What a Quoted Wireless Range Is Measuring

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

The edge of its range is not a place. A quoted wireless range is a link budget written in the wrong units, and once you convert it back two things fall out that change how you shop. The whole spread of published range figures across every no-WiFi baby monitor in this catalogue is 6.02 decibels, which is about two plasterboard walls. And moving the parent unit one room closer buys you the same 6.02 decibels for nothing.

The walkie talkie case is stranger still, and it is the clearest demonstration in consumer electronics of a specification that is true and useless at the same time.

Thirty miles is a statement about the Earth

A Midland LXT630VP3 three pack of FRS walkie talkies publishes "up to 30 miles" and, to its credit, immediately qualifies it. Our own record quotes the asterisk and reads it correctly: Midland states "longer range communication in open areas with little or no obstruction", and the record adds that "FRS is line of sight. Hills, buildings and trees all cut it, and in ordinary terrain real usable range is a small fraction of the headline."

That is right, and the useful version of it is a number. The distance to the horizon from a height h is the square root of twice the Earth's radius times h, which for a 6,371 kilometre Earth comes to 3.5696 times the square root of the height in metres, in kilometres. Radio bends very slightly downward in the atmosphere, which raises the constant to about 4.124.

Put two people holding radios at hand height, say 1.5 metres. Each one can see 4.371 kilometres of horizon geometrically, so the pair reaches 8.742 kilometres, which is 5.432 miles. With refraction it is 10.101 kilometres, or 6.277 miles.

Now solve the other way. For 30 miles, which is 48.28 kilometres, with one radio still at hand height, the other has to be 151.3 metres up on the geometric figure or 109.9 metres with refraction. That is a ridge, a tower, a tall building or an aircraft. It is not two people in a field, and it is not a campsite, a ski area or a worksite.

The surplus the radio is not allowed to spend

Here is the part that makes the figure genuinely strange rather than merely optimistic. At 30 miles the radio is not struggling. It has an enormous amount of signal left over.

FRS runs at 2 watts, which is plus 33 decibel-milliwatts. Free space path loss at the 462 megahertz band centre over 48.28 kilometres is 119.42 decibels. So a receiver with nothing in the way would be getting minus 86.42 decibel-milliwatts, against a sensitivity for a narrowband handheld of around minus 118. That is 31.58 decibels of unspent margin, which is a factor of 1,440 in power.

Spend all of it on distance instead and set the received level to minus 118. The path loss allowed becomes 151 decibels, which puts the distance at 1,833 kilometres, or 1,139 miles.

So the published 30 miles is 1.6 percent of what the electronics can actually reach. The other 98.4 percent is blocked by the ground. Which means the headline figure is not measuring the radio at all, and no choice between two handhelds can move it, because what is limiting it is the shape of the planet.

Our record's framing turns out to be exactly the right one: "The practical way to read it is as a ranking tool between models rather than a distance you will achieve." The arithmetic says it is not even much of a ranking tool, because every radio in the service is pinned at the same 2 watt ceiling and therefore at the same horizon.

A published figure that does not agree with itself

While converting those records something else fell out, and it is in both of them.

A Midland LXT600BBX4 eight pack of business FRS radios publishes, as the three pack does, "36 FRS channels with 121 privacy codes", and then separately that those give "up to 2,662 channel options".

Thirty six times 121 is 4,356. The published combination count is 2,662, and 2,662 is exactly 22 times 121. Twenty two is the real channel count of the FRS service.

So the two figures in the same record disagree by a factor of 1.636, and the one that is internally consistent with the service points to 22 channels rather than 36. Both numbers cannot be right, and the combination count is the one that checks out. Nothing about either record would reveal that without the multiplication, and we carry the same pair of claims on two separate products.

The eight pack is otherwise the better documented of the two and worth saying so. It publishes a 3 year warranty on the radios against 90 days on the headsets, a twelve-fold gap, and our record reads it exactly right: "that gap tells you exactly which component the manufacturer expects to fail first." It also publishes four desktop chargers for eight radios, which the record flags as a planning problem rather than burying. Those are real specifications that mean something, and they are on the listing where the 30 miles also is.

Now the baby monitor, where the number does convert into something useful

A quoted range indoors is a different animal, because the limit is not the horizon. It is walls, and walls are a budget you can actually compute.

A 5 inch 720p no-WiFi pan tilt zoom baby monitor publishes the highest figure in our catalogue and publishes it carefully: "Open-area range is rated up to 2,000 feet, though walls and interference will reduce that in practice", over "a private FHSS direct link that keeps video off the internet".

Convert it. At the 2.44 gigahertz band centre, with a typical 100 milliwatt transmitter, the free space path loss over 609.6 metres is 95.89 decibels, so the receiver sees minus 75.89 decibel-milliwatts. Against a sensitivity of around minus 90 for a video link that is 14.1 decibels of margin, which is what an honest open-field measurement looks like: enough to be repeatable, not enough to be fantasy.

Now bring the parent unit to where you actually sit. Every halving of distance buys 6.02 decibels, so the margin grows as you come in: 20.2 decibels at 300 metres, 29.8 at 100, 40.2 at 30, 46.3 at 15 and 49.8 at 10.

And then spend it on walls. An interior plasterboard wall costs about 3 decibels at this frequency and a solid brick or concrete one about 12. So at 15 metres, which is a realistic distance from a nursery to a living room, you have 46.3 decibels, which is 15 plasterboard walls or 3.9 concrete ones. At 30 metres it is 13 and 3.3.

That is the number worth having. It explains the common experience that a monitor rated for 2,000 feet drops out through two floors of a Victorian terrace while a monitor rated for a third of that works fine across a modern house. The floors are concrete and the house is plasterboard, and the range figure never mentioned either.

The spread between products is two walls

Here is the result that should change what anyone does with a monitor listing.

A VTech no-WiFi baby monitor publishes a 1,000 foot range, and is admirably direct about what that is: our record quotes it as "The 1000ft range is a specification, though actual range varies with walls and interference", with a drawback simply reading "Range may be affected by walls".

A Buradin 1080p no-WiFi pan tilt monitor publishes a 1,200 foot figure alongside a 5000mAh battery, with its own record noting that "Range figures like 1200 feet are maker-stated and depend on walls and interference".

An ALIOBC AI video baby monitor with a 5 inch parent unit publishes 1,000 feet and goes further than any of them on disclosure: "Detection accuracy and range figures are maker-stated, not lab-verified, and range drops with walls and interference."

Convert those three against the 2,000 foot machine. Twenty times the logarithm of 1,200 over 1,000 is 1.584 decibels. Twenty times the logarithm of 2,000 over 1,000 is 6.021.

So the entire top-to-bottom spread of the specification buyers compare across this category is 6.02 decibels. Two plasterboard walls. And the 1,000 versus 1,200 foot difference, which looks like a 20 percent advantage on a listing page, is 1.58 decibels, which is half a wall.

Meanwhile halving the distance between the two units buys 6.02 decibels on its own. One room closer is worth the whole product difference, and it is free. Our guide to choosing a baby monitor that works without Wi-Fi already says the qualitative half of this correctly, noting that "open-area range figures assume clear line of sight. walls and floors cut it down." Three decibels and twelve decibels are the sizes of "cut it down".

The four quantities nobody publishes

A range figure is made of four things: the transmit power, the antenna gain at each end, the receiver sensitivity, and the loss between them. The listing gives you a consequence of all four and none of the four.

Not one baby monitor in this catalogue publishes a transmit power. Not one publishes an antenna gain. Not one publishes a receiver sensitivity. Not one publishes a wall figure. Neither walkie talkie publishes a sensitivity either. Which means the conversion above had to supply all of them from typical values, and that is the honest limit of this page rather than a complaint about the products.

What is striking is how good the disclosure is on everything else. Four of these records explicitly label the range figure as unverified and warn about walls, and one of them, a GoodBaby 5 inch no-WiFi monitor with a stated 1,000 foot range, states the general principle in one line: "All specs here, including battery hours and range, are maker-stated rather than tested by us." That is the right thing to say. The gap is that four records disclaim the number and none of them converts it, and converting it is the only way to make it actionable.

The same shape appears one order of magnitude down. A Pebblebee Clip 5 Bluetooth tracker tag publishes a range "quoted at up to 500 feet or 150 metres", and our page on how Bluetooth item trackers actually work notes the category figure as "usually within about 200 feet on a clear line of sight". Twenty times the logarithm of 500 over 200 is 7.96 decibels, so those two quoted ranges differ by less than three plasterboard walls, in a device whose whole job is to be findable from the other side of a sofa.

Why the antenna cannot rescue it either

One tempting response is to improve the antenna rather than the distance, and it has a ceiling that is smaller than people expect. Our page on why foil behind a router helps one band and nulls the other works out the limit properly: a flat conducting sheet behind an antenna can add at most 6.02 decibels, and only at exactly a quarter wavelength, which is 30.7 millimetres at 2.4 gigahertz.

Notice that 6.02 appears again, and it is not a coincidence. Six decibels is a factor of two in voltage and four in power, and it is the natural size of almost every easy intervention in radio: a perfect reflector behind an antenna, a halving of distance, and the entire spread of published range figures in this category are all the same 6.02 decibels. Two plasterboard walls is the unit of currency, and it is worth knowing that everything on the table trades at roughly one or two of them.

Which is why the honest advice is boring. A monitor two rooms away through plasterboard has about fifteen walls of margin and will not drop out. The same monitor through a concrete floor has under four, and the fix is not a different monitor, because a different monitor is half a wall to two walls better at most. The fix is a shorter path: the parent unit on the stair landing rather than in the basement, which is worth more than every range figure in the category added together.

The honest limit, and what to ask instead

What is exact here is the geometry and the conversions. The horizon is the square root of twice the Earth's radius times the height, and the 3.5696 and 4.124 constants follow from it with no product input. Free space path loss is the standard formula at the published band centres. And the decibel differences between published range figures are exact, because twenty times the logarithm of a distance ratio is what a free space link does whatever the hardware is.

What is representative is every radio parameter, and there are four of them. The 2 watts is the regulatory ceiling for FRS rather than a published figure. The minus 118 dBm handheld sensitivity and the minus 90 dBm video sensitivity are typical rather than stated by any listing we carry. The 100 milliwatt monitor transmitter is typical. And the 3 and 12 decibel wall losses are the middle of commonly published ranges, which vary with construction by several decibels either way.

Those matter asymmetrically, which is the useful thing about them. Move the sensitivity by 10 decibels and every margin in the third panel moves by 10, which is three plasterboard walls, and that is a big change to the absolute answer. But the 6.02 decibel spread between products does not move at all, because it is a ratio of two published distances and nothing else enters it. So the absolute margins here should be read as order-of-magnitude and the comparison between products should be read as exact.

What this page deliberately does not compute is anything about interference, which is the other half of real-world dropouts and is not a path loss problem at all. A neighbour's network, a microwave oven and a cordless phone all raise the noise floor rather than lowering the signal, and no listing publishes a signal to noise requirement, so the margin figures above are against thermal noise alone and are therefore optimistic in a busy block.

Which leaves the question worth asking a seller, and it replaces the range figure entirely. Not how far does it reach. Ask what it transmits at and what the receiver needs, because those two numbers and a tape measure will tell you whether it works in your house, and the quoted distance never will. And if nobody will tell you, the free experiment is better than any specification: put the parent unit where you intend to sit and walk the camera to the nursery. Whatever it does on that walk is the only measurement that matters.