Panorama Overlap for a Seamless Row: A Worked Example

A lot of panorama guidance defaults to talking about a full 360° loop, but most panoramas people actually shoot are simpler than that: a wide horizontal row covering a landscape, a cityscape, or a room, sweeping through some portion of a full circle rather than all of it. The planning math is identical either way — it's just applied to a smaller target sweep — and running it through the real panorama overlap calculator beforehand turns "how many shots do I need" from a guess into a number.

Setting up the problem: your lens, your sensor, your target sweep

Three numbers drive the whole calculation: your lens's focal length, your sensor's width, and how much of the scene you're sweeping across in degrees. From focal length and sensor width, the angle of view follows directly: twice the arctangent of sensor width divided by twice the focal length. From there, your target overlap percentage — commonly 20-30% — determines how far the camera can rotate between frames while still leaving that much shared area for stitching software to lock onto.

Worked example one: a wide-angle row

Shooting a 35mm lens on a full-frame sensor (36mm wide) gives an angle of view of about 54.4°. Aiming for a 180° row — a broad sweep across a landscape, not a full loop — at a 30% overlap target, the calculator returns a 36° rotation step and 5 shots total, landing at an actual overlap of 33.9% once that step is spread evenly across the sweep. Five frames for a full 180° is a manageable shoot even handheld on a panoramic head, and each frame carries enough of the scene that even a couple of missed matches between two adjacent shots would still leave plenty of overlap for the stitcher to work with.

Worked example two: a telephoto row, same target overlap

Switch to an 85mm lens on the same sensor for a tighter 120° row — say, a distant ridge line you want more detail on than the 35mm frame would capture — and the angle of view drops to about 23.9°. At the same 30% overlap target, the step shrinks to 15° and the shot count climbs to 8, for an actual overlap of 37.3%. Notice what happened: a narrower total sweep (120° instead of 180°) still needed more frames (8 instead of 5), because the telephoto's narrower angle of view drives shot count far more than the size of the sweep does. This is the single most useful intuition to take from running actual numbers rather than guessing: focal length, not sweep size, is usually the bigger factor in how many frames a row panorama needs.

Why the actual overlap always runs slightly higher than your target

In both examples above, the actual overlap the calculator returns (33.9% and 37.3%) came in higher than the 30% target. That's because the shot count has to be a whole number, and the calculator rounds up rather than down — spreading a few more degrees of overlap across each step rather than risking a final frame that falls short of covering the full sweep. A slightly higher-than-planned overlap costs a little extra shooting and file storage; a gap from rounding down costs the whole stitch failing at that one seam. It's a deliberately conservative design choice worth knowing about so you're not surprised when the "actual overlap" figure doesn't match your target exactly.

Turning the numbers into a shooting plan

Once you have a step angle and shot count, the actual shoot is mechanical: with a panoramic head with degree markings, set the detent to your calculated step angle and take one frame at each position. With a plain ball head lacking markings, count clicks if the head has consistent detents, or use a simple trick — place a piece of tape or a mental landmark at your starting position and estimate the rotation visually using the frame's own edges as a guide, checking that each new frame includes roughly the expected fraction of the previous one. Either way, knowing the shot count in advance means you know when you've covered the full sweep, rather than stopping early or shooting extra frames you don't need.

Parallax is just as much a row problem as a full-loop one

The nodal-point issue that affects full 360° panoramas applies identically to a row: rotating around the tripod's center rather than the lens's entrance pupil shifts a close foreground object against the background between frames. A row shot across a landscape with nothing near the camera is essentially immune to this, which is part of why rows over distant scenery are such a forgiving, beginner-friendly panorama format. A row that includes a nearby foreground element — a fence post, a low wall, a person standing close to the camera — carries the same risk a full loop does, and the fix is identical: a nodal/panoramic head that pivots around the entrance pupil rather than the tripod's own axis, or simply keeping the foreground far enough away that the parallax stays too small for the stitcher to notice.

Leveling still matters as much for a row as for a full loop

A row panorama is just as vulnerable to vertical drift as a full 360° one — if the camera tilts up or down even slightly across the sweep, the horizon curves and stitching software has to fight to compensate, exactly as covered in our broader piece on how much overlap you need for a seamless panorama. Checking the tripod's base level before starting the row, not just eyeballing the first frame's horizon, catches this before it becomes a problem across all five or eight frames rather than just one.

Sensor size changes the numbers even at the same focal length

A 24mm lens behaves differently depending on the sensor behind it, because angle of view depends on sensor width, not focal length alone. On a full-frame sensor (36mm wide), a 24mm lens covering a 180° row at 25% overlap needs just 4 shots, at a 45° step, landing at 39.0% actual overlap. Put the same 24mm lens on an APS-C sensor (roughly 23.5mm wide, giving a narrower angle of view for the same focal length) and the same 180° row at 25% overlap needs 5 shots at a 36° step, for 31.0% actual overlap. The lens didn't change — only the sensor behind it did — and that alone shifted the shot count. Anyone switching bodies between a full-frame and a crop-sensor camera, while keeping the same lenses, should expect their usual panorama shot counts to shift accordingly rather than assuming a lens's "usual" numbers travel with it across sensor formats.

What choosing a lower or higher overlap target actually costs you

Sticking with the 35mm lens and 180° row from the first worked example, dropping the overlap target from 30% down to 15% reduces the shot count from 5 to 4, at a 45° step, for an actual overlap of 17.3% — fewer frames, but right at the edge of the range where stitching software starts to struggle, especially over a featureless sky or water. Pushing the target up to 40% instead raises the shot count to 6, at a 30° step, for an actual overlap of 44.9% — meaningfully more frames for a stitch that wasn't struggling to begin with. This is the concrete version of the general 20-30% guidance: the cost of going too low is stitch reliability, and the cost of going too high is only time and storage, which is exactly why guides skew toward erring high rather than low when in doubt.

Planning shoot time from the shot count

Once you know the shot count for a planned row, estimating shoot time is straightforward: multiply the shot count by however long you need per frame — the time to recompose, wait for a locked-off setup to settle after any vibration, and fire the shutter, plus any extra time for a longer exposure in low light. An 8-shot telephoto row at a few seconds of settling time per frame is a very different time commitment on location than a 5-shot wide-angle row, and knowing the count in advance is what makes that estimate possible before you're standing in the field watching the light change.

When a row is the better choice over a full loop

A row makes sense whenever the interesting part of the scene doesn't actually wrap all the way around you — a coastline, a mountain range, a building facade, an audience — and shooting a full 360° would mean capturing a lot of frames of, say, the parking lot behind you that never make it into the final crop anyway. Deciding on the row's actual angular width first, rather than defaulting to a full loop and cropping later, is usually the more efficient shoot: fewer frames, less time on location, and a target sweep that matches what you actually intend to publish, without extra frames of a parking lot or a tripod bag sitting just outside the interesting part of the scene.