A new blog on the wearables-era.com web site features an interview with Kopin CEO John Fan. It provides a very nice summary of the thinking of Fan and Kopin and their approach to developing the VR market. The full-length interview is reprinted below with their permission.
I want to highlight a couple of key elements in their discussion, however. First, Fan believes that microOLED displays will be the dominant display technology for these headsets for some time as it will take years for microLED technology to become cost competitive, and for VR, the brightness of MicroOLED displays would be fine.
Secondly, he says there is a growing debate as to the best display architecture to use for a VR headset. For example, Kopin has developed a silicon backplane for their 1.3” microOLED display that includes display drivers in the backplane with the OLED layers on top. Kopin’s Display-on-Chip approach adds the drivers, MIPI interface and 10-bit HDR color management all in the silicon backplane.
Alternatively, companies like Sony prefer a larger (>2.0”) OLED display on a glass substrate or a silicon-based display with a separate display driver IC that they flip-chip bond to the microOLED. This may sound like a subtle difference, but Fan thinks it can have a big impact on cost, performance and ergonomics of the headset design.
Kopin’s approach of putting a lot of functionality in the backplane eliminates the tricky bonding step, lowers cost by eliminating a separate chip and allows for more programmability in the way the display can be used. Separating the display driving elements into a separate IC lowers yield risk but requires high-volume orders to lower the cost of this IC and limits its flexibility. According to Fan, key manufacturers in this VR supply chain may be coming around to Kopin’s way of thinking. This is really a debate between a hybrid or a fully integrated approach.
Another important piece of the commercialization puzzle is the optics. Kopin has now developed all-plastic Pancake optics that do not cause birefringence issues that have plagued previous attempts at all-plastic Pancake designs. This breakthrough with plastic enables thinner and lower weight solutions consumers have been asking for. While the Pancake designs are less optically efficient that Fresnel designs, this may be outweighed by their advantages, especially as developers adopt dual stack OLED displays and even microlens arrays that can really boost brightness. Again, for VR applications, brightness issues may not be a concern.
This is a debate worth following. – Chris Chinnock, Insight Media
It has been about a year since our last discussion, so I thought this would be a good time to connect with Kopin CEO, Dr. John Fan and capture his current thinking on VR, AR, and Metaverse Glasses. I caught up with John in January following CES 2022 where Kopin showcased a number of new developments. Lots of great information here, please enjoy.
So Pancake optics is an interesting area. When we spoke in 2021, you highlighted the importance of optics for microdisplays in AR and VR. Since then, Kopin has announced patented All-Plastic Pancake optics including a new P80 version at CES 2022. Actually, Kopin has registered Pancake optics a while ago as a trademark. So, how should we think about what’s different and unique about this new class of near-eye optics?
We’ve been talking about Pancake optics for some time and, as you noted, Kopin registered a trademark for the name “Pancake” optics in 2019. I think it is also important to note that we have deployed Pancake optics for quite a long time in Defense programs. In fact, the US Military’s FWS-I weapon sight, our largest production Defense program right now, uses glass Pancake optics.
The whole idea of Pancake optics is actually quite simple: as you magnify an image, you need a certain optical length. So, in order to reduce the optical length so that the optics do not become too thick, you use a polarizing technique to reflect the light rays back and forth and fold them like an accordion. So, that’s the general idea which is now well known to many in the VR industry. I like to point out that one of the original patent inventors for this approach is working here at Kopin and we have successfully used Pancake optics for our defense display modules to make them thin.
All-Plastic Pancake Optics. Photo courtesy Kopin Corp.
Now for Consumers, the current class of VR devices are quite heavy, bulky, and kind of front-end loaded due to the use of larger displays and the thick Fresnel lenses. From our experience in defense programs, we believe one should transition to microdisplays and Pancake optics for Consumer devices in order to have wider user adoption. However, the glass Pancake solution used in our Defense devices, can be quite heavy – this is fine for the specific use by our soldiers, but not for a head-worn device for Consumers.
Interestingly, I think there’s more consensus right now on optics than on MicroOLED as the display of choice for the Metaverse. The industry now understands that Pancake optics would be ideal for VR and the Metaverse. The remaining requirement from the industry was that these Pancake lenses should be all plastic, which has previously been a huge obstacle as there had been no plastic material and processing techniques available to use in plastic for Pancake optics without getting unwanted birefringence aberrations. Kopin has now solved that problem too. When compared to glass lenses, all-plastic Pancake optics provide a number of advantages: flexibility of design, scalability of manufacturing, lighter and lower cost. The very difficult optics puzzle for VR and the Metaverse is now resolved.
So that brings me to my question about what is the “sweet spot” or best combination of display panel size and optics? We hear Sony proposing large 2” panels for their new PSVR2, while the new Shiftall/Panasonic VR device is using your 1.3” MicroOLED panels. From Kopin’s perspective, let’s say for that Personal IMAX Theater experience, what is the optimal panel and optics combination?
I love this Personal IMAX Theater concept and it is a dream that I myself have had for over 10 years. Interestingly, there is a concept that some refer to as the “Emperor’s Seat” inside every IMAX theater – that one seat is where the combined audio and visual experience is optimal or perfect.
I’ve always felt you should have a Personal IMAX Theater where you always get to sit in the Emperor’s Seat and experience the best combination of sights and sounds. On top of that, with a headset you can get 3D effects, head tracking, spatial sound portability, and other things that IMAX cannot provide.
In order to accomplish this, your display has certain requirements. We know the screen has to be very large. We also know the image has to be very, very sharp. And, whether it is for a Personal IMAX Theater, gaming or Metaverse, the display and the associated signal processing have to be very fast with low latency.
This is where I would highlight an observation from the evolution of the smartphone. When you think back to the first iPhone and how Apple disrupted the market, the touchscreen display was a key enabling technology for its massive success. This idea that you could replace a physical keyboard with a much larger touchscreen and multi-touch really changed the game for the smartphone market. Similarly, fast forward to today, and we see the right combination of microdisplay and optics as a key enabling technology for Metaverse Glasses as well as AR and VR headsets. This new class of microdisplay-based devices could put Consumer adoption on a different trajectory compared to what we have seen. Perhaps Apple will again change the game using microdisplays as the key enabling technology for VR? Of course, everyone is waiting to see what they do.
So, back to your question, what is the perfect match? Right now, we think it is a small OLED microdisplay combined with all-plastic Pancake optics that provide magnification of 30,000X to 50,000X while maintaining a very sharp image. The small display with excellent optics performance addresses the issues that many users complain about today – mainly size and weight.
Pancake advantages are: it allows you to get a very sharp image with pretty good eye relief and eye box. A disadvantage is that the optical efficiency is relatively low at between 8% to 10%, so you lose a fair amount of light, and therefore need very bright displays.
However, remember that the Personal IMAX Theater experience and Metaverse Glasses are meant to be quite immersive, so MicroOLED panel brightness of 2,000 nits matched to Pancake optics provides 160 to 200 nits to the eye which is more than enough for immersive VR experiences. This is why MicroOLED and Pancake optics are a perfect match for these applications.
DuoStack MicroOLED. Photo courtesy Kopin Corp.
My dream of the perfect VR Glass in a few years would be based on Duo-Stack MicroOLED panels that are 4K X 4K resolution, run at 120Hz, with HDR 10 bit color. We have actually laid the foundation for all of this in our 1.3” 2.6K X 2.6K panels that Panasonic chose for their VR glasses. Here, I would like to stress why we have picked 1.3” diagonal size. For excellent metaverse experiences, the things that can make a great visual experience, big displays and big optics – run counter to what is required for comfort and style – small and sleek. Therefore, you need a display and optics combination that is big enough to provide excellent resolution, but small enough to keep the headset design small, lightweight and stylish. For MicroOLED on Si displays, 1.3” size is the largest possible because of Si foundry equipment limitations and display costs reasons. Hence, we started with the maximum 1.3” size. With that 1.3” size, one would like to further increase resolution. The next step is 4K by 4K. With such superhigh-resolution microdisplays and further developing Pancake optics with even larger FOV, we may approach the ultimate goal of a digital universe almost indistinguishable from the physical universe – in a small and compact form-factor.
We believe the combination of our 1.3” 4K X 4K Duo-Stack MicroOLED panel paired to wide FOV all-plastic Pancake optics will enable my dream Personal IMAX Theater and VR glasses in just a few years.
Let’s talk more about your production model for MicroOLED and your ability to scale production when demand starts to ramp. Has your fabless approach utilizing OLED deposition partners turned out as successfully as you had hoped?
It is true that the anticipated wave of MicroOLED demand has taken longer than many of us had hoped, mostly because of serious technical and production challenges, but it remains very clear that it is coming and I think our approach remains the right one.
We learned a lot from our work several years ago where we held the leading market share position in supplying HBT transistors that were inside power amplifiers for smartphones. We have since sold that business, but that past experience helped us focus on the segments of the MicroOLED supply chain where we can add the most value while remaining very flexible in how we serve our customers. For example, we believe there are two areas of the MicroOLED supply chain we do not need to be directly involved in: (1) fabricating the silicon wafer itself, and (2) OLED deposition. Where Kopin can add a lot of value is in the design of specialized silicon backplane wafers and our multi-stack OLED architecture – this is the so-called “brain” of the display. This strategy was decided by us a number of years ago and levers our experience and IP that we expect will be in high demand by display manufacturers.
Currently we work with three silicon foundries and three publicly announced OLED deposition foundries (BOE, Lakeside, and Olightek). Interestingly, this approach gives us the flexibility to work with the most appropriate OLED foundry and set of equipment for each display without the high capital investment while maintaining our IP that is critical to the display performance. In fact, I think our ability to work with multiple silicon and OLED foundries is a very valuable asset.
Another point I should note, the current array of 8” fabs globally is fine for MicroOLED panels, but there is a transition to 12” fabs underway that will make production of 1.3” size 2.6K x 2.6K and higher resolution panels much more economical. Remember, we think 4K X 4K resolution in a 1.3” size MicroOLED panel would be quite optimal for VR and Metaverse Glasses.
The component supply chain and network of Original Design Manufacturers (ODM) for VR, AR, and Metaverse devices continue to reside in China and other Asia regions for a variety of reasons. In fact, we see investment increasing there, so we feel it is likely that MicroOLED volume production for consumer applications will follow a similar path as occurred with LED lighting, solar cells, smartphone display panels, and other capital-intensive technologies. When you consider all of this, we remain confident that our fabless multi-foundry approach is the right one and it will allow us to leverage the growing manufacturing scale in Asia.
It has been a long journey to get to this point, but it feels like microdisplay solutions may finally see wider adoption given renewed momentum in VR and the emergence of a Metaverse. Certainly, all eyes are on Apple and the head-worn VR device they are expected to introduce later this year. What is next for Kopin?
Our goal is very simple: continue to lead in the two critical technology areas that are needed for great VR/AR experiences – microdisplays and optics. Optics is an area of opportunity for us because the industry doesn’t see any alternatives to all-plastic Pancake on the VR side. There is so much flexibility in our all-plastic approach and we have strong IP in this area. The ability to refine, modify, and mass produce a plastic lens quickly is far superior to what can be done with glass lenses. All-plastic Pancake also has significant cost and low weight advantages. Once the industry moves to this design, they will not go back to glass.
On the display side, we are now focused on getting our 2.6K x 2.6K panels ready for production and then plan to transition to 4K x 4K resolution. Remember, our MicroOLED architecture is a full Display-on-Chip as we have integrated DDIC, MIPI, and HDR 10-bit color all into the Si backplane. It turns out to be a monumental task and has required much of our resources and innovations. It is really a full display system integrated together — display and IC in one integrated unit. It has performance, power saving and eventually cost-saving advantages. But! What a challenge. Like climbing Mount Everest. We are delighted that we are currently leading the noble charge. However, for the ultimate Metaverse glass, the resolution march should continue. We need to reduce further the pixel size down to achieve full 4K X 4K resolution in a 1.3” diagonal panel size (it has to remain in 1.3” size for reasons described earlier). New innovations are required. We are quite confident that in due time, we will deliver such ultra-high resolution DoC..
There is a competing technical solution using a 2.0” or larger LCD or OLED on glass display, and hybridly bonding a DDIC chip to such a panel. This is what currently leading VR headsets are using. (As an aside, even for these hybrid displays, Pancake optics instead of currently used Fresnel optics would be preferred). Hybrid approach is certainly easier and quicker to go to market. However, full integration such as our Display-on-Chip has many performance advantages – including the always-important smaller size. It may be similar to a hybrid car vs. a fully electric car. Hybrid cars did get to market earlier, but electric cars are the ultimate solution.
Observers should know we have long been a leader in the microdisplay and near eye optics spaces and we also possess the extraordinary benefit of our HBT experience, which was a radical innovation. What did we learn from our success with HBT? The most important lesson is that just like in HBT case, Kopin should use our native disruptive DNA to address and invent the advanced technologies needed for the new Metaverse platform. We focus on our inventions so technology and production capability would come together at just the right moment to meet consumer demand for the new transformation to emerge. We feel such pattern is about to repeat with our all-plastic Pancake optics and MicroOLED displays. The technology is finally here, and a number of manufacturers are starting to build out capacity. If Apple and others deliver MicroOLED devices soon, the demand for our OLED DoC and all-plastic Pancake optics will arrive. In this way, we now see a good path for Kopin to achieve HBT-style success with much higher value components including microdisplays and optics.