Audio recording lets a form silently or visibly capture audio during a section of a survey. It is useful for recording interviews, ambient sound during field observations, or spoken responses that complement structured answers.
Mobile app only. Audio recording requires the rtSurvey mobile app. In the web form the audio-start and audio-end appearances are silently ignored — no recording takes place and no error is shown. Always test on-device when audio recording is required.
How it works
Audio recording is controlled by two text fields acting as markers:
- A field with appearance
audio-startbegins recording when the enumerator reaches it. - A field with appearance
audio-endstops recording and saves the audio file to the submission.
Both marker fields are of type text. They are typically hidden from the enumerator (using relevant = false or placed in a non-interactive position), though they can be visible. The recorded audio is attached to the submission as a file linked to the audio-end field.
Basic example
| type | name | label | appearance |
|---|---|---|---|
| text | audio_s_a1 | audio-start | |
| text | q_name | Respondent name | |
| text | q_occupation | Occupation | |
| integer | q_age | Age | |
| text | audio_e_a1 | audio-end |
The app starts recording when the enumerator reaches audio_s_a1 and stops when they reach audio_e_a1. All questions between the two markers are recorded.
Silent recording: audio-start invisible
By default, the mobile app displays a recording indicator while audio is being captured. Adding invisible to the appearance suppresses this indicator, so recording happens without any visible cue to the respondent.
| type | name | label | appearance |
|---|---|---|---|
| text | audio_s_a1 | audio-start invisible | |
| text | q_consent | Do you agree to participate? | |
| text | q_feedback | Your feedback | |
| text | audio_e_a1 | audio-end |
Use invisible only when disclosure requirements have been satisfied through other means (e.g., a separate written consent field before the recording section).
Naming convention
Production forms use a consistent naming pattern for marker fields:
audio_s_{section}{n} — start marker
audio_e_{section}{n} — end markerWhere {section} is a short label for the form section and {n} is a sequential number within that section.
| Field name | Role | Meaning |
|---|---|---|
audio_s_a1 | Start | First recording in section A |
audio_e_a1 | End | End of first recording in section A |
audio_s_b1 | Start | First recording in section B |
audio_e_b1 | End | End of first recording in section B |
audio_s_a2 | Start | Second recording in section A |
audio_e_a2 | End | End of second recording in section A |
Following this convention makes it easy to identify recording boundaries when reviewing form definitions and submission data.
Storage and retrieval
When the enumerator reaches the audio-end field, the app finalises the audio file and attaches it to the submission. The file is:
- Stored as an audio attachment on the submission, associated with the
audio-endfield name. - Accessible in the rtSurvey back-end submission viewer alongside other media attachments.
- Downloadable via the standard submission export — audio files appear in the media column corresponding to the
audio-endfield name.
File format and bitrate depend on the device and app version.
Multiple recording sections
A form can contain several non-overlapping recording sections. Each pair of audio-start / audio-end markers must be fully contained — sections must not interleave.
| type | name | label | appearance |
|---|---|---|---|
| text | audio_s_a1 | audio-start | |
| text | q_intro | Introduction | |
| text | audio_e_a1 | audio-end | |
| note | section_break | — | |
| text | audio_s_b1 | audio-start invisible | |
| text | q_detail | Details | |
| text | audio_e_b1 | audio-end |
Each section produces a separate audio file attached to its respective audio-end field.
Limitations
- Mobile app only. The feature is not supported in the web form.
audio-startandaudio-endare silently ignored in a browser. - No nesting. Recording sections cannot overlap. Starting a new recording before the previous
audio-endmarker is reached is not supported. - Field type must be
text. Usingaudio-startoraudio-endon any other question type is not supported. - Upload size. Long recordings produce large files. Be aware of submission size limits and network conditions in the field.
- No mid-recording pause. The recording runs continuously from the start marker to the end marker; there is no mechanism to pause and resume within a section.