Simular VR: Lenses and Schematics | Hacker News Reader

Simular VR: Lenses and Schematics

simulavr.com

98 pointsBiteCode_dev53 comments

I've worked full time for the last 3 days on a quest 2 using Immersed. Although I'm liking the experience, it's still not as good as my regular 3 monitor setup. But there are already a lot of advantages and after this brief period, I firmly believe that this is going to be a very common setup in 2 or 3 years at most. I really hope Simula succeeds as I'd love an open and not-attached-to-facebook alternative to my current setup. Sadly, their current offerings are a little out of my price range.

I've been in the same position as you, also on Immersed, and I just can't quite get used to reading text and editing code in VR.

I think part of it is that you have to move your head not your eyes, and that's not how I tend to read.

I have been really impressed with the hardware though. Games are good. Tracking is super precise. The way passthrough video is handled is really clever.

I want it to be my multi-monitor setup but it's just not quite doing it for me yet.

> I think part of it is that you have to move your head not your eyes, and that's not how I tend to read.

We're pushing hard on high pixel density to get around this issue. For comparison to the Quest 2, we have basically 70%+ better pixel density, which allows you to compress more text into a smaller area without having to crank your neck around to read everything.[1][2]

[1] For specs comparison: https://news.ycombinator.com/item?id=29699702

[2] For technical details on how we achieve such high PPD: https://simulavr.com/blog/ppd-optics/

How big is the 'sweet spot' of the lenses? I run an OG Samsung Odyssey HMD and while it's nicer than any other headset I've used (although a mate just got an Index so I'll get to try that sometime soon) I can still only look left/right by ~20° before I start losing visual quality.

You can gaze anywhere without significant quality loss (slight PPD reduction past 20 degrees but it's not really noticeable and it's uncomfortable to have your eyes that far away from center)

This could be the key. In the oculus you can't really move away from centre (for text) and I'm finding it weird.

Appreciate your support!

> Sadly, their current offerings are a little out of my price range.

Yeah, our pricing is one of our project's biggest weaknesses. Unfortunately, upfront capital costs and low scale basically force things to be this way.

With that said, our VR computer is still reasonably competitive against other premium office laptops and ultra-premium (high PPD) VR/AR gaming headsets. We made a post about this here.[1] TLDR: we're offering a headset as good as the tethered Varjo Aero (priced at $2K), with a premium-spec computer as good as a top-shelf Lenovo X1 Carbon bolted on as well (making up for the remaining $699 in price).

Finally: if it helps, we made a $100 off discount code you can apply at checkout: "DISCOUNT_HN". This applies even if you opt for the partial deposit ($1,499 now; $1,499 later) pre-order option.

[1] https://simulavr.com/blog/why-is-the-simula-one-so-expensive...

I totally agree with you. To clarify, your device is too expensive for me because I live in a country which has huge import taxes and a currency which is really devaluated in relation to the US dollar. If I could purchase it without paying those taxes , I would have already made a preorder. This was not a complaint about the price .

This recent article by Rick Casteel (XRforWork on Facebook) is a good overview of the current software/hardware options for Personal Productivity Using Virtual Reality [1]

Immersed is excellent now with the keyboard visibility and I also like vSpatial but it's all still highly dependant on your individual app behaviors like how they do scrolling/zooming/spawning windows etc. so lots of trial and error.

[1] https://medium.com/xr4work/personal-productivity-using-virtu...

> Immersed is excellent now with the keyboard visibility

Is it though? the overlay keyboard never seemed quite accurate enough, and with passthrough you cant see what the keys are. Maybe for touch typing its ok, but for us mortals its simply not good enough.

If this concept takes off it could be an interesting leveler (and level-up) for productivity, especially with people working at home.

I'm sure there are loads of people out there working off a single laptop screen, maybe with an extra monitor if their employer has funded one and they have the space and technical ability to set that up correctly [0]. They're spending all day switching their tiny screen betwen Excel and Word and a browser and email and chat, and it's just killing productivity.

If all it takes is a single piece of equipment, with one or two cables, to give any person an enormous virtual workspace - that could be a huge change. Everyone gets a laptop and a VR headset, so everyone is on the same level for that resource. Everyone has space for VR, so it's not lucky me with the full home office against unlucky you who works at their dining room table or on the couch.

[0] Don't underestimate how complex this stuff is for many people - if they're even aware that it's possible!

We agree :) Our goal is for the Simula One to be so good that it's the strictly-more-productive choice over a single-screen laptop (and also over most PC setups too). Here's our pitch for why VR Computers can boost your productivity:

- Provide unlimited screens of any size

- Improve work focus & immersion (like "noise-cancelling headphones for your eyes")

- Fully controllable with just a keyboard (with hand gestures, eye gaze tracking, etc)

- Allow for fully portable, multi-screen computing

- Usable outdoors (no laptop glare) in places like backyards, urban balconies, & parks

- Provide more privacy (no one around you can snoop on your screen)

- Compact design takes up less desk space than PCs/Laptops

- Promote better posture & freedom of movement: with a VR Computer (VRC) you can change positions, sit up, lean back, stand, lie down, or even walk while you compute

Whats the point of being outdoors if you can't see anything?

That's what the AR passthrough cameras are for!

Fresh air, feel the weather, hear the sounds. Some people can probably crack open a window and get that, but some might need to go for a walk to the park or something like that. It would also be a game changer to be able to take a fully fledged setup to co-working spaces, cafes and so on.

I have aesthetic hesitations toward all this which is really shallow I know, you will probably look like a dork sitting with VR on at a cafe. But imagine where this could lead, if the technology keeps getting more compact or even just popular enough that it's the norm.

> Fully controllable with just a keyboard (with hand gestures, eye gaze tracking, etc)

If it is precise enough, this would be very pleasant to work with, to the extent that you describe, a productivity boost over working with monitors and a mouse.

Glancing at the start button, it being detected with no latency and high precision, and then I press say CTRL + ENTER and I get a left click.

> Don't underestimate how complex this stuff is for many people - if they're even aware that it's possible!

If pluging a monitor to a laptop is too complex, how would pluging/synching a VR headset be simpler ?

I see benefits to VR, but I don't think it will have mass appeal or bring productivity gains in any simple matter (put differently, "no pain no gain")

It's not just plugging in the monitor, it's setting everything up - are you doing screen cloning or expanding your desktop? How do you move windows between screens? How do you configure your OS to know where one display is positioned relative to another (so that windows can be dragged).

I don't want to imply that most people are incapable of figuring this out. But I do think it's potentially hard to find and properly configure your display settings in Windows (which is where corporate is) - if you even have access to such.

This is kind of thing is why people buy iMacs, instead of a laptop and nice external monitor.

The Quest 2 is wireless, and I think VR will have to be wireless to really succeed. At that point it'll be more like "just put on the headset" than "connect the headset to the laptop". The applications will run on-board or stream from the cloud, which is already a thing that works well.

Setting up VR rigs is indeed painful. Just in case this isn't clear though: the Simula One is a portable VR headset, designed to be immediately usable out of the box (with no/minimal setup pain). It features a detacachable compute pack running our Linux VR distro.

With that said, it's admittedly a bleeding edge early adopter product, so won't necessarily be as user friendly as mature products on the market (e.g. office laptops running Windows).

I would hope to use something similar to i3wm or other tiling window managers, so that I could concisely control everything from the keyboard. That might be leaving some user experience on the table though, as I'm not yet sure what's possible with VR and gestures. Could I work like it's minority report but with a keyboard in front of me for input and shortcuts? Because that would actually be rad.

If I'm not mistaken, their included software distro has some opinions on all this, I'm sure they're working just as hard on the UX of the software as they are on the hardware.

For those that didn't know what this is:

> The Simula One is an office-focused, standalone VR headset built on top of Linux Desktop. It provides comparable functionality to any Linux laptop or PC, but with the power of VR.

> 2448x2448 per eye display resolution

That's a fair bit higher than a Quest 2. That might be enough for productivity based on my own subjective tests.

I hope they succeed. There's a lot of potential.

I want it to succeed too, VR recently has become really quite impressive in many regards.

I do wonder if it's overly ambitious. There's a Linux distro to maintain as part of this project as well as a huge hardware project.

Thanks for your support!

RE our ambition: the bulk of our work lies in hardware stuff going forward. Our VR compositor is already pretty stable, and people have been using it on older Linux compatible headsets (HTC Vive, Valve Index) for a few years now.[1] Our Linux distro will be a NixOS fork which uses nix as our package manager, and which bolts our VR compositor on top as the window manager. This choice will allow users to roll back to working configurations pretty easily if some sort of disaster ever occurs.

With that said, this is an ambitious project. We're posting weekly (mostly technical) updates to help build trust that this is a real project making progress towards its target ship date (Q4 2022 - Q1 2023; review units incoming in a month or two).

If there's a topic or question in the back of your mind of what's going on behind the scenes, let me know and I'll add it to our queue!

[1] https://github.com/SimulaVR/Simula

While I love Linux and run it on my home stuff, at work it’s going to basically have to run Windows. Is that going to be an option ever?

If Microsoft decides to give us access to the APIs to implement a window manager (however you'd do that on Windows...), possible. As is, would require a significant redesign of our approach.

That being said, tethered mode will be Windows compatible.

By and large, having access to Windows APIs is rarely a problem. Since Vista they have locked down the OS a bit so you may have to ask for more permissions (or straight admin). More importantly is: are all APIs you need documented? You can still use undocumented calls but they won't be stable.

Replacement 'window managers' were a bit of a craze years ago. I remember running a version of OpenSTEP.

Why the need to fork NixOS? Why not contribute?

We're 2448x2448 and use optics that magnify the center more than the edges, leading to even higher resolution (36 PPD, about equivalent to a 27" 1920x1080 monitor 60cm away)[1]

[1] https://simulavr.com/blog/ppd-optics/

Typo in title, it’s simula. No r.

Yes. Would be helpful if OP or a mod could fix this =] Title should ideally be "SimulaVR: Lenses and Schematics".

I recently tried the Quest 2 and was blown away by the level of isolation from the external world it provides. It's pure concentration. I think VR has lot of potential for productivity (and education).

Regarding the price point of Simula VR my thoughts are that they are only high for a yet unproved product. On the other hand, if the product delivers I could see myself expending 2500€ for personal use. After all I'm evaluating laptops right now for 1/2 the price. And then of course a company can recover that amount multiple times if people is just 5% more productive.

Forgot to mention: I love it's Linux based, as it's my daily driver. I have seen the light but most of the people I know at work have switched to Mac, so I suppose that at some point it would make sense to give that option.

I like the immersive environment concept. And it seems like a good way to take advantage of the way human memory maps to 3D space.

My concern about VR in general is: what distance are your eyes actually focused at?

If they’re focused fairly close, isn’t that bad for eye health / cause eye strain?

You can set the focus wherever you want. For the Quest 2, it's 1.2m [1], for most comfort at arm length. The future is varifocal systems though, since we use focus/depth of blur to judge size [2]. That's a really neat talk. As said there, you want the focal distance to be the same as the object, for least disagreement.

1. https://twitter.com/id_aa_carmack/status/1371485209603022853

2. https://www.youtube.com/watch?v=YWA4gVibKJE

> since we use focus/depth of blur to judge size

I’ve come to the conclusion that this is the factor holding back VR for me. I just don’t feel immersed and everything feels kind of flat and small. Hard to describe.

Yeah, distance gets confusing. Things don't seem as big as they should.

But, I found a workaround for this. In VR chat, I walked around several worlds with my characters forward movement speed set to the same as my treadmill. Slowly (a mile or so), the worlds became huge, it was really surreal. It fixed the scaling for everything, and it lasted about a day afterward. Mountains were distant and huge. Buildings looked far away. It was a significant change.

Things seeming like they are the wrong size is usually a sign of incorrect calibration for inter-pupilary distance.

Adjusting IPD is really annoying at times. Probably one of the main reasons I pushed for auto IPD in our feature set.

I had a huge struggle debugging an issue on HoloLens that turned out to just be mis-calibrated IPD.

Unbeknownst to myself, a coworker had ran the calibration on himself the night before and when I got the device back, suddenly all my graphics of 3D models overlapping real-world objects looked "wrong". Every time I tried to visually adjust the registration of the model to the object, it would be completely unaligned in the z-axis with respect to my view when I performed the alignment. Midway through debugging, I tried adjusting from multiple angles to get the model absolutely centered on the object, which then made it look like it was swimming in the air whenever I walked around it.

I started to think there was something wrong with the device tracking, so I went through every calibration process on the device. I left IPD for last because I was used to working in VR where it doesn't really matter all that much, so I didn't expect it to be that important. IPD calibration on the original HoloLens was also a gigantic pain in the ass, as it took well over a full minute if clicking on objects with that interminable "air tap". Suddenly everything locked into place and was rock solid.

Immediately bought IPD measuring tools, had my own measurements confirmed by my optometrist, and it's now the first thing I check on every new headset. Every auto calibration system I've used regularly gets my IPD be wrong by +-5mm. I don't know if that's a significant difference (it's almost 10%, so seems like it), but I don't take any chances anymore.

Not sure how other headsets do IPD, but with eye tracking it's pretty hard to get it wrong. If the eye's off center it's easily noticeable in the image.

I get this on flat monitors actually. Sometimes I'll come home from a day of being outside, and it's really really clear that it's just a picture on a screen, and it feels really flat, small and hard to immerse in. But usually I can feel like I'm basically in the world, with a good sense of scale and depth.

It makes me think that perhaps it's not physical but an internal interpolation occurring. I am way out of my depth in that discussion, but it feels like a suspension of disbelief but for interpreting visual input.

Yeah, it kind of sucks that every current VR headset has a focal range somewhere between 1.3 and 2m (depending on which headset.) I wish there was some kind of aftermarket lens swap to make that infinity; it would make things like Space Engine in VR way more impressive I think.

Infinite focal distance isn't really doable with the space constraints you have in a headset. We tried, it massively increases complexity for no real gain.

Was there a perceivable difference when looking at something at hand length and "infinity"?

I have a friend with perfect vision, and he has trouble seeing things in video games on a flat screen. I have an eye that does not participate in the general effort, and so I don't have true binocular vision, yet I've always had great hand-eye coordination both in real life and in games, and never had motion sickness. I quietly theorize that this is because I use a multitude of other cues to judge distance, not the "6th sense" of depth that binocular vision provides.

I would love to hear how other people experience sensing depth as it's a really fascinating topic, I have no idea what it's supposed to be like. Do you find you 'feel' depth, like if a ball were coming toward your face does it give you a feeling of something speeding toward you?

About 1-2m in our design.

I'm quite excited about the concept of working in VR. I've tried out Immersed with Quest 2. I really wish I could try SimularVR headset on physically before committing to it though. Quest 2 was cheap enough so I could risk it.

Yeah, I understand your concern here. Unfortunately this isn't possible currently since we're still making the headset =]

We're targeting (stripped down) review units to be available in May-June time frame.[1] While this won't allow you to personally test a Simula One, it'll allow you to see it more tangibly tested by bloggers/YouTubers who can report their experience online.

Also, if there's some particular aspect of the headset you're curious/have concern about, LMK and I'll see if I can provide more detail in an update or over DM.

[1] https://simulavr.com/blog/review-units-timeline/

Thanks for letting me know! The only concern is being able to see text clearly without having to move my head because that quickly got annoying. It does look like SimularVR's resolution + lenses may solve this issue though, so I'm keeping my hopes up.

> We're targeting (stripped down) review units to be available in May-June time frame.

That is very exciting. I'll keep my eyes peeled.

Are there any VR headsets with actual curved displays that closely track the curvature of the eyeball?

Similar to curved monitors…

In general no, the optics in front of the displays creates the curved focal plane. In fact due to pin cushion effects of these optics, the corners of the displays are largely unused/invisible to the user.

See the lens distortion on page 37.

https://stanford.edu/class/ee267/lectures/lecture7.pdf

There's a patent by Samsung on it, but no actual product.

There's always optics in front, so it's not immediately helpful. But a curved surface might simplify lens design.

>There's a patent by Samsung on it, but no actual product.

In a justifiable way or did they get to patent the equivalent of 'small curved display for vr purposes'?

https://uploadvr.com/samsung-wide-fov-patent/

VR purposes. This might be what was used in the Odyssey, come to think of it.

Reply on news.ycombinator.com