Ask an inspector where a roof leaks and you will not hear "the middle of the south slope." You will hear vent, stack, chimney, skylight, valley, wall tie-in. The field of a slope is the boring part. Everything that punches through it is where water gets an invitation.
Documentation habits do not always follow that. It is easy to shoot forty photos of shingle surface and four of the one flashing detail that will actually be argued about in March.
Shoot every penetration three times
Wide, medium, tight. Not because three is a magic number, but because each one answers a different question later.
- Wide puts the penetration on a slope. Six months on, a tight photo of a cracked boot is a photo of a cracked boot somewhere. A wide frame that catches the ridge line and the neighbor's garage makes it that boot, on that roof.
- Medium shows the flashing relationship — how the shingle courses run into the detail, whether the counterflashing laps correctly, whether somebody caulked their way out of a problem.
- Tight is the condition itself: the split, the lifted lap, the exposed fastener, the failed sealant.
Skip the wide shot and you have evidence with no address. That is the same failure the slope-by-slope method is built to prevent, applied at a smaller scale.
Flashing is a system, not a part
A penetration fails as an assembly. The boot is fine but the shingle above it was cut short. The counterflashing is sound but the reglet was never sealed. The skylight itself is perfect and the head flashing is buried under a course that should have lapped over it.
So document the assembly, not the component. When you narrate it, name the parts in the order water travels: what sheds onto what, and where that sequence breaks. "Water coming down this slope hits the chimney, and the counterflashing here stops about an inch short of the step flashing behind it" is a sentence a homeowner can follow and a sentence that still means something when read back later.
That is also the difference between a note and a record. A note says "chimney flashing — poor." A record says what specifically is wrong with it and why that matters.
Say what you could not see
Half of penetration work is inference. You can see that a boot is cracked. You usually cannot see whether the deck under it is wet without opening something up.
State that boundary out loud and put it in the write-up. "The boot is split and there is staining on the shingle below it. Whether that has reached the deck I can't tell from up here — that needs an attic look or a small opening." Nobody is harmed by a stated limit, and a record that quietly implies more certainty than the inspector had is the kind that falls apart under scrutiny. There is a longer version of this argument in what a record cannot prove.
Count them before you leave
The quiet failure mode is not photographing a penetration badly. It is missing one entirely — the second plumbing stack behind the ridge, the abandoned satellite mount, the bath fan that vents into the soffit.
Count penetrations from the ground or from the imagery before you climb, then count them again in the record before you drive away. Two numbers that match is a thirty-second check that closes the most common gap in roof documentation. If they do not match, you know exactly what you are going back for.
This is where inspecting with somebody watching earns its keep. A homeowner or an office reviewer following the live walk asks "what about the one on the other side?" while you are still on the roof, which is the only moment that question is cheap to answer. The mechanics of that — the link, the recording, what the viewer can and cannot do — are on the how it works page, and the honest way to find out whether it changes your walk is to run one: every plan opens with a free week.
InspectStream records what an inspector observed. It is not a public adjusting service and does not prepare, negotiate, or advise on insurance claims.