3D tours and video: what you can check when buying from afar

A growing share of people buying a home in Trieste do so from another country. Often the first "real" visit happens when the choice is nearly made, and all the shortlisting takes place on a screen. That is a healthy change: fewer wasted trips, and you arrive at the viewing with better questions. But it only works if you know what a screen can prove and what it can't.

This article sets out what we have learned producing and publishing 3D tours, videos and computed data on our property pages — including the places where we found our own tools could mislead.

The 3D tour: geometry doesn't lie, dates do

A Matterport tour comes from a scan: a camera with a depth sensor records the home room by room, producing a model you can walk through, with a "dollhouse" view from outside and a floor plan.

What it shows well.

• The layout: how you get from one room to the next, where the windows are, how deep a corridor runs. This is exactly what photographs, by their nature, hide.
• Proportions: is the living room that looks "spacious" in wide-angle photos still spacious in the tour? Often not.
• Measured area: the model carries a measurement of the floor area. It is worth comparing it with the figure on the listing; when the two diverge significantly, something needs clearing up before going further.

What it doesn't show.

• How things are today. A tour is a three-dimensional snapshot of the day it was scanned. We have found tours a few years old showing furniture that was long gone. So the first question to ask is the month of the scan: viewers should know when the home they are looking at dates from.
• That the tour is of that home. It sounds absurd, but it happens: in an archive of models, two different spaces can share a name — two flats on the same street, say, scanned in different years. We learned to match them by street number and comparison of images, never by name.
• That the tour still exists. A good share of the tour links we found in our own records pointed to models that had been archived or deleted. A dead link on a listing is worse than no tour: it suggests someone has something to hide.

A practical detail: the "full screen" button on many embedded tours is unreliable and hard to find, especially on a phone. We added our own, and where the browser won't allow full screen (this happens on iPhone) the tour opens in a new tab. It sounds minor; for someone viewing a home on their phone thousands of kilometres away, it is the difference between exploring it and giving up.

Video: the story, with its limits

Video does a different job. It conveys a home's rhythm, the light as you move through it, the sound of a floor, the view from a window that no photo captures. It is also the format in which it is easiest, without meaning to, to pick the best angle and leave out the rest.

Our advice to buyers at a distance is to ask, alongside the edited video, for a live video call: ask for a window to be opened, for the bathroom ceiling to be shown, for a walk out onto the landing, for the street to be heard. A live call can't be edited.

"Living facts": sun, sea, steps

On our property pages we added a strip of data computed from the property's real coordinates. The rule we set ourselves is strict: a fact appears only if the data support it. Better two true facts than four forced ones; if nothing holds up, the strip isn't shown.

• The sun. Today's arc of the sun over that spot — sunrise, sunset, golden hour — computed in the browser with standard astronomical formulas, always in Trieste time even for someone reading in New York. We also show the difference between the longest and shortest day of the year.
• The sea. The straight-line distance to the coast, measured against a coastline taken from OpenStreetMap. That line is simplified to a tolerance of a few dozen metres, so we round to the nearest ten: we don't pretend to a precision the data doesn't have. Beyond a kilometre and a half, the figure isn't shown.
• Steps. Minutes on foot to the area's main landmark: straight-line distance multiplied by 1.3 (the classic factor for real streets), at 4.8 km/h, rounded up. Better to promise one minute too many than one too few. Beyond 25 minutes it isn't a stroll, and we don't call it one.

What these figures don't say. The sun's arc is computed against the geometric horizon: it doesn't know whether there is a six-storey building opposite the windows, or a hill that swallows the afternoon in December. Walking times don't know that many of Trieste's streets climb, and that 800 metres on the flat is not 800 metres of steps. They are honest estimates, labelled as such, with the source written underneath. They don't replace a visit.

What you still need to see in person

After all that, the list of what a screen can't convey is shorter, but not empty:

• noise: traffic, neighbours, a bar downstairs, church bells;
• smells and damp: mould, drains, cellars;
• real light at different hours, with the real obstructions;
• wind: in Trieste the Bora is not a folkloric detail, and some exposures feel it far more than others;
• the common parts: the state of the stairs, roof and façade, which say a lot about future costs;
• the neighbourhood at ten in the evening, not only at noon.

And then there is what no viewing settles on its own: planning and land-registry compliance, restrictions, mortgages. That is a matter for the notary and the surveyors, not for tours.

In short

A screen is excellent at ruling out: homes that are too small, layouts that don't work, wrong locations. It is far less reliable at confirming. The right way to use it is to arrive at the viewing with a short list and precise questions — and always to ask when the tour you are looking at was made.