Audio guides for city walks: how they actually work
How do audio guides work when walking around a city?
The audio guide has been reinvented three times, and each version solved a problem the last one created. Understanding that sequence explains why the app on your phone behaves the way it does — and why some of them are so much better than others.
Three generations of the same idea
The first was the rented handset: a plastic wand at the museum entrance with a numeric keypad. You typed the number printed next to the painting. It worked precisely because the museum controlled both the numbers and the route. It does not survive outdoors, where nobody has numbered the buildings.
The second was the downloadable tour — an MP3 and a PDF map, or a podcast episode per district. Cheap to produce and genuinely good when the writing was good, but it demanded that you walk the author's exact route in the author's exact order. Step off it and the narration is describing a corner you passed ten minutes ago.
The third is the location-aware app, which drops the fixed route entirely. It knows where you are and offers what is around you. This is the version that finally matches how people actually move through a city, which is to say: badly, distractedly, and off the route.
GPS triggering versus tapping: the real trade-off
Automatic playback sounds obviously better. In practice it is the feature most likely to make a guide feel broken, and the reason is physics. In a dense city centre, GPS signals bounce off building faces before reaching your phone — the multipath problem, sometimes called the urban canyon effect. Your reported position can drift by tens of metres, which is the difference between two entirely different buildings.
The failure modes are specific and irritating: audio that starts for a building you have already walked past, audio that starts for the building behind you, or two stops firing over each other on a narrow street. Continuous location tracking is also the fastest way to drain a phone battery on a day out.
Tapping is less magical and more reliable. You decide what you are looking at; the app does not have to guess. The best compromise, and the one most well-built apps land on, is to surface what is nearby automatically but let you choose what plays.
Why listening beats reading on the street
This is the part that gets underrated. Reading about a building requires you to look at a screen. Looking at a screen is exactly what you should not be doing when standing in front of a facade someone spent four years carving.
Audio decouples the information from your eyes. You can listen to a description of a tympanum while actually examining the tympanum, which is the entire point of having gone there. It also works while walking between stops, so the transitions stop being dead time.
There is an accessibility dimension too: audio serves people who find sustained reading on a bright screen difficult, and it works in full sunlight, where phone screens do not.
What makes a good stop
- Two to four minutes. Under two and it is a caption; over five and people walk off mid-sentence.
- One idea per stop. A stop that tries to cover the architecture, the founder, the fire and the restoration will be remembered for none of them.
- Written for the ear. Long subordinate clauses that read fine on a page fall apart when spoken. Short sentences. Concrete nouns.
- Specific to where you are standing. 'Look at the third window from the left' is worth more than a paragraph of dates.
- Honest about uncertainty. 'Probably', 'the records disagree' and 'nobody knows' are better than confident invention.
Practical notes for a day of walking
- Download offline before you leave. Mobile data in old city centres is unreliable, and streaming audio between stone buildings fails often.
- Use one earbud, not two. You need the other ear for traffic, and city sound is part of the place.
- Start with the building you are actually curious about, not the first stop on a list. Curiosity holds attention; obligation does not.
- Let yourself skip. A tour you abandon halfway through because something else caught your eye is a successful walk, not a failed one.
How Globo approaches it
Globo skips the fixed route. You point your camera at whatever caught your eye, it identifies the place, and the story is there to read or to play. Because it starts from what you are looking at rather than from a predetermined list of stops, it covers neighbourhood buildings as readily as monuments — and there is no route to fall off.
Questions about this
- Do audio guide apps work without internet?
- Some do and some do not. Apps that pre-download a city or a tour will play offline; apps that stream each stop on demand will stall in areas with poor coverage. Old city centres frequently have weak mobile data, so downloading before you set out is worth the two minutes.
- Are GPS-triggered audio guides accurate in city centres?
- Often not. Signals reflect off building faces before reaching the phone, a problem called multipath or the urban canyon effect, which can shift your reported position by tens of metres. In narrow streets this is enough to trigger the wrong building, which is why many apps let you tap a place instead of relying on automatic playback.
- How long should an audio guide stop be?
- Two to four minutes. Shorter than two minutes and there is no room for more than a caption; longer than five and most listeners walk away before the end. Each stop should make a single point rather than attempt a full history.
- Is there a free audio guide app for walking around a city?
- Globo, from Hot Forge Studio, is free to start and includes audio stories for the places it identifies, with an optional Globo Pro subscription for unlimited scanning. Many museums and city tourist offices also publish free audio tours as podcast episodes or downloadable files.
