Peter Smith | Smith Displays Limited: Can a transparent display dynamically change its focal depth to solve convergence-accommodation conflict?
08:02 - 09:11
Other snippets from this talk
Summary of the clip:
Can a transparent display dynamically change its focal depth to solve convergence-accommodation conflict?
But if this light coming in here, and this diagram is slightly confusing deliberately, if I put curvature on this wave, then the image which is formed appears to come from a finite depth. This is some work we did for UK. Here we have a display focused on where it says Thor Labs in the distance. We viewed that with a camera, slowly pulling the camera focus forward, bringing this to 30 centimeters, and this goes out of focus.
There is some fringes, but that was in focus. What we then do is we change the radius of curvature of the other beam, and then we are bringing that into focus. This is out of focus and is about to come back into focus. You can see it changing. What we are doing is changing the accommodation depth. This is now a display where we can change the apparent distance, giving a possible solution to the convergence-accommodation conflict.
When we went out in 2025, our first demos, we had about six or eight degree field of view. We now have 35. Everybody said we don't want any free space beams because. The reported objection to free space beams accompanies the field-of-view progression from about six or eight degrees to 35, in a display that can change the radius of curvature of the other beam.
In this short video, you can learn:
* Putting curvature on the incoming wave makes the formed image appear to come from a finite depth.
* Changing the radius of curvature of the other beam brings the image back into focus and changes accommodation depth.
* First 2025 demos had about six or eight degree field of view; the current figure is 35.
📋 **Clip Abstract** A display focused on a distant Thor Labs target was viewed with a camera, then its focus was pulled forward to 30 centimeters, causing defocus and fringes. By changing the radius of curvature of the other beam, the system brings the image back into focus, changing accommodation depth and apparent distance as a possible solution to the convergence-accommodation conflict.
About the speaker:
* Speaker: Peter Smith
* Company: Smith Displays Limited
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#ThorLabs, #RadiusOfCurvature, #AccommodationDepth, #FreeSpaceBeams, #ARDisplays, #DisplayOptics
This is a highlight of the presentation:
The Transparency Paradox: Achieving AR in Crystal-Clear Displays
More Highlights from the same talk.
03:19 - 04:50
How can a non-linear lens turn two infrared beams into a transparent, in-focus AR display?
How can a non-linear lens turn two infrared beams into a transparent, in-focus AR display?
This frequency-doubled green laser pointer emits about a milliwatt of green light: starting with infrared, two photons join together and make visible light. The crazy idea is to make the lens of the glasses out of something non-linear, bring the infrared light in, and get that directed towards the eye. In RF, that is exactly what phased array radar does — you create dipoles, steer and do stuff.
The optical challenge is that the frequencies are 10 to 15 hertz, so making this work needs spacing down to the nanometer. The game is to bring two infrared beams in, get a non-linear response like in that laser pointer, and produce an output like a phased array radar output. In a phrase, that is a beamforming display.
There it is: lithium niobate put into a system, with lasers in the background and a big optical setup. What you see is video, live video, playing on a fully transparent display — there are no diffraction gratings, it is just fully transparent material, and as if by magic visible light comes out. A 3D printed frame holding a pair of glasses with the same material inside makes the point.
In this short video, you can learn:
* A frequency-doubled green laser pointer puts out about a milliwatt of green light by joining two infrared photons into visible light.
* The concept needs two infrared beams and a non-linear response to give a phased-array-like output, described as a beamforming display.
* Lithium niobate in a big optical setup with background lasers showed live video on a fully transparent display with no diffraction gratings.
📋 **Clip Abstract** A non-linear lens brings two infrared beams together to produce a phased-array-like output — a beamforming display — that can be directed towards the eye. Lithium niobate in a large optical setup with background lasers played live video on a fully transparent display with no diffraction gratings, and the same material was placed inside a 3D printed glasses frame.
About the speaker:
* Speaker: Peter Smith
* Company: Smith Displays Limited
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#LithiumNiobate, #NonlinearOptics, #BeamformingDisplay, #InfraredUpconversion, #ARDisplays, #AugmentedReality
12:12 - 14:38
Can a cubic-millimetre wafer-scale laser, a 10-micron organic film and sub-1.4 mW drive make AR glasses Class 1 safe?
Can a cubic-millimetre wafer-scale laser, a 10-micron organic film and sub-1.4 mW drive make AR glasses Class 1 safe?
Four problems define the path to glasses-like AR: miniaturisation, laser choice, spatial light modulator optimisation, and selection of the nonlinear material. Lab and science lasers have given way to a 50-kilowatt peak power miniature laser, one cubic millimetre in size and wafer-scale produced — the kind of device you can actually put into a pair of glasses. In the lab, the glowing element is the laser; the surrounding metalwork merely holds it.
A total-internal-reflection display carried a University of Southampton logo and a replicated eye box — proof that the laser moves from laboratory bench to something small enough to use. Specifications shown at Photonics West covered a single exit pupil; 35 degrees was demonstrated with a route seen to 70, and the organic variant reached 120 pixels per degree. Transparency has always been perfect: anti-reflection coated, it is as good as you could want.
Power is the deciding constraint. A green laser point measured roughly two milliwatts; a display needs only a few microwatts because laser light is the highest on viewpoint and does not pass through a scattering process. Getting below 1.4 milliwatts would make the system automatically Class 1, and with optimisation that looks achievable. Ten microns of organic thin film also looks attractive for weight, and may be the right route to manufacture.
In this short video, you can learn:
* Four hurdles must be cleared: miniaturisation, laser selection, spatial light modulator optimisation and nonlinear material choice.
* A wafer-scale, cubic-millimetre, 50-kilowatt peak power miniature laser makes a glasses-mounted light engine feasible.
* Dropping drive power below 1.4 milliwatts would confer automatic Class 1 laser safety.
📋 **Clip Abstract** The route to glasses-like AR rests on four fronts: miniaturising the light engine, choosing the laser, optimising the spatial light modulator and selecting the nonlinear material. A 50-kilowatt peak power cubic-millimetre wafer-scale laser, TIR displays demonstrated at 35 degrees with a path to 70 and 120 pixels per degree, and a sub-1.4 milliwatt drive for automatic Class 1 safety mark the reported progress.
About the speaker:
* Speaker: Peter Smith
* Company: Smith Displays Limited
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#TotalInternalReflection, #WaferScaleLaser, #Class1LaserSafety, #OrganicThinFilm, #ARDisplayOptics, #NonlinearPhotonics
05:34 - 07:50
How does a non-linear 4F imaging system create a full image without laser beam scanning?
How does a non-linear 4F imaging system create a full image without laser beam scanning?
Start with a laser and combine light through two paths into a non-linear material. The chosen material can be PPLN, though other materials exist. Two infrared photons are joined to make visible light, so the source is infrared lasers rather than visible lasers. One beam is marked DLP but acts as a spatial light modulator; the other arrives from the opposite side and they join to produce an image.
The system forms the whole image at once, not by laser beam scanning. It is closer to laser illuminated LCOS with a non-linear twist. In a 4F imaging system, points in the object transform into points in the Fourier plane, or into angle; the second lens transforms angle back into points in the image. The eye sits close to the glasses, 16 millimeters away, converting that angular information.
To create the image, points on the spatial light modulator are illuminated with a laser and pass through the first lens of the 4F system. That lens converts position into angle in the Fourier plane. Angular spectrum information then travels through the non-linear material, and green light emerges containing the angles placed into the system. Feeding a simple video signal therefore generates the whole image at the same time.
In this short video, you can learn:
* Two infrared photons are combined in a non-linear material to produce visible green light.
* A spatial light modulator, illuminated by a laser, feeds angular information into a 4F imaging system that creates the whole image simultaneously.
* Because the image is not formed by laser beam scanning, the approach resembles laser illuminated LCOS with a non-linear twist.
📋 **Clip Abstract** Infrared laser light is routed through two paths into a non-linear material, where two infrared photons combine to generate visible light. A laser-illuminated spatial light modulator then feeds angular information through a 4F imaging system, producing a full image at once rather than by laser beam scanning.
About the speaker:
* Speaker: Peter Smith
* Company: Smith Displays Limited
* Event: Eindhoven 2026
* Location: High Tech Campus, Eindhoven
#PPLN, #SpatialLightModulator, #InfraredPhotons, #4FImaging, #ARDisplays, #LaserDisplays