Why does my alignment drift apart on long drone flights?

Nov 01, 2026 Last reply: 3 hours ago 4 Replies

First reply

CopperLisa90

That seam between flights is almost always a tie-in problem, not a key point problem. When you fly two seprate batteries, the batches usally share some photos in the middle if you planned it right, and Metashape ties the whole block togehter through those overlap images. If the…

Read the full reply

I've been using Metashape for about six months now, mostly small sites around my property and a neighbor's acreage. Short flights with a few hundred photos work fine, but last week I flew a bigger area, maybe 600 photos in two separate batteries, and when I hit align the whole thing comes out warped. One half of the model sits noticeably lower than the other, like the two flights didn't agree on where the ground is.



I tried bumping the key point limit up and rerunning it, which helped a little, but the seam is still there. There's some tree canopy in the middle of the site and not a lot of man-made features, so I'm wondering if that's part of it. Should I be overlapping the two flight paths more, or is there a way to tell Metashape that both batches belong to the same survey? Also wondering if I'm wasting my time fixing this in software when I could just plan better flights. I don't have any ground control points yet, been putting off buying markers. Is this the kind of problem they'd actually solve, or am I barking up the wrong tree?


That seam between flights is almost always a tie-in problem, not a key point problem. When you fly two seprate batteries, the batches usally share some photos in the middle if you planned it right, and Metashape ties the whole block togehter through those overlap images. If the overlap strip is thin, or if most of the images in it are looking at tree canopy, alignment gets shaky because canopy moves between shots. Leaves shift, shadows shift, and the matcher grabs on to stuff that isn't actually fixed to the ground. That's why one half ends up sitting lower than the other, it's basially floating on a weak connection.

Markers will help you here, yeah. I put out 8 on a 90 acre job last fall and the seam between my second and third battery basically vanished once I hit optimize with them checked. You don't need anything fancy, painted plywood squares or even a bucket lid with an X on it reads fine if they're spread across the site and not all in one corner. For the flight itself, I'd plan the two runs so they cross over each other for a couple of passes near the middle instead of just butting end to end. Also try merging the chunks and running camera alignment on the merged block rather than aligning each batch on its own first.

Been doing this since the film and theodolite days, and one habit from back then still works. Fly a short strip perpendicular to your two runs, right across the seam area. Gives the software a diffrent view of the ground and the connection between batteries gets a lot stiffer. Costs you maybe ten minutes of flying. Also check your camera calibration settings didn't reset between flights, I've seen a mismatch there cause exactly the kind of step you're describing. And don't sweat the markers, a couple cheap squares and a borrowed tape is all it takes to get started. Fixing a warped block after the fact never fully goes away, you just chase it around.

Update: tried the perpendicular strip across the seam like the film-days guy suggested and it defintely stiffened things up, the two halves are closer but there's still a slight height difference. So now I'm leaning toward trying markers. Question though: do they need to show up in photos from both batteries, or is it enough if they're scattered around the site?

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required