Starlink Obstruction Guide: Trees, Buildings, Chimneys & How to Fix Them

Starlink is unique among satellite internet services because it talks to a constantly moving swarm of low-earth-orbit satellites rather than one fixed point in space. That is what gives it low latency and real broadband speed — but it also means the dish needs a genuinely clear view of the sky, not just a window pointed south. The rule of thumb is simple: from your mount point, nothing should block a cone reaching 25 degrees above the horizon all the way up to zenith, in every direction. Break that rule and you get dropouts, slow speeds, and the kind of intermittent problems that are maddening to debug. Break it badly and the dish will not lock on at all. If you want a precise read on your own site, our sky photo obstruction analyzer converts a phone photo into a real obstruction percentage in about ten seconds.

This guide walks through exactly what counts as an obstruction, how to survey a site before you spend $599 on hardware, how to read the Starlink app's obstruction map, and the mount choices that fix 95 percent of real-world problems. It also covers the hard cases — dense forest, urban canyon, high rise apartments — where no amount of hardware will save you.

What actually counts as an obstruction

The short version: anything opaque to the Ku-band radio signal Starlink uses. The slightly longer version is that there are two categories, and they behave differently in the app score and in real life.

Hard obstructions are permanent, solid objects: buildings, rooflines, chimneys, rock faces, mountain ridges. When a satellite flies behind one, it is gone — there is no partial signal to work with. The app marks these as solid red arcs, and the only remedy is to move the dish or raise it until the arc clears. A neighboring two-story house 6 meters away blocks roughly 20 degrees of elevation on that side, which is 5 degrees below the required horizon and will cost you 10 to 25 percent obstruction on its own.

Soft obstructionsare trees, branches, foliage, power lines, and distant tree lines on the horizon. These attenuate rather than block, which sounds better but is actually worse for user experience — the dish can see a satellite well enough to attempt a handoff, fail partway, and drop the connection. A single mature pine tree 10 meters away and 12 meters tall will consume roughly 6 to 12 degrees of sky from your dish's perspective. Multiple trees compound fast. A full forest canopy gives you essentially zero usable sky, even though in theory some signal still gets through.

Chimneys are the sneaky third category. They look small from the ground, but a standard 1.2 meter tall chimney sitting 3 meters from a roof-mounted dish can block 15 to 20 degrees of sky on its side. This is why most roof ridge installs fail on the first try — the installer sees the dish clearing the roofline, does not realize the chimney behind it is now in the critical northern sky view, and the customer gets a 22 percent obstruction score instead of 3.

Mountains and distant ridges are the benign-looking killer. If a ridge sits 10 kilometers away but rises 500 meters above your elevation, it consumes about 3 degrees of horizon from your site. That sounds fine until you realize the minimum elevation is 25 degrees, so you have no margin. On the Starlink app this shows up as a thin red stripe near the bottom of the sky map — easy to dismiss, costly to ignore.

How to survey your site before you buy

The single best decision any prospective Starlink owner can make is to check their sky before the $599 hardware arrives. Refund windows are 30 days and shipping eats another $50, so a bad install is real money. Three methods, from easiest to most rigorous.

Method 1: sky photo analyzer. Stand at the exact spot where you plan to mount the dish. Point your phone camera straight up. Use the widest lens you have (0.5x on newer iPhones, ultra-wide on Pixels). Take one photo of the full sky above you. Drop it into our obstruction analyzer and it returns your obstruction percentage plus a heatmap of where the blocks are. This is what we recommend first because it takes 30 seconds and uses actual pixels, not estimates.

Method 2: free mobile apps.Sun Surveyor, PhotoPills, and Google Earth Pro all let you overlay the sun's path on your view. A rough proxy for Starlink coverage: if the sun at solar noon summer is clear from your mount point, you probably have a usable northern sky view too (for northern hemisphere installs). The Starlink app itself also has a manual scan mode that works without the dish — it uses your phone's camera and gyroscope to build a sky map. Run it at the exact mount height using a selfie stick if needed.

Method 3: manual protractor. This is old school and costs nothing. Print a paper protractor, tape a string to the center with a small weight, and sight along the 25 degree line. Slowly rotate a full 360 degrees at your mount height. Anything that touches or crosses the line is a problem. This method will not tell you the exact percentage, but it tells you definitively whether you are in the green zone or need to move the mount.

Fixing common obstruction problems

Almost every obstruction problem has one of four fixes: raise the dish, move the dish, trim the obstruction, or accept reduced service. In order of how often they work.

Raise the dish with a pole mount. This is the most common fix by a wide margin. Standard Starlink pole mounts come in 3, 6 and 9 meter heights. A 3 meter pole clears most single-story suburban obstructions. A 6 meter pole handles two-story buildings and small trees. A 9 meter pole is what you need to rise above mature residential canopy. Pole mounts cost $150 to $600 depending on height and wind rating, and they need concrete footing below the frost line in most climates. The rule of thumb: add 2 meters above your worst-case obstruction height to give the 25 degree cone genuine margin.

Move the dish to a different spot. Often the cheapest fix, almost always overlooked. Moving the mount 3 to 5 meters can change which tree or corner of the roof becomes your limiting obstruction. Our coverage and placement map helps you visualize which direction the problematic satellites are crossing, so you know which side of the house to favor. For most US homes, a spot slightly north of the roof peak gives the clearest usable sky because more satellite passes happen in the northern arc than the southern.

Roof ridge install. The highest natural point on most houses is the ridge line of the main roof. A ridge mount uses a non-penetrating clamp that saddles the peak — no holes in the roof. This adds 1.5 to 2 meters over an eave mount at zero additional wind load concerns. Downside: ugly cable run, and you absolutely must survey the chimney problem we flagged above before committing.

Tree trimming. Sometimes the right answer. A single 30-minute arborist visit to remove a few offending upper branches runs $150 to $400 and can drop obstruction from 20 percent to under 5 percent. This is far cheaper than a 6 meter pole in most cases. The catch is that trees grow back. Plan to re-trim every 2 to 3 years, or budget for the pole you should have bought in the first place.

Accept reduced service. If your obstruction score lands between 10 and 25 percent after every reasonable fix, the dish will still work — just with more dropouts than average. For light browsing and email this is usable. For video calls or gaming it is a constant irritation. If you fall in this zone, consider whether your expected workload actually needs Starlink, or whether a cheaper plan matches your reality. Our plan picker will also flag whether Residential or Roam Regional makes more sense given your site constraints.

Reading the Starlink app's obstruction map

Once the dish is powered on, the Starlink app draws a live obstruction map as it accumulates 12 to 24 hours of actual satellite pass data. Understanding what you are looking at matters because the colors and numbers are not labeled well.

The map is a fisheye view of your sky, looking straight up. Center is zenith. The outer edge is the horizon. North is at the top by default (it may rotate based on dish heading). Solid red arcs mean a hard obstruction was detected during a satellite pass — these are buildings, chimneys, roof edges. Partial red or orange blobs are soft obstructions — trees, branches. White areas are clear.

The percentage number at the top is the total fraction of critical sky that registered a blocked satellite handoff in the last 12 hours. Here is what the bands mean in practice: under 2 percent is optimal, and most first-day installs at good sites score between 0 and 1 percent. 2 to 5 percent is acceptable and typical of suburban yards with nearby trees. 5 to 15 percent starts costing real speed. Over 15 percent, the app itself will recommend you move the dish. The recommendation is worth taking seriously — Starlink has more satellite pass data than any individual user.

One gotcha: the map only accumulates data as satellites actually fly over. A fresh install in the first hour shows very little red because not enough passes have occurred yet. Wait a full day before making placement decisions based on the map. The first-day reading can be wildly optimistic.

Extreme cases: forest, urban canyon, high-rise

Dense forest. A mature deciduous or coniferous forest with 30 meter canopy is effectively opaque to Starlink. The only workable solution is a pole that clears the canopy — typically 35 to 40 meters, which is structural-engineer territory and costs $5,000 and up. The realistic path for forest properties is finding a natural clearing, a ridge top, or a pond-side spot where the canopy opens up. A small clearing 8 meters across with surrounding 20 meter trees gives you a 40 degree cone — usable, not great. 15 meters across gives you a 60 degree cone — actually good.

Urban canyon. Row houses, brownstones, narrow European streets all create this problem. Two 4-story buildings 8 meters apart block everything below about 55 degrees elevation on both sides. The only usable view is a narrow strip directly overhead. In this case you need either a rooftop install (which often requires HOA or landlord approval) or you need to accept that Starlink is not the right service. A fiber connection from a normal ISP beats Starlink in every urban case where fiber exists.

High-rise apartments. Counterintuitively, being on floor 20 of a tower does not help if your balcony faces a wall. You need a balcony with genuine open sky above the 25 degree line in most directions, not just one. North-facing balconies in the northern hemisphere are often the best choice because more Starlink satellites cross the northern sky. A suction-cup balcony mount is the cleanest solution here, and many buildings ban drilling anyway.

Pole mounts and roof installations: specs that matter

Not all mounts are equal. The Starlink-branded options are overpriced but well engineered. Third-party mounts from companies like Proxicast, Flagpole Buddy, and Poynting are often better value. The specs that matter are height, wall thickness, base type, and wind rating.

One consideration most DIY installers miss: wind loading on a pole is not linear. Doubling pole height quadruples the bending moment at the base. A 3 meter pole needs a modest concrete footing; a 9 meter pole needs 2 to 3 cubic feet of poured concrete below the frost line, plus guy wires if you are in an exposed location. The dish itself is not heavy, but catches surprising wind at angle. Do not cheap out on the base — a toppled pole can destroy a $599 dish and take a roof with it.

FAQ

Survey before you spend

The single biggest mistake new Starlink owners make is assuming their yard “looks clear enough” and discovering 30 percent obstruction after hardware arrives. A ten-second sky photo survey prevents this. If the survey returns a red zone, you can either plan the pole mount correctly the first time or choose a different service — either outcome beats returning the dish and eating $50 in restocking fees. Use the tools below before you click order.