Photography Lighting & Exposure

How Focal Length and Subject Distance Change Perspective Compression

A backdrop pattern can grow behind a portrait even when the face stays the same size in the frame. This guide separates the effect of moving the camera from the effect of changing focal length. With 35mm, 50mm, and 85mm lenses, you’ll match subject framing, measure two targets in pixels, and predict how their image sizes compare before moving the tripod.

Same subject framing, three camera positions Side view to scale · angles show nominal full-frame horizontal coverage Lens Subject Backdrop Backdrop / subject 35 mm 54.4° 50 mm 39.6° 85 mm 23.9° 41% 50% 63% 1.40 m 2.00 m 3.40 m 2.00 m 2.00 m 2.00 m Camera viewpoint Ratio measured in image pixels
Read each row from the camera dot to the subject and then to the backdrop. The backdrop stays 2.00 m behind the subject, and the percentage gives the projected height of its target relative to an equal-height subject target.

Set the viewpoint before choosing the lens

If a portrait needs a quieter sense of depth, move the camera back and fit a longer lens to recover the original framing. The move changes the relative apparent size of the sitter and the set behind them. At the same viewpoint, switching from 35mm to 85mm narrows the field of view and enlarges everything within it by the same factor in an ideal rectilinear image. A centered crop of the 35mm shot can reproduce the 85mm view, with fewer pixels across the sitter.

The consequence shows up anywhere there is depth, including within the subject. Suppose the nose is 8 cm closer to the camera than the plane of the ears. At 0.80 m from the ear plane, a millimeter at the nose projects about 11% larger than a millimeter at the ears. At 2.00 m, the difference is about 4%. These are local scale comparisons, so they describe the effect of moving the camera without assuming a particular face shape.

Predict the size of the backdrop in the frame

Call the axial distance from the camera viewpoint to the subject plane D and the subject-to-backdrop gap G. For upright targets parallel to the sensor, image height is proportional to physical height divided by axial distance. Two targets of equal height therefore give a backdrop-to-subject image height ratio of D / (D + G). The common projection scale cancels. Measure along the camera axis between planes, even if the targets sit a little to either side of the centerline.

Put a 20 cm ruler beside the subject and a second one 2.00 m farther back. At a subject distance of 2.00 m, the rear ruler spans half as many pixels as the front one. If the rear ruler is 30 cm tall, multiply the distance ratio by 30 / 20. It then spans 75% of the front ruler’s image height. This correction matters when you compare a printed pattern in the background with a feature on a person or object.

To plan the set in reverse, divide the desired ratio by one minus that ratio and multiply by the gap. A 60% ratio with a 2.00 m gap requires a 3.00 m camera distance. If the room stops you at 2.00 m, a 1.33 m gap gets you the same backdrop ratio. That change preserves the camera’s view of the subject, so the facial proportions still differ from the shot at 3.00 m.

Match framing at 35mm, 50mm, and 85mm

Use one camera in full-frame stills mode, with a 36 × 24 mm image area in landscape orientation. Start with the 50mm lens at 2.00 m. In the ideal projection, subject image height scales with focal length divided by subject distance. The 35mm starting mark is 2.00 × 35 / 50 = 1.40 m. The 85mm mark is 2.00 × 85 / 50 = 3.40 m. All three have the same nominal 0.025 image-to-object scale at the subject plane. A 20 cm target then projects to about 5 mm on the sensor, or 500 pixels in a 2,400-pixel-high export. The angles in the figure are nominal horizontal fields of view, calculated as twice the arctangent of 18 mm divided by focal length. Real coverage can shift at closer focus.

Predicted target sizes with matched subject framing and a fixed 2.00 m gap
LensCamera to subjectCamera to backdropSize ratioBackdrop target at 500 px subject height
35mm1.40 m3.40 m41.2%206 px
50mm2.00 m4.00 m50.0%250 px
85mm3.40 m5.40 m63.0%315 px

The same pattern element appears about 53% taller in the 85mm frame than in the 35mm frame after you move the camera to keep the subject size fixed. If the backdrop is a wall of evenly spaced tiles, fewer tiles fit behind the sitter in the 85mm shot. The 5.40 m setup also needs clearance for the tripod and operator. When you run out of floor, decide whether subject perspective or backdrop scale matters more, then adjust either the camera position or the gap accordingly.

Run two controlled comparisons on the studio floor

Keep the camera in place

Tape the subject plane and backdrop plane on the floor. Mount two matte cards with high-contrast marks exactly 20 cm apart vertically, one card on each plane. Offset the back card just enough to keep both marks visible, and keep both targets near the center of the narrowest frame. Square the cards to the camera, level the head, and keep camera height and tilt fixed. Use steady light across the cards so shadows and glare leave the mark centers easy to identify.

Photograph both cards with each lens from the 2.00 m station. Focus on the front card and stop down enough to read the rear marks. Export with consistent aspect ratio and distortion correction settings. Crop the wider shots around the optical center and scale them until the front target matches the height in the 85mm image. Read the pixel distance between mark centers on each card. With a fixed viewpoint, the rear-to-front result should stay near 50% across all three shots.

Keep the front target the same size

Take a fresh 50mm reference and write down the front target’s native image height. With the 35mm and 85mm lenses, start on the calculated marks, refocus, and slide the tripod until the front target is within 1% of the reference height in the uncropped image. Record the actual distances and use those to calculate fresh predictions. For a front target near 500 px, a 5 px difference is a reasonable framing tolerance. Record the ratio from each image using its own two measurements.

Diagnose unexpected ratios before changing the set

If the ratio shifts during the fixed-camera comparison, check the card alignment and lens corrections first. An off-center crop can spoil a visual overlay, although cropping alone preserves the height ratio if both marks remain in frame. Viewpoint is at the entrance pupil, whose position can vary when you swap lenses even on a locked tripod. The camera’s sensor-plane mark gives you a repeatable tape reference, with a small offset from the viewpoint. To check the effect, line up a near and a far vertical edge off center in live view. Match their alignment after a lens swap by sliding the camera along its axis and keeping its aim fixed. This matters most with close tabletop work.

Focus breathing can shift the framing when you refocus and is one reason to treat the floor marks as starting estimates. Trim the camera position to recover the intended framing and record where it lands. If the rear marks are soft, measure the centers of the blurred strokes, or repeat with a smaller aperture and more light. A soft background may look quieter even as its pattern gets bigger. Judge blur and pattern size separately when you choose the final aperture and camera position.

Verify and record the setup

  • Mark the camera station, subject plane, and backdrop plane. Record the axial distances, camera height, and any change in tilt.
  • Check that both targets are upright, parallel to the sensor, and readable in the narrowest frame.
  • Compare the fixed-viewpoint crops at matched front-target size. Then measure the uncropped set made from the three camera stations.
  • Divide the rear pixel span by the front span and compare it with the prediction from the measured distances. For a 500 px front target, a difference of two percentage points corresponds to 10 px at the rear. Recheck the setup if the result exceeds that working tolerance.
  • Save the lens, focus setting, aperture, frame dimensions, distance readings, and both pixel spans with the reference frame so you can rebuild the shot.