noise suppression and echo cancellation
Beamforming dynamically focus audio capture on the active speaker while reducing background noise from the room.
The following microphones have demonstrated reliable performance in ambient documentation workflows:
  • Philips SpeechMike Ambient
  • Logitech Rally Mic Pod
  • Jabra Speak2 55
  • Jabra Speak2 40
Some webcams include built‑in microphone arrays and can also work well in small rooms for 1x1 consultations or telehealth visits:
  • Logitech Brio
  • Logitech 9000 Pro
Do not use a dictation, directional microphone for ambient, conversational audio capture.

Choosing the Right Microphone

Selecting the correct microphone depends on the workflow you are supporting:
WorkflowRecommended microphone typeExample devices
DictationDirectional, near‑field microphonePhilips SpeechMike, OM SYSTEM RecMic
Ambient scribingBeamforming microphone arrayPhilips SmartMike, Jabra Speak2, Logitech Rally Mic Pod
TelehealthDirectional headset or dedicated channel per speakerUSB headset microphones
Directional dictation microphones are optimized for a single audio source, while ambient microphones are designed to capture multiple speakers within a room.

Additional Audio Capture Best Practices

Regardless of microphone choice, follow these guidelines for best transcription performance:
  • Dictation:
    • Keep microphones close to the speaker for dictation workflows.
    • Keep microphones at the side of your mouth so you don’t breathe directly into it.
    • Speak in a natural voice - don’t over annunciate, slow down, or speed up your speaking cadence.
  • Ambient:
    • Position ambient microphones centrally in the room when capturing multiple participants.
    • Maintain background noise below ~40 dBA where possible.
    • Avoid reflective rooms that produce echo or reverberation.
  • Both:
    • Use wired connections when possible for stable audio capture.
    • Mobile devices are convenient, but ensure the primary microphone at the bottom of the device is not obstructed - talk into it like you would for a phone call!
For additional guidance on configuring audio capture and streaming settings, see Recording Audio Best Practices.