3D convergence and interaxial distance are the two camera settings that decide how deep a shot feels and where its subjects sit relative to the screen. Interaxial is the gap between cameras, and convergence is where their views cross.
This page explains both settings and how crews use them, as part of the wider 3d film production guide. It is an independent informational page, not affiliated with any studio, brand or company.
Interaxial: how much depth
A training handout on stereoscopic 3D defines the interaxial as the distance between the left and right camera. By adjusting it, the handout says, crews can dynamically increase and decrease the depth in a scene, and the wider the interaxial, the more depth is captured.
The reference point is the human eye. The same handout notes that eyes are approximately two-and-a-half inches, or 6cm, apart and that matched cameras are typically spaced at roughly that distance. Setting the gap near eye spacing gives a natural sense of depth for ordinary scenes, while a smaller or larger gap changes the feel of the space.
A wider gap makes a scene seem deeper and the world seem smaller, like looking at a miniature. A narrower gap flattens the scene and makes things feel larger. The handout’s glossary describes the cardboard effect, in which the viewer’s eye distance exceeds the interaxial and objects look like cut-outs. Crews therefore change the gap to suit subject distance: small for close subjects, larger for distant landscapes.
The gap also matters because it is hard to repair. The handout says too much depth in a shot, from too much interaxial, is very difficult or sometimes impossible to fix in post-production. For that reason crews tend to be conservative on set and add depth later if needed.
Convergence: where things sit
Convergence is the second setting. The handout says the convergence point determines where an object appears in relation to the screen. If the cameras converge on an object, it appears at the screen plane. If they converge in front of the object, it appears behind the screen, and if the cameras are parallel the picture sits entirely in front of the screen.
The handout lists two ways of adjusting it: toeing in the cameras, meaning angling them inward, or horizontal image translation in post-production, in which the left and right images are shifted sideways relative to one another. Moving the two images apart or together slides the whole scene forward or back relative to the screen.
This links to the idea of screen parallax. When left and right images coincide, there is zero parallax and the object appears on the screen. When the left image sits to the left of the right image, there is positive parallax and the object appears behind the screen. When the order is reversed, there is negative parallax and the object appears in front of it.
Using both settings together
The two settings act on different things. Interaxial sets the total amount of depth in the scene, while convergence moves that depth range forward or back. The handout says that simultaneously manipulating both gives control over the depth and the placement of objects within the 3D space.
A common approach is to choose the interaxial first, based on the nearest and farthest objects, then set convergence so the important action sits at a comfortable place relative to the screen. Most of a scene is usually placed at or behind the screen, with only brief moments in front of it for effect.
- Close subject: a small interaxial keeps near objects fusible, and convergence places the subject near the screen plane.
- Deep landscape: a larger interaxial gives distant objects some depth, though a very large one can make the scene look like a model.
- Moving subject: both settings may change during the shot, which is the work of a convergence puller.
The handout lists the convergence puller as the person who dynamically adjusts live convergence and interaxial separation. The mechanics of the equipment that makes this possible are in the guide to the 3d camera rig.
Limits that protect comfort
Depth is limited by how the eyes work. The handout explains that when viewing 3D the eyes always focus on the screen but may converge anywhere along the depth axis. Extreme convergence can turn the eyes too far inward so the viewer cannot fuse the image, and background divergence beyond the distance between the eyes forces them outward, which is unnatural and uncomfortable.
To manage this, productions set a depth budget, defined in the handout as the maximum positive and negative parallax allowed for comfortable viewing, expressed as a percentage of screen width. A depth script then sets out the budget for each shot. Interaxial and convergence are the tools used to stay inside it.
The screen matters too. Because the budget is a share of screen width, the same shot can feel different on a small display and a large one, and playback on a larger screen can push distant objects beyond what the eyes can comfortably fuse. How pictures reach each eye is covered in the comparison of polarized vs active shutter 3d glasses.
Setting the values on set
On a shoot, the stereographer chooses starting values from the depth script and the lens, then checks them on a monitor that can overlay the two eyes. Looking at the combined picture shows how far apart matching points sit, and that separation is the parallax the audience will see. The nearest object should not exceed the negative limit, and the farthest should not exceed the positive limit.
The lens matters because focal length changes how much of the scene fills the frame, and therefore how strongly the same gap reads as depth. A longer lens compresses space, so a given interaxial produces a flatter result, while a wide lens exaggerates it. Crews usually adjust the gap whenever the lens changes, rather than reusing the same number.
Finally, the crew keeps notes. Recording the interaxial and convergence for each setup helps the editor and the depth grader later, because they know what was intended and can match neighbouring shots. A simple log pays for itself when two shots from different days need to cut together smoothly.
Fixing depth in editing
Convergence can be changed after shooting, which makes it more forgiving than interaxial. The handout defines depth grading as a post-production process in which the convergence point is adjusted by horizontal image translation of the left and right images. It also says that cuts between shots with strong negative and strong positive parallax can be uncomfortable, so editors use depth grading to keep depth continuous.
An active depth cut, the handout adds, dynamically adjusts the convergence point to match the depth position of the outgoing and incoming shots. This is why 3D edits generally run slower: viewers need time to converge on each new shot.
Frequently asked questions
What is the difference between interaxial and convergence?
Show answer
Interaxial is the distance between the cameras and sets how much depth is captured. Convergence sets where objects sit relative to the screen.
Can convergence be changed after the shoot?
Show answer
Yes, using horizontal image translation in post-production. Too much interaxial, however, is very difficult or sometimes impossible to fix.
What is a depth budget?
Show answer
It is the maximum positive and negative parallax allowed for comfortable 3D viewing, expressed as a percentage of screen width.
Why is the cardboard effect a problem?
Show answer
It happens when the viewer’s eye distance exceeds the interaxial, so objects look like flat cut-outs.
