Its true I think with Quest 3 we reached to the diminshing returns of VR hardware where more payment doesnt mean alot better experience.... however VR need more work in Software.
With headsets thing are doing okay. But we need better haptics. Once someone makes a good affordable vr glove, that will change everything. At the moment VR is kind of stuck, you can see it but you can't properly interact with it and better visuals and more processing power don't change that. And zuck is only pushing VR this hard so he can get AR small enough that everyone will wear AR glasses all the time. Then he can sell you a load of virtual garbage to overlay on the real world, and that's where the most money will be made.
What is wrong with Valves approach of using a diamond shape of lights/sensors around the eye, and then measuring the brightness of the whites of your eyes? patents.google.com/patent/US20230393654A1/en
I attribute it to having an outlet (my new second channel) to post content/rants without worrying if it will cater to the algorithm. Ever since I came to the conclusion I could do this, I felt way better in the mental health department. And sleeping department lol
Shiftall is a terrible name. EVER SINGLE PERSON that i show my phone to with an article about Shiftall.......when i take the phone away and ask them what the name of the product was and they ALL said SHITFALL.
I use my Quest 3 to project life sized parts from Fusion into where they're supposed to fit in the real world, and using joysticks to manipulate and modify models works pretty well and I like the tactile feedback. That said the only good MR 3D modeling apps cost as much as a vision pro, so the touchpads might work better in a professional setting. Ngl I respect the opinion on touchpads but dont agree
I'd like to know some apps like this aswell. Having 3d printer you can visualise what its gonna look like before you set it for printing for next couple of days lol
2:25 one of the main goals of the project is being able to call people with VR/XR mainly for technicians so that they can see what the other person sees at a site, especially in cases of emergencies/when the site is far, being able to see everything and instruct person on the other end what has to be done to fix the issue and guide him/her trough every step live
Event based cameras work fundamentally different than normal cameras in the method used to produce a "picture". They don't have the concept of "frames" at all, they directly report a change from a pixel. They are more akin to a measuring device. And that's the reason why they are so expensive, every pixel needs its own dedicated processing. Frame based cameras just collect the current pixel state and write it to memory.
@@yvan2563 Yes, that's essentially what's happening now. The drawback is you always have to use the entire picture, since you don't know what's changed. Video compression algorithms are basically sending what's changed each frame and then compress that further. And this results again in a full frame. This produces a boatload of latency for full frame encoding -> full frame decoding -> filtering change again. Event based cameras don't stream full images, you get an uncompressed addressed pixel array what changed without any notable latency whatsoever. Oculus DK2 used USB tracking cameras which had to send the image uncompressed in order to reduce the latency to something usable. This is fine in theory but since the cameras are seldom the only thing plugged in, it often resulted in literally headaches because tracking was spotty because the USB was fully saturated. Fun times 😑 So you had to get a PCI USB host card, just for those cameras alone, preferable with separate host controllers for each port.
I'm not worried at all tbh. My Quest 3 with the kiwi battery strap and the qgo is just fantastic. The Zuck has recently stated he's totally into spending $15,000,000,000 a year for as long as it takes for it to be where it needs to be. The upcoming arms race with Apple will hopefully get us 20 years of progress in the next 5/7. Like i said, having an absolute blast with my pimped out Quest 3. Anyone for Racket Club👌
It's nice seeing more people talk about event cameras. There's two technologies for ideal MR sensors. Event camera hardware is one piece. (Ideal for eye tracking as you saw, but also for getting subtle face movements). Canon has the other technology, a SPAD sensor. (A device that can count individual photons per pixel). Their current MS-500 product is not an event camera configuration and is 21K USD. Combining a SPAD sensor with event hardware to output intensity changes allows for nearly flawless data. (Can also control the output rate variably if they wanted per pixel). At ~10K Hz output it would suffer from no exposure, motion blur, or noise while being low-powered. (Can be modified with metalenses in theory also to extract material data from surfaces, but that's not critical). The big picture is miniaturizing such sensors and mass producing them for MR, robotics, and any application that needs perfect SLAM (even in darkness) and room scanning. If you ever talk to Canon, could see if they're working towards that.
you're gonna be waiting for a very very long time. i been waiting almost 4 years so dont get your hopes up that its coming any time soon. if you see the index go on sale which it never does it may be coming soon after and thats a very big may at this point.
Deckard is Schrodinger's headset, the only reason the Deckard die-hards care so much about it is by virtue of it not being officially announced, no official specs being published and committed to, people can let their own imaginations run wild. Someone who wants Wigig 2 can imagine that Deckard will have it. Someone who wants full human FOV can imagine Deckard will have it. Someone who wants 60-70PPD can imagine Deckard will have it. And so on... There have never been so many good choices for PCVR but for anyone who doesn't see 100% of their dreamed up feature set in anything on their market, they can hold out hope that Deckard will be the thing, even though no one should trust Valve at this point farther than they can throw them.
i used to say the same thing than after 3 years i just started to say it arrives when it does. im sure it will one day find its way to the domestic sales table if other companies dont do something revolutionary enough to send it back to the drawing board@@Shadow_banned_by_RU-vid
A true BradFan* cannot distinguish between those options. Or, in other words, a true BradFan sees "being thanked by Bradley" as a superset of "being slapped by Bradley"? Do BradFans call themselves: MadlyForBradley, GodlyIsBradley, OddlyFondlyOfBradley, BroadlySupportiveOfBradley...? 🤔 *May or my not also be a true Scotsman.
I love my xreal/nreal glasses but I don't see them as VR they are pretty much just an over glorified floating monitor to display content on but they work brilliantly with my steamdeck
@@SadlyItsBradley with a ducky plug in "xr gaming" they do work for so-called VR lite mode "3dof mouse look". But frankly I find that to be annoying as it doesn't handle tilting the head side to side.
I'm genuinely surprised it took this long for someone to FINALLY do event based eye tracking. The first thing I said when I first saw eye tracking emerge in VR years ago was, "Why are they tracking the whole eye? That's computationally expensive. They should be doing event based tracking and just looking for changes." I had no idea that the hardware just hadn't seen much more need for miniaturization from the aerospace and military uses that made me familiar with this method.
Same. CV papers based on event sensors started to show up from 2014 or so but it took 5 more years for the first paper using event sensors to come out for eye tracking. The high price point of an event sensor was quite a turn off for independent investigators like me but I'm super surprised the big 5 haven't released a viable event-based eye tracking product yet.
I'll save you the trouble. AI buzz words for every product. Golf clubs, refrigerators, tvs, etc. Speaking of TVs the big news was a transparent tv tech. Very boring CES imo.
11:30 Makes perfect sense to track pixel changes instead of full camera snapshots. Video compression encoders like Divx have been doing this since the 2000’s
I'm really annoyed with Immersed's management of their product. First they ditch 2.5k because they didn't get enough pre-orders-- What!? The only people pre-ordering aren't the people you would be targeting with a 2.5k, so you've just abandoned an entire underserved market sector for a flimsy reason. Then they say "uhhh, well, that wasn't enough so now we're promising standalone and it's way more expensive" thus taking them from having been an actually interesting product to an overblown project that'll probably never finish. It's Agile management process run amok over there and it's annoying as all hell because I was going to buy the damn thing.
I’m still not sold that eye tracking is “necessary”. Feels like with all of the non gaming headsets coming we’re forgetting that most of us just do that or like VRchat. And is it cool for that stuff? Sure, but I don’t think we should just start including it in every headset if it’s gonna make the prices go up by a ton across the board. My quest 3 is totally fine without it.
Also fun fact that might have flew under your radar - pimax did a live stream from ces, and it was all about 2 main white guys - Kevin and Joshua - leaving the company to start Ingenious VR and become the middle man for pimax headsets. Things must be going extremely downhill there.
I cancelled Vision Pro order after being so excited for it..... its simply overpriced for what you get, I cant figure the purpose of it at all, its for people to say "wow look at that" then forget about it to me....
WHERE ARE YOU! I was worried youtube had done a channel ghost on you but I visited your channel page and you have been MIA! Please let us know you’re ok?!
User experience, hand/eye tracking, and more development needs to be the main focus for XR the hardware is getting there with companies like XREAL and big screen
Now here's a solution: how about UV cured *(LED lightform technology)* user-moldable facial interfaces? The included cushion would come in a "universally neutral" shape but would mold and cure in seconds to the exact shape and contours of the wearer's face once the headset is properly fitted. I imagine the cushion could also be user removable so that another user can swap-in their custom-molded facial cushion (each additional facial interfaces would cost extra of course).
I know this is because of the cognitive gap I have with what I want VR to "be" vs what it is, but it really does demonstrate that computing as a metaphor is built on layers of sand that weren't really designed for what we're doing. What I hate about VR mostly is its pushed as a system primarily for content consumption over content creation and so it feels like a snake eating its own tail to me. Touchpads are fine but I get a lot more use out of sticks. My major complaint with VR is it doesn't have input parity with controllers: I want to have full input plus arm and head position as a game-developer but we're just not in that world and it means making a good 6dof VR experience doesn't really work in existing controllers. We're at a point where tool developers don't have much of an advantage to doing their work in VR 90% of the time, and game-developers can't really add it in a way which is compelling without adding abstractions which create unwanted latency which collapses even slower games into shadows of themselves. I love the concept of VR but thus far its been nothing but a huge disappointment for me, which is why only revisit it periodically and I'm disappointed every time.
For all its faults, I think the WIndows MR control scheme with both a touchpad and a joystick (in addition to grip and triggers ofc) is just peak VR controls tbh. Valve knew what they were doing with the Index control scheme, but I would've dropped the buttons
Event-based cameras coming to VR headsets?! 😮 I so can't wait to get my hands on one, but they cost like $5k right now and resolutions are tiny, hopefully their wider use can change that!
I don't understand why you want event based camera. To me that's add latency instead of removing it. (processing on the camera + processing on the headset)
From what I understand, you need a lot of processing required to use current cameras too. Train heavy models to filter out the unimportant parts of the full image. And at a much higher power/bandwidth as they try to push the refresh rates higher for that
I hope Shiftall pulls off their "beyond 1.5" of sorts. The headstrap and facepad are sad, but that can be changed. The lenses are super high stakes though. Heard something about a DPVR PC focused HMD, so here's hoping it's actually [good]
damn, I only see huge vr headsets, where i see the big problem is that VR is really unconfortable, maybe the only confortable vr headsets are the Bigscreen beyond, but its too expensive for me...
Heya Brad, not really sure where to put this. But Disney just recently put out something called a "Holotile" its a new type of vr omnidirectional treadmill, seems like it flew under the radar. Do you have any thoughts on it?
Heads-up on the Pimax Crystal wireless module. It’s not running at 72hz. The SteamVR performance graph showed 60hz. Pimax claims it’s 72, but it’s not.
I was thinking like "Did he even go to CES?" cause the video wasnt out yet xD But still thanks for covering it! Love to hear your opinions on this tech :D
there were 30 venues and over 130k people attended. i highly doubt that it was lame or lackluster or boring. there are a few videos of some really neat (and really goofy/corny) tech coming out. like ai pillows. you might consider lowering the high and mid frequencies in your mixer a bit and loosening up your compressor a bit. the high pitch frequencies combined with the reverb it captures and then stomped on to immediate nothingness by the compressor or by your editing is quite jarring and amplify the contrast of one another. just a recommendation. great video though, thanks~
got a big hope for that holographic stuff Meta has been working on to get to that Safety Goggles size you saw in Westworld some time soon. If the next big leap takes too long it'll be bad. Also stick all the compute in a battery holster thing. Apple have made that an OK thing, at this point nothing needs to be bigger than a Bigscreen Beyond with a few extra inside out cameras which aren't that heavy. Bung the rest of it into the waist worn battery and processor and be done with it. Also, GREAT WORK!
Where you said that Meganex light is still far from a good fitting product, I saw another video where two Germans said that it was the best headset so far and were really crazy about it.
Maybe if they fix the optics and facial gasket, and strack: I can enjoy it. But it's too early of a prototype I feel, and not was probably more destructive for them to try and demo it to the public
Agree that "AI" is getting tossed around way too much. It's like a rehash of the whole iPhone thing that brought us the "iBox" "iCard" "iCup" etc etc etc. Don't get me started on "Smart" devices.
Dude if you think the "spatial computing" thing was going away keep in mind if Apple took a crap on the sidewalk EVERY tech company would follow suit. ANYTHING Apple does is gonna get cloned. Except making content for your hardware 😂
Right after the 19:00 mark you mentioned that you want a great image if you're going to wear something on your face. What device would you recommend for watching movies on the go like on a road trip, a cruise ship, an airplane flight? Price doesn't matter to me. Which device has image quality that truly WOWS you???
not sure why people putting in AI think venture capitalists will invest, the market is imploding at the moment.. because of venture capitalists borrowing so much money and no returns
