Creating in stereo – stereoscopic.mov

Today, both stereoscopic photos and videos can be created. While most stereoscopic photos use a “normal” field of view (like most 2D photos) some stereoscopic videos use a much wider field of view, close to 180°. Apple uses the term Spatial for a stereoscopic photo or video with a “normal” field of view (and extra metadata) while using the term Immersive for stereoscopic video with a more expansive 180° field of view. Immersive stereoscopic photos are rare, but possible. Here’s a primer on creating stereo content for the Apple Vision Pro.

Today, spatial is hard, but immersive is harder, mostly because the resolutions required to capture 180° in 3D are truly huge, and the cameras you can use to capture these images are rare and expensive. Here’s a look at the best available today.

Spatial photos and videos on iPhone

To capture a stereo photo or video, the easiest solution is to capture a Spatial Photo or Video with the iPhone 15 Pro, 15 Pro Max, or any iPhone 16 except the new 16e. The wide and ultra-wide lenses each capture an image at the same time, creating a 3D effect.

Though spatial photos appear 2D on an iPhone, they are natively viewable in 3D in an Apple Vision Pro, and can be exported to a side-by-side stereograph using freely available apps. It’s possible to export spatial videos into a side-by-side format, but viewing with a dedicated device (like the Apple Vision Pro) is more common.

Stereoscopic capture on other devices

The iPhone is probably the easiest way to capture in 3D, but many other apps and devices can create stereoscopic images, including:

Use a third-party camera app on iPhone. These allow for higher resolutions but can come with compromises in image quality.
Use a dedicated 3D camera with two lenses. There are very few modern 3D cameras available.
Use a regular mirrorless camera with a special lens that contains two internal lenses. These lenses are rare, but I’ve used such lenses from Lumix (Panasonic) and Canon.
Use two cameras in sync with one another. This process can produce great results, but is more complex, and footage can be ruined if processes aren’t carefully followed.
For still photos of static subjects, take two images with a single camera, moving position slightly between the two shots. This is known as the “cha-cha” method.
If capturing in 3D is impossible or impractical, you could capture in 2D and convert to 3D. So far, this process is imperfect but improving — I took a look at it here.

Stereoscopic alignment is an issue

Spatial photos and videos shot on iPhone with the native Camera app are normally perfectly aligned, but that’s not always the case with other devices — precise positioning of the left and right images is critical.

If the left and right images aren’t exactly aligned, you’ll need to align them, or the results will be uncomfortable to view. This is a more serious problem in video, and especially in a headset, where misalignment can cause nausea.

The most common problem is vertical disparity, in which one eye is positioned slightly higher than the other, but there can also be minor differences in scale, exposure, sharpness and more. It’s very hard to perfectly align two separate cameras, and even the two separate smaller lenses in a single stereo mirrorless lens can be misaligned. This has been the case in both such lenses I’ve used, and I’ve seen it in footage from other cameras too.

Fixing stereoscopic alignment

To fix disparity problems and other issues in still stereoscopic images, the most commonly used app is StereoPhotoMaker. It handles alignment automatically and allows exporting to a variety of formats. Recommended, and though it’s natively a Windows app, a version for modern Macs is also available.

For video, standards are higher, and you must also consider convergence, which defines the apparent depth position of objects in a stereoscopic image. Because the left and right images are captured from different points, only some of the objects in the image will appear to be positioned neutrally (that is, in the position they would be on a 2D viewing screen) and other objects in the frame will appear to be in front of or behind the screen.

A zero convergence value means that the object appears to be in line with the screen.
A negative convergence value means that the image appears to protrude from the screen.
A positive convergence recedes into the distance beyond the screen.

It’s important that the subject of a shot maintains a similar convergence between subsequent clips, or a viewer will have to refocus each time a shot changes. To do this, you should adjust the Convergence value for each shot, moving it backwards or forwards in depth.

Final Cut Pro 11 offers convergence controls and other tools for editing Spatial 3D videos, and it’s easy to get started. Note that Immersive productions are not officially supported.

Editing Stereoscopic Video in Final Cut Pro 11

First you should import your clips and create a Spatial/Stereo 3D timeline. The next steps depend on what kind of stereo footage you’re working with.

Native Spatial video clips can be added to the timeline now.
Clips from a mirrorless camera with a stereo lens need to be tagged first. In the Info Inspector, set the Stereoscopic Conform setting to Side-by-side. When clips are added to a timeline, you should also check the Swap Eyes checkbox.
Clips shot with two separate cameras should have their Stereoscopic Conform set to Monoscopic. Making sure they’re time-synced, position them one above the other at the same point in time, on the left and right of the stereo frame. Add an Adjustment Layer above both clips so you can adjust them together.

While FCP 11 offers control over convergence, it doesn’t offer any controls to correct vertical disparity, or scale, color or sharpness differences between the two angles. Although Canon does offer subscription software to correct these issues for its own lenses, it has several limitations. If you’re editing in Final Cut Pro, there’s no automated solution for correcting these issues for any other lens or camera setup.

To solve the issue, I created Spatial Kit, a plug-in for stereo 3D creators using Final Cut Pro 11! You can read about it here and buy it at FxFactory here.

The editing process for 3D clips is much like the editing process for 2D clips, with the additional factor of convergence to consider. When you’re finished, you can export to the Apple Vision Pro preset, then view it on device, or upload it to Vimeo for wider sharing in 3D. While it is possible to upload 3D video to YouTube, there’s no official YouTube app for Apple Vision Pro. If you do want to pursue this, you’ll need to export to a double-wide SBS format, then recompress in Handbrake with the “frame-packing=3” option — check out my Q&A article linked below for more information.

Spatial photos and videos on iPhone

Stereoscopic capture on other devices

Stereoscopic alignment is an issue

Fixing stereoscopic alignment

Editing Stereoscopic Video in Final Cut Pro 11

Further reading