How Walls Affect Wi-Fi: Signal Loss by Building Material

Your router is twenty feet from the back bedroom, and the back bedroom does not work. Distance is rarely what is wrong. In open air, every doubling of distance costs about 6 dB — one modest step down. What sits between the router and that bedroom can cost 30 dB in a single stride, and from the hallway a wall that costs 3 dB looks exactly like a wall that costs 30.

Here is the loss per building material, why the angle you cross a wall at changes the number, and how to add it up for your own floor plan.

What a wall actually does to a radio wave#

Three separate things happen, and it pays to keep them apart.

Absorption. Dense, damp materials turn radio energy into a tiny amount of heat. Water content is the biggest single variable, which is why fresh plaster, damp brick, a curing concrete slab and a packed bookcase all cost more than their thickness suggests. This is the mechanism that scales: twice the wall, roughly twice the dB.

Reflection. Conductive surfaces bounce the signal back instead of letting it through, and thickness barely matters. A layer of aluminum foil a few thousandths of an inch thick blocks more than four inches of brick, because what stops the wave is the continuous conductive sheet, not the mass behind it.

Frequency. Loss through dense material climbs as frequency rises. As a rule of thumb, whatever a wall costs you at 2.4 GHz, expect one and a half to two times that at 5 GHz and a little more again at 6 GHz. That one fact decides which band reaches which room, and it is the whole argument of the band-by-band range comparison.

Distance is cheap and predictable; walls are expensive and wildly variable. One mirrored closet door can cost four times what doubling your distance does.

Signal loss by building material#

Every row is a typical range and a rule of thumb, not a measurement. The spread is real: moisture, plaster thickness, what is stapled inside the cavity and the crossing angle all move the number by several dB. For 6 GHz, add roughly 1 to 3 dB to the 5 GHz column on ordinary materials and change nothing on the metal rows — those were already opaque.

The four materials people underestimate#

• Mirrors. A mirror is a sheet of metal with glass in front of it. A mirrored closet door across a bedroom wall is a 20 dB obstacle sitting exactly where you would have aimed.
• Low-E and coated glass. Energy-efficient glazing carries a microscopically thin metallic oxide layer, which is why a garden room with glass on three sides can be worse than the brick room next to it. It is also why open-plan homes still have dead zones despite having almost no internal walls.
• Foil-backed insulation. Foil-faced plasterboard and radiant barrier sheeting are common in loft, garage and new-build exterior walls. A converted attic lined with it behaves like a shielded box, and the stairwell is the only way in.
• Electric underfloor heating. The mat is a conductive grid under the tile, so a bathroom or kitchen with one installed can put a 25 dB ceiling over the room below. It is the usual explanation for a floor that worked until a remodel.

The angle changes the number#

Table values assume the signal crosses the wall head-on. It usually does not. A wall crossed at a slant is a longer path through the same material, and since absorption scales with path length, the dB scales with it too.

Two things compound at a shallow angle: the path through the material gets longer, and more energy reflects off the surface instead of entering it.

That is the missing explanation for the room diagonally opposite the router. It is barely further away in feet, but every wall on the path is crossed at 50 or 60 degrees, so a pair of 6 dB stud walls behaves like 24 dB.

How to count the walls between your router and the room#

You need a floor plan, roughly to scale, and ten minutes.

1. Sketch the floor, room by room. Proportions matter; exact dimensions do not.
2. Mark the router where it actually stands, not where you wish it stood.
3. Mark the spot in the problem room where you actually use the device — the desk, the pillow, the end of the sofa.
4. Draw the straight line between those two marks.
5. List every surface the line crosses, in order. Include the things that are not walls: the chimney breast, the refrigerator, the wardrobe against the wall, the tiled shower, the water heater closet.
6. Give each one a number from the table above, then apply the angle multiplier where the crossing is clearly slanted.
7. Add them up. That total is your wall loss.
8. Now draw a second line: the route you would walk, through doorways. Score it the same way. In most homes it crosses far less material even though it covers more ground.

Your budget for all of this is small. A few feet from the router you are around -30 dBm, and 5 GHz stops being reliable somewhere near -67 dBm, so you have roughly 37 dB to spend on distance, walls, furniture and people. What each level along that scale supports is in the dBm signal strength chart.

One bedroom, three paths, three answers#

A worked example, using the numbers above rather than anything measured. Same house, same router, same back bedroom — only the route differs. Distance loss assumes about 6 dB per doubling from -30 dBm a few feet out.

The longest path wins by 8 dB over the shortest and by 26 dB over the worst. Your device does not choose one of these — it hears all three superimposed and lives on the best. Your job is not to pick the path but to make sure a good one exists. That is the whole logic behind router placement.

When the walls mean an extender will not help#

An extender repeats what it hears. Run the wall count for its candidate positions and the answer is often no.

• Nothing worth repeating. If every outlet on the router side of the problem already sits below -67 dBm, the extender rebroadcasts a weak link at high volume and your phone clings to it. The target window for the extender's own spot is in where to place a Wi-Fi extender.
• A metal or foil boundary. A room lined with foil-backed board, or above a heating mat, is not a range problem. No position on your side of that boundary fixes it; the second radio has to be on the far side, fed by cable, coax or powerline.
• Two masonry walls with no doorway route. At 5 GHz that is 20 to 32 dB before you spend a foot on distance — most of your budget. Solid-wall homes have their own playbook in Wi-Fi through brick and concrete.
• The room needs real throughput. If the count lands the far room near -80 dBm, even a well-placed single-radio extender relays it at roughly half throughput. No position recovers bandwidth the room never received — see what actually causes a dead zone.

Saying no here is cheap. Buying the wrong box twice is not.

Doing the arithmetic on a floor plan instead of on paper#

Adding up one path by hand is fine. Comparing five router positions against four rooms is forty sums, and that is where people give up and guess.

That arithmetic is the whole job Range Up does, and the settings are the part that matters for this article. You sketch the floor plan on your phone or open one of the starter layouts, mark the walls and doors, then set the Material preset — Wood frame, Brick or Concrete — and tune the Wall loss and Door loss values until they describe the rows you just read off the table above. The engine is attenuation-aware, so those figures genuinely change the map rather than decorating it. Drop the router where it really stands and the heatmap does every path sum at once; drag it three feet along the shelf and watch which rooms recover, which is the crossing-angle effect from earlier in this article, made visible.

When you want the second radio question answered, mark the room you are fighting over and the app weighs outlets against that room and puts the two maps side by side. Every level is predicted from your drawing, not read off your live network — which is exactly what you need while the hardware is still hypothetical. Draw your plan and find out which wall is costing you the room, or start with the placement rules.