1. Concept Frame: What 3D Video Without Glass Actually Means

Stereoscopic 3D video has existed for decades in two dominant forms: anaglyph and polarized projection. Both require the viewer to wear something on their face. Anaglyph uses red/cyan filters; polarized systems use passive glasses whose lenses match the polarization state of two projectors. In every case, the glasses are doing the work of separating left-eye and right-eye views.

3D video without glass removes that step. The display itself separates the two views, so the viewer perceives depth with no wearable hardware. Two categories of technology are commonly grouped under this label:

• Autostereoscopic displays with a fixed optical layer such as a lenticular lens sheet or parallax barrier. These direct interleaved views toward specific viewing zones.
• Eye-tracked autostereoscopic displays, which add structured-light eye tracking and display-side FPGA processing so the optical layer actively follows the viewer’s pupils.

The glasses-free label is sometimes confused with light-field or holographic displays. Those are related research categories, but commercial glasses-free 3D video products in the current 3DV portfolio fall under autostereoscopic architecture, not full light-field rendering.

2. Architecture Notes: Autostereoscopic Optics, Eye Tracking, and View Mapping

An autostereoscopic display is structurally a conventional flat panel with an added optical layer in front of the LCD or OLED surface. Two implementations dominate:

• Lenticular lens array: vertical cylindrical lenses refract the interleaved columns of pixels so that alternating columns reach the left and right eyes.
• Parallax barrier: an occluding mask blocks alternating columns so only one set is visible from each viewing angle.

Both approaches require the source signal to be a stereo pair, not a single 2D image. The panel’s column layout interleaves the left-eye and right-eye views; the optics then steer each interleaved strip toward the correct eye.

Eye-tracked autostereoscopic displays extend this architecture with three additional components:

1. Structured-light eye tracking projects an invisible IR pattern and reads the viewer’s pupil position via onboard cameras.
2. Display-side FPGA processing rewrites the interleaved column mapping in real time as the viewer moves.
3. Dynamic stereo view mapping delivers a single optimized sweet spot instead of a fixed set of viewing zones.

The result is a monitor-style workflow: the viewer sits at normal desk distance, moves their head naturally, and the display keeps the left and right views aligned with their pupils. There is no headset, no controller, and no forced head pose.

3. Content Pipeline: SBS, Stereo Exports, and 3D-Ready Video Sources

3D video without glass is a delivery format, not a content format. The display still requires a stereo source. Three content families fit the workflow well:

• Side-by-side (SBS) stereo video, where left-eye and right-eye frames are packed into a single frame side by side. This is the most common interchange format.
• Stereo exports from CAD, medical, or industrial 3D viewers, including DICOM 3D volumes, CT reconstructions, and CAD assemblies exported with left/right camera pairs.
• Real-time stereo output from Unity, Unreal, or WebGL pipelines when the application is configured with a stereo camera rig.

Content that does not work natively:

• Ordinary single-view 2D video.
• Flat images or screenshots that do not carry left/right depth information.
• 3D applications that only output a single 2D view to the GPU.

For teams evaluating content fit before procurement, the Content-to-3D Path Checker and the Spatial Display Content Compatibility article describe how SBS, stereo exports, and real-time pipelines map onto the display. The 3DV Spatial Player is the practical SBS preview tool, and the Spatial Display Simulator shows how the eye-tracked view mapping behaves.

4. Workflow Implications for Professional Review Teams

Because 3D video without glass keeps a monitor-style interaction model, the workflow impact is mostly upstream and downstream of the display itself.

4.1 Upstream: Content Preparation

Source pipelines must be able to emit stereo. For medical visualization this typically means volume rendering with left/right camera pairs or stereo DICOM export. For CAD it means a viewer with stereo output or a stereo-aware export to SBS. For Unity and Unreal workflows it means a configured stereo camera rig and a display target list that includes the spatial display.

4.2 At the Display: Viewing Geometry

Eye-tracked autostereoscopic displays expect a single primary viewer in the sweet spot. This maps cleanly onto individual review tasks such as a radiologist reading a CT volume, a design engineer inspecting a CAD assembly, or an NDT technician evaluating a defect. Multi-viewer review is supported by fixed-zone autostereoscopic models, where several viewers share a smaller sweet spot each; eye-tracked models prioritize depth fidelity for one viewer.

4.3 Downstream: Decision Capture

Because the workflow stays on a desk monitor, annotation, screenshot capture, and side-by-side comparison with a 2D reference monitor remain practical. Teams do not have to abandon their existing review software, mouse, and keyboard. This is one of the core differences from VR-based stereo review, where the interaction model is rebuilt around controllers and headset tracking.

5. Limits and Constraints of Glasses-Free Stereo Video

Several constraints should be visible before procurement:

• Single optimized viewer for eye-tracked models. The dynamic view mapping concentrates resolution into one sweet spot. Multiple simultaneous high-fidelity viewers require fixed-zone displays.
• Source content must be stereo. 2D video, flat images, and single-view applications cannot be depth-converted at the display. They will play as 2D.
• Resolution trade-off. Interleaved columns mean each eye receives roughly half the horizontal pixel count of the panel. 4K panels mitigate this; lower-resolution panels show softer per-eye detail.
• Viewer position assumptions. Eye-tracked models assume the viewer’s face is visible to the onboard cameras. Bright backlighting, hats, or unusual glasses can reduce tracking reliability.
• Non-touch interaction model. Current 3DV Spatial Display models are positioned as non-touch review monitors. Touch-driven interaction belongs to the separate Interactive Technology product line.

6. Implementation Notes and Next Technical Steps

For teams moving from concept to evaluation, a practical sequence is:

1. Confirm that one or more content sources can emit stereo. CAD viewers, medical 3D exports, Unity/Unreal rigs, and SBS video are the four most common starting points.
2. Use the Content-to-3D Path Checker to map each source to the display’s expected input format.
3. Run a short evaluation through the 3DV Spatial Player with a representative SBS clip before ordering.
4. Match panel size to room geometry using the Display Selector. Compact demos and dedicated review desks favor 14–15.6 inch models; shared review rooms and labs typically move to 27–32 inch panels.
5. Before procurement, route any compatibility or workflow questions to the Ask Before Ordering path so the source pipeline is confirmed against the chosen display.

3D video without glass is a delivery architecture, not a content shortcut. The display removes the wearable, but the source pipeline still has to carry left and right views. Teams that can already emit stereo from CAD, medical 3D, or real-time engines will find the workflow drop-in; teams working with flat 2D sources should plan a stereo pipeline before expecting spatial depth.