For accurate voice-to-text transcription, achieving good audio quality is essential. Here’s a guide to creating an optimal recording environment, covering microphone selection, minimising room noise, and using the microphone correctly for clear, consistent audio.
1. Choosing the Right Microphone
PracSuite’s voice-to-text transcription performs well even with basic, low-cost microphones, such as built-in device mics and affordable computer microphones. For practitioners looking to optimise audio quality, we recommend choosing an external microphone with a digital connection, such as USB or USB Type-C, as these provide better compatibility and clarity compared to older phono-type connectors.
When selecting a microphone, consider your recording needs:
Directional (Cardioid) Microphones: Ideal for practitioners specialising in dictation, as they focus on capturing sound from a single direction while minimising background noise.
Omnidirectional Microphones: Recommended for practitioners recording patient conversations, as they capture sound equally from all directions, ensuring clarity for all voices in the room.
By selecting the right microphone, you can enhance the accuracy and reliability of your transcriptions.
2. Minimising Room Noise
Room noise can disrupt transcription accuracy, so choose a quiet space with minimal echo. Keep windows and doors closed to limit outside sounds, and turn off any sources of ambient noise, such as fans, air conditioning, and music. Additionally, mute device notifications to avoid sudden interruptions.
3. Microphone Setup
Place the microphone on a stable stand to prevent handling noise, positioning it slightly angled towards your mouth. This setup minimises breath sounds and produces a smoother, more professional recording.
4. Proper Microphone Use
Using your microphone correctly is as important as having the right equipment. Follow these practices for clear, consistent audio:
Positioning: Keep the microphone about 15-50 cm from your mouth. Too close may cause harsh sounds, while too far makes your voice sound distant.
Speak Clearly and Consistently: Maintain a steady volume and pace. Natural pauses and a moderate tone will make your speech easier for transcription software to capture accurately.
Avoid Handling the Microphone: Avoid moving or adjusting the microphone during recording, as vibrations can create unwanted noise, pausing the recording if you need to make adjustments.
5. Testing and Adjusting
We recommend setting your audio level in PracSuite so that none or only 1 bar lights up when you're not speaking. When speaking, the audio meter should consistently light up the 3rd bar and occasionally light up the 4th and last bar.
Before a transcription session, record a short sample. Listen for any background noise or distortion, making adjustments as necessary. Testing helps ensure your audio is optimised for accurate transcription.
6. Managing the Recording Session
To maintain clarity and prevent unnecessary transcription errors, manage your recording actively:
Pause During Long Silences: If there are extended pauses or breaks in the conversation, pause the recording. This avoids capturing ambient noise or adding filler text in the transcription and reduces the need for editing later.
End the Recording Promptly: Once you've finished your transcription, stop the recording immediately. This prevents any unintended background sounds from being included in the transcription, ensuring a concise and accurate final document.
7. Recommended Microphone
We have tested several microphones to determine a good option for voice transcription and have identified one that we can recommend: the Audio-Technica Tabletop Boundary USB-C Microphone (ATR4697).
This microphone is a good choice for both practitioner dictation and patient conversations, delivering clear and consistent results in various scenarios. It supports both USB-C and USB-A computer connections and we have tested it on Windows and Mac computers. Its key features include:
Excellent sound pickup: Captures voices effectively, even in room environments.
Low-profile design: Discreet and practical for use on desks or tabletops.
Affordability: Provides good performance at a reasonable price point.
You can purchase this microphone from the following retailers:
8. Notes about Microphones in Video Conferencing Cameras
While convenient, built-in microphones on videoconferencing cameras are often optimised for enhancing audio clarity for human listeners and not for transcription. These devices frequently apply processing techniques such as automatic gain control (AGC), noise suppression, echo cancellation, and audio compression. Although these features can make voices sound clearer during live video calls, they can distort or flatten the audio signal in ways that make it more difficult for AI transcription systems to accurately interpret speech. This is particularly problematic in clinical settings where precise wording and speaker identification are essential.
For accurate AI transcription, we recommend avoiding the use of videoconferencing camera microphones unless they can be configured to disable all audio processing features. Dedicated external microphones typically offer better control and higher audio fidelity.
9. Capturing Audio in Patient Conversation Workflows
When recording conversations between a practitioner and a patient, especially in face-to-face clinical environments it is important to ensure that both voices are captured clearly and consistently. A cardioid or unidirectional microphone, while effective for solo dictation, may only capture the practitioner’s voice clearly, missing or muffling the patient’s responses if they are not directly in front of the microphone.
In these scenarios, an omnidirectional microphone is recommended. Omnidirectional microphones are designed to capture sound equally from all directions, making them more suitable for two-person conversations where participants may be seated at varying angles or move during the session. Position the microphone centrally between the practitioner and patient on a stable surface, and ensure minimal background noise.
It’s worth noting that capturing quality audio in this type of workflow is inherently more difficult than recording dictated notes. The presence of multiple speakers, room acoustics, microphone placement, and background noise can all impact transcription accuracy. For this reason, careful attention must be given to both your working environment and the type of microphone being used to ensure consistent and reliable transcription results.
10. Voice-to-Text Transcriptions in PracSuite
PracSuite’s transcription uses an advanced voice recognition AI ideal for health and medical transcription due to its extensive vocabulary and ability to capture specialised medical terms in real-world environments. This AI strives for accuracy but relies on clear audio input. If background noise or unclear audio is present, it may “hallucinate,” adding non-existent or repeated words. Following the above steps will enhance your voice transcriptions and reduce errors due to poor input.