Limbak Bought by “Large US Company”

Yesterday’s article Meta Buys Luxexcel resulted in another tip I could confirm: Spanish optics design company Limbak was bought by “a large US company.” The speculation is that it might be either Apple or Meta, but I don’t have a specific company. Limbak website used to be (the last WaybackMachine Capture of the Website was July 3 2022 — very little still works), but that website will now give a DNS error (consistent with a buyout like we saw when Meta bought Imagine Optics). The fact that the website stopped updating on the WaybackMachine, may indicate that the deal was completed some time ago (like the Luxexcel deal).


Limbak developed compound (reflective & refractive) optics and some “different” refractive optics. They are probably best known in the industry as being used by the startup Lynx for their AR Passthrough headset.

With the website gone, you can still find out about some of what Limbak was working on via their AR/VR/MR 2019 presentation on YouTube and their 2021 Photonics Media Webinar and YouTube Presentation about their “Thin Eye” multiple mini-lenses foveated optics.

Limbak has touched on some key topics with their design. Pancake optics are a hot topic with many companies, including Meta and (reportedly) Apple, using it. Limbak claims to have a “super” pancake optical design. They are probably best known for their compound optics (mix reflective and refractive in a single device) ThinEyes clover used in Lynx. The Thin Eye Clover is an alternative to pancake optics. They showed ways to combine multiple displays for a wide FOV with their “Thin Eyes Crystal” technology. A bit more recently, they showed the Trenza multi-lens-array plus “pixel interlacing” for increasing the center resolution of a “foveal” display.

Lynx Using Limbak

Lynx used Limbak’s ThinEyes® Clover optics in its new AR headset. I discussed the Lynx prototypes in a video with SadlyItsBradley about AR passthrough (the link points to the Lynx discussion in the 30-minute video).

Frankly, I have a hard time seeing an advantage of Clover Optics over pancake optics other than light/power efficiency. The image quality suffers from how the image must be replicated and combined. You can see the seams with the discontinuous optics (discussed in the SadlyItsBradley video).

Following Up On My Comment From Last Time

Yet another set of optical technology disappears into a big company, showing part of the problem of these big companies monopolizing new technology developments I discussed in my comments last time. Lynx is just going to market, and a big competitor now owns the technology they are using.

Lynx probably has the right to keep making their product, but who will support them in future generations with new and improved designs? They will be left to look elsewhere for optics and hope whatever they decide next does not get bought too. As I wrote last time, it helps explain some of Snap’s behavior in buying companies before Meta or someone else buys them.

Karl Guttag
Karl Guttag
Articles: 244


  1. Glad to see you writing on the blog again lately 🙂
    Another advantage (not technical but still important) compared to pancake optics is the cost. Pancake lenses are expensive to make!

    • Stan, I have the pieces of the two lenses used in the iFixit teardown of the Meta Quest Pro (MQP). They look reasonably simple to me, especially compared to the Limbak Cloverleaf, but I have not seen an analysis of the comparative cost to make. The only complication is that they have what looks like a film wire grid polarizer that supports both reflective and transmissive polarized light and some quart waveplates.

      The obvious downside is pancake optics’ efficiency, as they throw away a lot of light with the passing and reflection of the light of the 50/50 mirror and polarizer. Also, there is a tendency to want to use smaller LCDs which lose efficiency due to the iris effect of the individual sub-pixels with the transistor, wire, and spacing structure blocking so much light, which becomes a higher percentage as pixels get smaller. OLEDs have different issues, but it starts with losing about 60% of the light due to polarization (LCDs already have polarization).

      Efficiency has an often overlooked set of costs. Most people think of the battery, but as big a problem can be heat removal.

  2. “As I wrote last time, it helps explain some of Snap’s behavior in buying companies before Meta or someone else buys them.” – note that even Meta’s behavior can be explained with similar reasoning. Boz mentioned that they heard Apple was considering buying Within, so they panicked and made their own bid. There’s always a bigger fish to look out for… Apple is 10x the size of Meta while Meta is 10x the size of Snap (… roughly at least). I’m sure there are some tense power games behind the scences.

    • If the FTC did not stop Facebook from buying Instagram, they would certainly not worry about Meta and Apple buying up a bunch of small companies almost nobody has heard of before.

      The FTC is feckless and does what politicians and their lobbyists tell them to do and collect a paycheck. This is backed up by the Supreme Court, which applies “The Consumer Welfare Standard” (, which uses a shortsighted standard of whether mergers will cause imminent harm to the consumer and ignores the long-term impact and undermining of competitiveness.

      Google effectively bought the US Patent Office and installed their patent lawyer as head of the US Patent Office. They successfully undermined much of the US Patent system to tilt it against small companies. When it comes to protecting American innovation via small companies, the foxes are guarding the hen house. The startups that benefitted from the anti-trust action in ATT’s breakup and its impact on IBM in the early 1980s to not crush all the PC hardware and software startups are now unfettered monopolies and oligopolies that are bigger and more powerful than the 1980’s corporate giants.

Leave a Reply