Audio is optional — a well-written trail with good images works perfectly well without it. But when audio is done well, it adds a dimension that text cannot replicate: the warmth of a human voice, the rhythm of a good story, the feeling of being guided rather than reading a notice board. This chapter explains how to produce audio that is genuinely good, using equipment most groups already have.
Do you need audio?
Audio is worth adding if you have the time and someone willing to narrate. It is not worth adding if it will delay getting the trail live, or if the recordings are going to be of poor quality. A trail launched without audio is better than one held back waiting for it.
Audio works particularly well for:
- Atmospheric storytelling — a ghost walk, a First World War trail, a trail about a specific individual whose story benefits from being told in a personal voice
- Accessibility — visitors with visual impairments or reading difficulties who find audio easier than text
- Families and children — a friendly narrated voice holds children's attention better than text on a screen
- Longer stops — if a particular stop has a rich story that would run to 250 words as text, audio lets you tell it fully without overwhelming the screen
Audio adds less value for trails where the stops are primarily navigational, or where the content is largely factual lists — dates, names, dimensions — that are easier to scan as text than to follow by ear.
Writing the script
Audio scripts need to be written differently from stop descriptions intended to be read. The guidance in Chapter 2 covers the key differences in detail — short sentences, no lists, signposted structure, and a direction cue at the end. Read that section before writing your scripts if you haven't already.
A few additional points specific to the recording process:
Write for your narrator's voice
If you know who will be recording, write with their natural speech patterns in mind. A script that sounds slightly formal on the page can sound stiff when spoken by someone who normally talks conversationally — and vice versa. The narrator should read through the script before recording and adjust any phrases that feel unnatural to say aloud.
Mark up the script for delivery
A few simple marks on the printed script help the narrator deliver it well:
- Underline words to be emphasised
- Use a forward slash / to mark a short pause
- Use a double slash // to mark a longer pause — at the end of a section, for example
- Write out numbers in full: "nineteen twenty-three" rather than "1923," which is easy to misread under pressure
- Spell out any unusual place names phonetically in brackets: Cogenhoe (COG-en-hoe), Woburn (WOO-bern)
Time the script before recording
Read the script aloud at a natural pace and time it. Aim for 60–120 seconds per stop. If it runs longer, cut rather than rush — a rushed delivery is harder to listen to than a shorter script.
Who should narrate
The narrator does not need to be a professional broadcaster. What matters is a clear, warm, confident voice and a willingness to do several takes until the recording is right. Many successful heritage audio guides are narrated by society members, local teachers, or community volunteers.
Qualities to look for
- Clear diction — not necessarily a received pronunciation accent, but clear enough that every word is intelligible through a phone speaker outdoors
- A natural pace — the most common mistake is speaking too quickly, especially when nervous. A slightly slower pace than normal conversation works well for audio guides.
- Confidence with the material — someone who knows the subject and finds it genuinely interesting will sound more engaging than someone reading unfamiliar text
- Willingness to re-record — good audio usually requires multiple takes. The narrator needs to be patient with themselves and comfortable doing a stop several times to get the best version.
Multiple voices
Using different voices for different stops — or for different sections within a stop — can add variety and interest. A trail about a specific individual might use one voice for the narrative and a different voice reading from contemporary letters or documents. Keep this simple, though: too many voices can feel disjointed.
Equipment
You do not need specialist equipment to record a usable audio guide. A modern smartphone, used correctly, produces recordings that are entirely adequate for a heritage trail. Better equipment helps, but the recording environment and technique matter more than the microphone.
Smartphone
Any modern iPhone or Android phone has a microphone capable of producing clear speech recordings. Use the built-in Voice Memos app (iPhone) or a free recorder app (Android). Hold the phone 20–30cm from the mouth, slightly to one side to avoid breath noise on the microphone.
USB microphone
A USB condenser microphone — available for £30–£60 — plugs into a laptop and produces noticeably better recordings than a phone, particularly in reducing room noise. The Blue Snowball and Samson Q2U are reliable choices at this price point. Requires a laptop and recording software.
Lavalier microphone
A clip-on lavalier microphone that plugs into a phone's headphone socket — around £15–£40 — is useful if you want to record in the field rather than indoors. It keeps the microphone close to the mouth and reduces wind noise better than holding the phone. Useful for recording ambient sound or on-location narration.
Recording technique
Good recording technique is the difference between audio that sounds professional and audio that sounds like it was recorded in a bathroom. None of it is complicated.
A quiet, small room with soft furnishings. Turn off fans, heating, and anything else that makes a continuous background noise. Close windows. Tell anyone else in the building you are recording. Wait for any traffic noise to pass before starting.
Record 30 seconds, then play it back through headphones — not speakers. Listen for background noise, echo, breath sounds, and whether the voice is clear and at a good level. Fix any problems before recording the full script.
Hold the phone or position the microphone 20–30cm from the mouth, slightly to one side — not directly in front, which picks up breath sounds on the microphone. Don't move around while recording; even small movements cause changes in volume and tone.
Begin recording, wait two seconds, then start speaking. At the end of the script, pause two seconds, then stop recording. This gives you clean audio at each end — essential for editing.
Never rely on a single take. The second or third take is almost always better than the first — the narrator is warmed up, more confident, and has had a chance to hear what works. Keep all takes until you have chosen the best one.
Don't stop the recording — just pause for two seconds and say the sentence again from the beginning. In editing, you can remove the mistake and keep the correct version. This is much faster than stopping, resetting, and recording the whole stop again.
Editing your recordings
Basic editing — removing mistakes, trimming silence from the beginning and end, and adjusting the volume — is all that most trail audio needs. You do not need to be an audio engineer.
Free editing software
| Tool | Platform | Best for |
|---|---|---|
| Audacity | Windows, Mac, Linux — free | The standard choice for beginners. Slightly old-fashioned interface but powerful and well-documented. Can record, edit, remove background noise, and export to MP3. Plenty of tutorial videos online. |
| GarageBand | Mac and iPhone — free | Cleaner interface than Audacity. Good for basic editing and built into all Apple devices. Can export to MP3 via a simple share option. |
| Descript | Windows, Mac — free tier available | Unusually, edits audio by editing the transcript — delete a word from the text and it removes it from the recording. Excellent for cutting mistakes without needing to find them on a waveform. The free tier is sufficient for trail audio. |
| Voice Memos (iPhone) | iPhone — free, built in | Records well and has basic trimming built in. Good enough if your recordings are clean and need minimal editing. Export as M4A then convert to MP3 using an online converter. |
The three edits every recording needs
- Trim the beginning and end — remove silence before the narrator starts and after they finish. Leave about half a second of silence at each end.
- Remove obvious mistakes — false starts, stumbles, and repeated sentences. Use the two-second pause gaps to find them on the waveform.
- Normalise the volume — most editors have a "normalise" function that sets the peak volume to a standard level. This ensures all your stops play at a consistent volume. In Audacity: Effect → Volume and Compression → Normalize.
File format and naming
Save your finished audio files as MP3. MP3 plays on every phone and browser without any issues, and produces small file sizes that load quickly on a mobile connection.
Export settings
- Format: MP3
- Bit rate: 128 kbps — sufficient for speech, small file size
- Sample rate: 44,100 Hz (the default in most software)
- Channels: Mono — speech audio does not benefit from stereo, and mono files are half the size
At 128 kbps mono, a 90-second audio clip will be approximately 1.4MB — small enough to load quickly even on a slow mobile connection.
Naming your files
Use the same naming convention as your image files — lowercase, hyphens, no spaces. Name audio files to match the stop they belong to:
audio/stop-01.mp3audio/stop-02.mp3audio/stop-03.mp3
Store all audio files in an audio/ subfolder within your trail folder, alongside the images/ folder.
Adding audio to the trail
In the map editor, each point of interest has an audio field. Enter the filename of the audio file for that stop — for example audio/stop-01.mp3 — and the trail app will play it automatically when a visitor reaches that point, subject to their browser's auto-play settings.
If the browser blocks auto-play — which some do until the visitor has interacted with the page — a play button will appear instead. This is a browser security feature, not a problem with the trail. Most visitors understand what a play button is.
Audio checklist
- Script written and timed — each stop script is 60–120 seconds at a natural reading pace
- Script marked up for delivery — emphasis, pauses, and phonetic spellings noted
- Recording environment prepared — quiet, soft furnishings, background noise eliminated
- Test recording done and checked — played back through headphones before the full session
- At least two takes recorded per stop — best take selected and others kept as backup
- Recordings edited — trimmed, mistakes removed, volume normalised
- Exported as MP3 — 128 kbps, mono, correctly named
- Files placed in audio/ folder — and linked correctly in the map editor
- Tested on a phone outdoors — audio plays correctly and is audible in the open air