CastAR - An Attempt To Dispel the Myths

  • Dimididales

    Posts: 357

    Posted:
    Edited:
    Posted:
    Edited:

    Does CastAR use shutter glasses in VR mode, or only in AR?

    Only in projected mode. No need for shutters in AR and VR mode.
    NoamLoop2.png
  • Woopate

    Posts: 2573

    Posted:
    Posted:
    Well, for better or worse I've decided to back at the pro level. That 80 dollars international shipping made me flinch (as a Canadian, less than 1/5 of a day's drive from the States, it seems excessive and I wish they had more options), but I'd like to get my hands on VR, and this seems like more bang for my buck than an Oculus Rift(even if the Rift does turn out to be better for VR, I can still use this for a bunch of things).
    Ztj7fnR.png
  • HydraShock

    Posts: 880

    Posted:
    Posted:

    Well, for better or worse I've decided to back at the pro level. That 80 dollars international shipping made me flinch (as a Canadian, less than 1/5 of a day's drive from the States, it seems excessive and I wish they had more options), but I'd like to get my hands on VR, and this seems like more bang for my buck than an Oculus Rift(even if the Rift does turn out to be better for VR, I can still use this for a bunch of things).

    Congrats :)

  • 0ni

    Posts: 1102

    Posted:
    Posted:
    I could see this being great for simpits if you can cut and shape that backdrop thing to fit where windows are. This killer for me though is the FoV. 90° from the rift was already less then I'd like. That 210° InfinitEye looks promising in that respect but it obviously still has some maturing to do.

    I do appreciate the link AR could potentially provide between real world physical control panels and the game screen through a "window" but without full field of view it seems like it'd be a serious immersion killer. The rift and things like it provide that uncompromising cut off factor from the real world that really works to help establish that immersed feeling, seems like it's a lot less FoV dependent in a way.

    Anyway, I'll ultimately go with whatever has the best FoV. As a simpit builder though I hope castAR can provide. Until I know more I'll remain skeptical as always though. Seems this device hasn't gotten that much exposure/tesing yet so I wait patiently.

    Highly deceptive thread title btw, definitely find it misleading. I read the threads before and I don't think y'all need to be so defensive ( and neither did they). This does make it seem like you have some underlying agenda. All concerns about any product are legitimate until they are in wide review. That applies to both castAR and Oculus Rift consumer equally. I'd appreciate it if you renamed the thread simply "castAR" or something more appropriate for the attitude the hardware subforum should be trying to cultivate. Not trying to attack you OP but I like my hardware to not be like general. :p
    Cascadia_Sig.png

    Ahhh, the internet. Where Men are Men, Women are Men, and kids are usually cops.

    -Captain_Trips
  • oninoshiko

    Posts: 940

    Posted:
    Posted:

    The down side of OR is that they are not promising 6DoF, only 3DoF (only rotation, not translation). While we expect they will get it to the 6DoF point eventually, we don't know if it will be in the first production version or not. (unless I've missed an update, which is quite possible)

    Consumer Version has always been promised with 6DoF tracking. Oculus people are smart and know without it its' dead in the water. Afaik they haven't decided fully what tech they will end up using. But let's be honest. The castAR tracking tech ain't exactly rocket science and easily replicated. Oculus just hired a new guy based on his work on camera based tracking:

    http://www.mtbs3d.com/phpBB/viewtopic.php?f=138&t=18148


    Consumer Version will have 6 DoF. That's a given!
    I stand corrected. Their website doesn't say anything to this effect (and I couldn't find anything useful in that forum link either), but their KS did say 6DoF. I apologize for my mistake.
  • HydraShock

    Posts: 880

    Posted:
    Posted:
    0ni | 0ni said:

    I could see this being great for simpits if you can cut and shape that backdrop thing to fit where windows are. This killer for me though is the FoV. 90° from the rift was already less then I'd like. That 210° InfinitEye looks promising in that respect but it obviously still has some maturing to do.

    I do appreciate the link AR could potentially provide between real world physical control panels and the game screen through a "window" but without full field of view it seems like it'd be a serious immersion killer. The rift and things like it provide that uncompromising cut off factor from the real world that really works to help establish that immersed feeling, seems like it's a lot less FoV dependent in a way.

    Anyway, I'll ultimately go with whatever has the best FoV. As a simpit builder though I hope castAR can provide. Until I know more I'll remain skeptical as always though. Seems this device hasn't gotten that much exposure/tesing yet so I wait patiently.

    Highly deceptive thread title btw, definitely find it misleading. I read the threads before and I don't think y'all need to be so defensive ( and neither did they). This does make it seem like you have some underlying agenda. All concerns about any product are legitimate until they are in wide review. That applies to both castAR and Oculus Rift consumer equally. I'd appreciate it if you renamed the thread simply "castAR" or something more appropriate for the attitude the hardware subforum should be trying to cultivate. Not trying to attack you OP but I like my hardware to not be like general. :p

    So I want you to be very careful about what people are reporting for field of view. The AR is going to have the same FOV as the Rift (Both are 90 degrees per eye). Lets also be careful to understand what FOV even means. To clarify, a standard monitor that displays 1280x720 resolution (as an example nothing more) gives you the SAME field of view as the the Rift (or AR). The amount of "in game world" you see does NOT change simply because you moved the image closer to your face. All that has happened is now your entire vision is filled with that image that was once on your monitor, only taking up 1/3rd of your "true" vision. So lets be honest here, your not going to SEE any more in any device unless they are able to put additional video feeds to a different display in your peripheral. This would be similar to taking a 3 monitor setup (the center being what we normally see, and the side monitors showing what you would see if you turned your head slightly to the left/right), then putting those right up in your face. In that case you would truly be seeing 3x as much as you were before (effectively 3840x720). NONE of the devices we are talking about can/are do that right now. Until you see a device claiming they have 3840x720, or some other hugely wide resolution, they arent doing anything different from anyone else. So that being said, the same image that the rift is placing in your face is going to look nearly identical to what you will see in the VR clipon for the AR. There is no significant difference in what you actually see. Both images will extend roughly the same distance left/right in your eye's ACTUAL FOV. There are SOOO many people who fail to conceptualize this. They keep seeing how the rift "fills your fov", but what they fail to understand is its the same damn image they always saw, just up close and in 3d. Again, no current device is adding in additional peripheral pixels. They are tricking your brain into thinking it has a additional peripheral when it really doesnt. Quite honestly, if you have a big enough monitor, and you sit close enough, your brain will be tricked in the same way (trust me I experience it every day on my 120" projector screen when I play games on it, if it were 3d it would be VERY immerisive). What makes these devices more immersive that a monitor which is large and up close is two fold. Its the fact that you now have 3D and the image tracks with your head movement. The only fundamental difference between the OR and AR is that the OR has a helmet/goggle like attachment which is meant to block out all light, and fits tightly to your face. The AR on the other hand will put the 3d image roughly the same distance from your eyes, but will be viewed through glasses (rather than the lenses that the OR uses). In the end, both devices are accomplishing the exact same thing....

    PHEW.....Hopefully someone understood that, because I have repeated this concept 1000x and very few seem to actually grasp what I am saying.
  • Dimididales

    Posts: 357

    Posted:
    Posted:
    0ni | 0ni said:

    I could see this being great for simpits if you can cut and shape that backdrop thing to fit where windows are. This killer for me though is the FoV. 90° from the rift was already less then I'd like. That 210° InfinitEye looks promising in that respect but it obviously still has some maturing to do.

    For FullHD (1920 pixel horizontally) 210 horizontal degrees would correspond to a pixel width of 2 mm in 1m distance - not sure whether this still counts as "immersive" :)
    With these angles countering distortion also likely becomes a huge problem. And many game engines don't allow such big camera angles, so it probably wouldn't work out of the box for most existing games.

    So we will probably have to wait a few years until glasses with these properties can be built in a way that it is really fun to use them.
    NoamLoop2.png
  • Dimididales

    Posts: 357

    Posted:
    Posted:
    @HydraShock: Not sure whether I understood your explanation. For a big FOV the angle of the virtual camera has to be correspondigly adjusted, so you don't see the same image as on your 1280x720 monitor - a much larger area is compressed to the same number of pixels, you trade resolution per degree with a higher number of degrees.
    NoamLoop2.png
  • HydraShock

    Posts: 880

    Posted:
    Edited:
    Posted:
    Edited:

    @HydraShock: Not sure whether I understood your explanation. For a big FOV the angle of the virtual camera has to be correspondigly adjusted, so you don't see the same image as on your 1280x720 monitor - a much larger area is compressed to the same number of pixels, you trade resolution per degree with a higher number of degrees.

    Are you sure they arent just adjusting the aspect...meaning pulling the camera back slightly? How would you increase the viewing angle and displaying it on the same screen w/out warping the view if the camera stayed in place. It would be like taking a panoramic picture and forcing it on a screen that is too narrow to view it unaltered. You would either have to put bars on the top/bottom, or stretch it so it looked like ass.....

    When you look at the image shape in the oculus, it certainly does NOT look like a panoramic view. It looks basically like a warped version of the same view you would get on a regular monitor.
  • HydraShock

    Posts: 880

    Posted:
    Edited:
    Posted:
    Edited:
    ..

  • Dimididales

    Posts: 357

    Posted:
    Posted:
    Just download Unity3D (or any other 3D engine of your choice), click on the camera, set projection to "Perspective" (which is the default setting) and change the field of view slider. You will see a graphical representation of the FOV in the 3D scene as well as a preview rendering of the resulting bitmap.
    The position of the camera doesn't move at all. This greatly differs from just moving back the camera as the angle between the rays changes, so you get a "fish-eye effect".
    NoamLoop2.png
  • PalmerLuckey

    Freelancer

    Posts: 2

    Posted:

    Freelancer

    Posted:


    Guess what, the CastAR doesn’t need a mat either. In fact this device can be used to make your NORMAL monitor into 3d as well, on top of the ability to go full VR (just like the rift), without the use of any projection mat/screen.


    -The tracking on the CastAR is curently far superior to the Rift. The Rift uses gyros, and induce a significant amount of lag (a serious issue the rift folks have promised to remedy). The Cast AR does not use gyros, it uses triangulation via IR LEDs. This triangulation is basically zero lag, and therefore causes much less motion sickness. The CastAR also does 6DOF tracking (which includes side to side/forward and back translation), which the Rift does not CURRENTLY support. There are rumors that the OR will support 6DOF tracking, but we have yet to see it in a working prototype. Either way, the castAR will feature it in their first release

    Full disclosure, I am the creator of the Rift. You are a bit off on a few things, I just wanted to clarify where I can.

    The gyros in the Rift do not introduce significant lag, they are actually one of the fastest things in the pipeline. We sample internally at 8000hz and report the smoothed samples over USB at 1000hz, so there is only 1ms between updates. The total latency of the tracker from motion to PC is about 2ms. It is definitely not accurate to say that CastAR has almost zero lag when compared to the Rift, the optical system they are using is innately 8 times slower than the Rift tracker, not including the time it will take to do image processing (which is minimal, since they are doing that in hardware). Full 6DOF tracking is definitely important, though, it does a lot to reduce simulator sickness! You are also partially wrong about being able to use CastAR without setting up a mat, you still need to set up markers for the optical system in whatever direction you plan on looking. Looking around 360x360 needs visible markers on all the walls, ceiling, and floor. I think they mentioned that they might build a gyro into the final product and use it in tandem with the optical system, which is a great way to do things.

    I am excited to see the final CastAR, I backed the Kickstarter for $10k. Can't wait to see what they end up with!
  • oninoshiko

    Posts: 940

    Posted:
    Posted:


    Guess what, the CastAR doesn’t need a mat either. In fact this device can be used to make your NORMAL monitor into 3d as well, on top of the ability to go full VR (just like the rift), without the use of any projection mat/screen.


    -The tracking on the CastAR is curently far superior to the Rift. The Rift uses gyros, and induce a significant amount of lag (a serious issue the rift folks have promised to remedy). The Cast AR does not use gyros, it uses triangulation via IR LEDs. This triangulation is basically zero lag, and therefore causes much less motion sickness. The CastAR also does 6DOF tracking (which includes side to side/forward and back translation), which the Rift does not CURRENTLY support. There are rumors that the OR will support 6DOF tracking, but we have yet to see it in a working prototype. Either way, the castAR will feature it in their first release

    Full disclosure, I am the creator of the Rift. You are a bit off on a few things, I just wanted to clarify where I can.

    The gyros in the Rift do not introduce significant lag, they are actually one of the fastest things in the pipeline. We sample internally at 8000hz and report the smoothed samples over USB at 1000hz, so there is only 1ms between updates. The total latency of the tracker from motion to PC is about 2ms. It is definitely not accurate to say that CastAR has almost zero lag when compared to the Rift, the optical system they are using is innately 8 times slower than the Rift tracker, not including the time it will take to do image processing (which is minimal, since they are doing that in hardware). Full 6DOF tracking is definitely important, though, it does a lot to reduce simulator sickness! You are also partially wrong about being able to use CastAR without setting up a mat, you still need to set up markers for the optical system in whatever direction you plan on looking. Looking around 360x360 needs visible markers on all the walls, ceiling, and floor. I think they mentioned that they might build a gyro into the final product and use it in tandem with the optical system, which is a great way to do things.

    I am excited to see the final CastAR, I backed the Kickstarter for $10k. Can't wait to see what they end up with!
    @PalmerLuckey: Thank you for clarifying. It's always good to hear from an authoritative, up-to-date source!

    Since you happen to be here (sorry for going a little OT, guys), the biggest concern I have with conflicting vr solutions we have at this stage in the life of VR is that from a game developer's perspetive getting support for all of the major players seems like it would be difficult. Has there been any talk of developing a common API that would allow developers to support muliple different brands of VR-devices?

  • Woopate

    Posts: 2573

    Posted:
    Posted:


    Guess what, the CastAR doesn’t need a mat either. In fact this device can be used to make your NORMAL monitor into 3d as well, on top of the ability to go full VR (just like the rift), without the use of any projection mat/screen.


    -The tracking on the CastAR is curently far superior to the Rift. The Rift uses gyros, and induce a significant amount of lag (a serious issue the rift folks have promised to remedy). The Cast AR does not use gyros, it uses triangulation via IR LEDs. This triangulation is basically zero lag, and therefore causes much less motion sickness. The CastAR also does 6DOF tracking (which includes side to side/forward and back translation), which the Rift does not CURRENTLY support. There are rumors that the OR will support 6DOF tracking, but we have yet to see it in a working prototype. Either way, the castAR will feature it in their first release

    Full disclosure, I am the creator of the Rift. You are a bit off on a few things, I just wanted to clarify where I can.

    The gyros in the Rift do not introduce significant lag, they are actually one of the fastest things in the pipeline. We sample internally at 8000hz and report the smoothed samples over USB at 1000hz, so there is only 1ms between updates. The total latency of the tracker from motion to PC is about 2ms. It is definitely not accurate to say that CastAR has almost zero lag when compared to the Rift, the optical system they are using is innately 8 times slower than the Rift tracker, not including the time it will take to do image processing (which is minimal, since they are doing that in hardware). Full 6DOF tracking is definitely important, though, it does a lot to reduce simulator sickness! You are also partially wrong about being able to use CastAR without setting up a mat, you still need to set up markers for the optical system in whatever direction you plan on looking. Looking around 360x360 needs visible markers on all the walls, ceiling, and floor. I think they mentioned that they might build a gyro into the final product and use it in tandem with the optical system, which is a great way to do things.

    I am excited to see the final CastAR, I backed the Kickstarter for $10k. Can't wait to see what they end up with!
    Whoa cool. I love what you guys are doing. I think it's likely to change a number of industries in significant ways. I also can apparently look forward to flying with you (maybe).

    I think that technology has a tendency to "want" to come about. Not necessarily that it has a will of it's own, but the state of the industry in general leads to certain options becoming feasible, and as soon as they are feasible, large numbers of people developing on those projects pop up. CastAR is an interesting subject in this regard though. It's using a bunch of different pieces than other VR setups. Off the top of my head the camera based tracking seems to be the closest thing to an "Inside the box" solution implemented by CastAR. We haven't seen retroreflectors being used in a significant way, and a lot of people dismissed pico projectors because of their poor image quality on regular surfaces. Combining the two really is an interesting innovation.
    Ztj7fnR.png
  • PalmerLuckey

    Freelancer

    Posts: 2

    Posted:

    Freelancer

    Posted:

    Has there been any talk of developing a common API that would allow developers to support muliple different brands of VR-devices?

    Certainly. Developing standards in such a young and fast moving industry is very difficult, though. Until the industry figures out the best way to do things, it is tough to decide what features should be included in the standard, and how they should be implemented. For example, look at how video format standards have held back television/monitor display technology; standardizing 1080P resolution across the industry was a good move at the time, but it really held back display advancement beyond that point. That is only a resolution standard, imagine how complicated things get when the standard needs to include input devices, head tracking, stereo rendering, physical eye/body measurements, and much more across a wide variety of engines! It will happen someday, but I don't know when that day is.

    I also can apparently look forward to flying with you (maybe).

    Definitely, I can't wait to fly my Hornet using the Rift. I have spent hundreds of dollars on Star Citizen already, I need to get the most out of it!

  • Dimididales

    Posts: 357

    Posted:
    Posted:
    @PalmerLuckey Good you jumped in! Being able to make an informed decision is so much more worthwhile than FUD-wars.

    I like @oninoshiko's proposal very much. One "Open VR" standard would make the life for developers much easier and help to spread native support for VR devices. And it would specifically help the small vr solution creators (what both Oculus and castAR people are): It is not unlikely that [enter your billion dollar company of choice] might release their own "EA VR glasses" as soon as it becomes obvious that this is a way to make money and others have done the R&D for them. And they will probably have an easy game etablishing their own proprtietary "standard" if the developer community is split between several incompatible solutions when they enter the field.
    NoamLoop2.png
  • HydraShock

    Posts: 880

    Posted:
    Posted:


    Guess what, the CastAR doesn’t need a mat either. In fact this device can be used to make your NORMAL monitor into 3d as well, on top of the ability to go full VR (just like the rift), without the use of any projection mat/screen.


    -The tracking on the CastAR is curently far superior to the Rift. The Rift uses gyros, and induce a significant amount of lag (a serious issue the rift folks have promised to remedy). The Cast AR does not use gyros it uses triangulation via IR LEDs. This triangulation is basically zero lag, and therefore causes much less motion sickness. The CastAR also does 6DOF tracking (which includes side to side/forward and back translation), which the Rift does not CURRENTLY support. There are rumors that the OR will support 6DOF tracking, but we have yet to see it in a working prototype. Either way, the castAR will feature it in their first release

    Full disclosure, I am the creator of the Rift. You are a bit off on a few things, I just wanted to clarify where I can.

    The gyros in the Rift do not introduce significant lag, they are actually one of the fastest things in the pipeline. We sample internally at 8000hz and report the smoothed samples over USB at 1000hz, so there is only 1ms between updates. The total latency of the tracker from motion to PC is about 2ms. It is definitely not accurate to say that CastAR has almost zero lag when compared to the Rift, the optical system they are using is innately 8 times slower than the Rift tracker, not including the time it will take to do image processing (which is minimal, since they are doing that in hardware). Full 6DOF tracking is definitely important, though, it does a lot to reduce simulator sickness! You are also partially wrong about being able to use CastAR without setting up a mat, you still need to set up markers for the optical system in whatever direction you plan on looking. Looking around 360x360 needs visible markers on all the walls, ceiling, and floor. I think they mentioned that they might build a gyro into the final product and use it in tandem with the optical system, which is a great way to do things.

    I am excited to see the final CastAR, I backed the Kickstarter for $10k. Can't wait to see what they end up with!
    Fair enough Palmer. That is awesome of you to come here and correct details. I am not trying to bash your product, in fact I think it will be amazing. I will reword that portion to more accurately reflect your specs. Unfortunately I dont have the specs of the cast AR tracking system, I was merely going by what I have heard them say, and that is that their tracking system was faster than gyro based tracking. I will try to pin them down on that :).

    As for the mats, I think what I was saying is still accurate. The mats are not the leds. The led panels are a separate part, which I agree IS necessary for the AR to work, but a projection surface is only needed when in "cast" mode.

    I also think that was awesome of you to back the AR kickstarter for $10k. Thats really awesome of you, and it does show that you are a good dude (not that anyone has ever said any different), and genuinely interested in moving VR forward. There is no denying, regardless of whos device does what, you are a huge reason for this recent leap forward in VR, and we all thank you for that!
  • HydraShock

    Posts: 880

    Posted:
    Posted:
    So now that we got folks like Palmer dropping by...I think its time for the Godfather himself to weigh in.

    We need CR to try this thing and give us a thumbs up or thumbs down. If he feels this is going to be awesome, I think all this back and forth is going to prove pointless :).

    We also need to know the likelihood of the AR being supported by SC. After all, if its not supported, there is really no point in furthering this discussion either eh?

    SO IM CALLING YOU OUT MR ROBERTS... WE NEED YOU!!!!
  • Soulfly

    Posts: 6

    Posted:
    Posted:
    Heck yeah! Next stretch goal could be CastAR support!

    Chris Robert's hear our plea!
  • 0ni

    Posts: 1102

    Posted:
    Posted:
    @Dimididales & @HydraShock
    I'm well aware that moving a screen closer to your face doesn't make magic pixels appear. This isn't my first rodeo as it were. :p
    I've in the past had a triple projector setup at just over 180° surround and I can tell you that FoV does more for immersion then pixel count. Obviously if the resolution is too low it's a bummer. I would never go back to gaming at 1024x768 for instance. Though filling your view is infinitely more important for immersion then pixel density, at least for me. Hence why I said I'll probably go with whatever has the widest FoV, because that's simply what matters to me. I always go with what makes the most sense for me, why should I buy something because lots of people think it's great when a different product has the features that matter to me? Barring of course some huge failure in another area. If the oculus came out with a 200° FoV but 12ms latency obviously I'd skip it, but as it's been apparently officially proclaimed to be 2 that really narrows down the margin for error,

    Price is also a factor, if anyone charges a crazy/unjust premium I vote that shit down with my wallet. So at the end of the day, if castAR can't beat Oculus for FoV I'm skipping it unless some hard facts come to light that prove it significantly better in other categories.

    I don't know why but y'all make me want to avoid the product for some reason. This thread is a real turn off for me, which is weird because I was really interested in the product before. I guess I still am, I dunno. Thread title still sounds like "castAR: I am right everyone else is wrong". Maybe I'm just getting the wrong vibe.
    Cascadia_Sig.png

    Ahhh, the internet. Where Men are Men, Women are Men, and kids are usually cops.

    -Captain_Trips
  • WavestarInteractive

    Civilian

    Posts: 24

    Posted:

    Civilian

    Posted:

    Why would you honestly go with active shutter technology, the could have easily made the glasses polarized and made it passive, with that resolution that would have been way better than any active shutters will be. I've used 120Hz 200Hz and even 240Hz shutter glasses, the all make me dizzy within a matter of seconds and make me vomit in under a minute. Many people don't notice but some do. Active shutters are old technology and need to die.

    Im going to say it again, this is NOT active shutter technology. The 3d is accomplished by polorization. The Shutter is used in the "projection" mode so it removes the cross talk that occurs with polarized 3d (common problem). This can also be disabled in VR mode.

    Then someone needs to tell the pitchers that since the kickstarter is quite clear that this is active shutter because they used a cheap reflective material for the glasses, having used many different forms of passive 3D, none of them require shutters unless they use a single projector because 2 projectors equals no crosstalk. If they use a single projector, then they are still active shutter technology simply moving the shutter to the projection end causing the same vertigo issues during any fast moving action that all other active shutter technology experiences.

    The data given on kickstarter is puzzling to say the least and makes me question the validity of the product. It's as if someone from a marketing department said people really like the occulus rift so I'm going to collect a bunch of specification for 3D and VR technology from a google search, splice them together using a little bit of knowledge but no understanding how it works and sell it as a product on crowdfunding where no one will do due diligence. I mean one of the photo's of the device is grabbed straight off the Vuzix website. I certainly would not buy one until I could test it hands on as I simply don't trust what I see so far.
    To clarify,

    It is and I will use ALL-CAPS:

    It uses:

    POLARIZED ACTIVE-SHUTTER technology.

    Second:
    Normal Active-shutter glasses (polarized or not) only has a duty cycle of 2-5% per eye and clocked at 60 Hz. Even those at 240 Hz which is effectively 5-12 FPS.

    castAR uses a much faster duty cycle where it drives each eye at 50% of the 240 Hz timing to drive both eyes and both projectors. So each projector is driven at 120 Hz and the shutters is timed to close the left shutter when the image on the right projector is projected and seen on the right eye (when the shutter is open on the right eye). The left eye's shutter lense opens when the right eye shutter closes. Each eye is open 120 times per second and closed 120 times per second. Right eye shutter is open, the left eye shutter is closed. Then the left eye shutter opens and the right eye shutter closes. It does it so fast that at 120 Hz rate per eye, you can't possibly see the flicker. The photo resists and optic timing rate biologically to reach your brain, the brain process the image and so forth... there is no way at 120 Hz you can see flicker. You can't see flicker on a 120 Hz TV or even on a 120 Hz LCD.

    You won't see it. Human brain doesn't process that kind of information that fast.

    By using this much faster shutter rate... you are effectively getting 120 fps. This is at a rate that is tough for PC video cards to drive at at full performance.

    Of course exact timing parameters are irrelevant here for all intents and purposes, the video is operating on intervals rates of effective referesh rate per eye at 120 Hz (cycles per second). This is extremely fast.

    The reason they don't do that rate for TVs and "3D" LED backlight LCD TVs has a lot to due with the LCD technology, long photo resists and god knows what reason. Jeri isn't playing by those low duty cycle non-sense. She is maximizing the tech to the fullest as of the available tech at the price mark.

    Why 'cripple' the shutter technology?

    Passive Polarized glasses is an older technique and although cheaper... you don't get the resolution. You lose resolution. Learn something about how the polarized glasses are made. You'll realize what I mean. Even big brand companies have explanatory video on that. When I find the videos, I'll point you to it. Learn it. Understand it. There is certain properties and attributes that Jeri wanted. To get 720p with passive polarized lenses, you would projectors projecting 1440p just so you lose half your pixels. Not exactly what Jeri is seeking with the technology. The price gets a bit pricey and there is also physical size.

    The current OR would be a neck killer when it comes to long periods of game play.

    The proposed new version is just nice graphics and mock up. Does it work as they claim? If it did, why is it not available as dev kits? Maybe, they'll get there but then at what price???? Hmm... Ok. It would be purely dedicated as a VR device and not very adaptable to many different game play and trying to use camera to give you the AR feel by displaying the outside world as a back ground does not exist that will have <1ms latency per frame rate that is small as your typical spy cam and have the same pixel resolution (hence same HSync) and genlock the computer display and the web cam and magically do all that without going over 1ms latency in total. Guess what, the technology to do it would drive the price upward to $1000 or more. Even professional grade genlocks for AMIGA back in the 1990s was expensive. Genlock for PC video is still not cheap. You have to do it at 120 ~ 240 Hz not 60 Hz. You need extremely low latency of probably 0.1ms on the genlock component. There is just a lot of things it would take to get the price down. You wouldn't want a camera that gives you 1080p, or Ultra 4K and it is grainy, full of artifacts and you need it to be constant and need the genlocking process immune from stuttering because the PC cpu decides to put priority to something else or hard drive activity or some crazy stuff that the OS wants to through the resource to. This means custom hardware to relieve the main computer from having to process.

    I doubt the consumer version of OR will have everything and will have to make trade offs. Don't get me wrong, OR is a cool project but it is largely as it currently is... a VR-only product suited for only VR.

    castAR, I can probably easily add inexpective side shields to block peripheral vision.

    Something like these but just paint them black if you need to:

    https://www.google.com/search?q=side+shields+for+glasses&newwindow=1&rlz=1C1DVCP_enUS552US552&espv=210&es_sm=122&source=lnms&tbm=isch&sa=X&ei=q9lxUvaPJMS0iQL8t4Fg&ved=0CAkQ_AUoAQ&biw=1143&bih=787#es_sm=122&espv=210&newwindow=1&q=clip+on+side+shields+for+glasses&tbm=isch

    Google for clip on side shields.


  • Soulfly

    Posts: 6

    Posted:
    Posted:
    Do not think of it as Oculus vs CastAR. I am sure each of them will excel in some areas and lack in others.

    What is important is that they are both of them are trying to bring Augmented and Virtual Reality to us through crowd-sourcing where the big corporations have not. We should fully support and thank both of them for making this a reality for us, and look forward to them continuing to innovate in this field. I sincerely hope there will be a competitive rivalry between them resulting in newer/better hardware being released every few months.
  • HydraShock

    Posts: 880

    Posted:
    Posted:



    castAR, I can probably easily add inexpective side shields to block peripheral vision.

    Something like these but just paint them black if you need to:

    https://www.google.com/search?q=side+shields+for+glasses&newwindow=1&rlz=1C1DVCP_enUS552US552&espv=210&es_sm=122&source=lnms&tbm=isch&sa=X&ei=q9lxUvaPJMS0iQL8t4Fg&ved=0CAkQ_AUoAQ&biw=1143&bih=787#es_sm=122&espv=210&newwindow=1&q=clip+on+side+shields+for+glasses&tbm=isch

    Google for clip on side shields.


    I was thinking the exact same thing. If people are really worried about light around the glasses, those would def help reduce that (assuming its even a problem worth fixing). You could also easily fabricate something out of a set of goggles im sure.

  • HydraShock

    Posts: 880

    Posted:
    Posted:
    0ni | 0ni said:

    @Dimididales & @HydraShock
    I'm well aware that moving a screen closer to your face doesn't make magic pixels appear. This isn't my first rodeo as it were. :p
    I've in the past had a triple projector setup at just over 180° surround and I can tell you that FoV does more for immersion then pixel count. Obviously if the resolution is too low it's a bummer. I would never go back to gaming at 1024x768 for instance. Though filling your view is infinitely more important for immersion then pixel density, at least for me. Hence why I said I'll probably go with whatever has the widest FoV, because that's simply what matters to me. I always go with what makes the most sense for me, why should I buy something because lots of people think it's great when a different product has the features that matter to me? Barring of course some huge failure in another area. If the oculus came out with a 200° FoV but 12ms latency obviously I'd skip it, but as it's been apparently officially proclaimed to be 2 that really narrows down the margin for error,

    Price is also a factor, if anyone charges a crazy/unjust premium I vote that shit down with my wallet. So at the end of the day, if castAR can't beat Oculus for FoV I'm skipping it unless some hard facts come to light that prove it significantly better in other categories.

    I don't know why but y'all make me want to avoid the product for some reason. This thread is a real turn off for me, which is weird because I was really interested in the product before. I guess I still am, I dunno. Thread title still sounds like "castAR: I am right everyone else is wrong". Maybe I'm just getting the wrong vibe.

    Oni,

    Im not trying to be a prick here. Perhaps I am being a bit defensive, mainly because I have been attacked, primarily out of ignorance. All I am trying to do is get this out there so others can learn about it, and get as excited bout it as I am :). I see it being a huge leap forward, just as the rift is. It just really got under my skin when just bringing it up brought forth a barage of hate against it...Ever since, I have been trying to educate people, thats it. If I have come off defensive, I apologize :)

  • WavestarInteractive

    Civilian

    Posts: 24

    Posted:

    Civilian

    Posted:


    Guess what, the CastAR doesn’t need a mat either. In fact this device can be used to make your NORMAL monitor into 3d as well, on top of the ability to go full VR (just like the rift), without the use of any projection mat/screen.


    -The tracking on the CastAR is curently far superior to the Rift. The Rift uses gyros, and induce a significant amount of lag (a serious issue the rift folks have promised to remedy). The Cast AR does not use gyros, it uses triangulation via IR LEDs. This triangulation is basically zero lag, and therefore causes much less motion sickness. The CastAR also does 6DOF tracking (which includes side to side/forward and back translation), which the Rift does not CURRENTLY support. There are rumors that the OR will support 6DOF tracking, but we have yet to see it in a working prototype. Either way, the castAR will feature it in their first release

    Full disclosure, I am the creator of the Rift. You are a bit off on a few things, I just wanted to clarify where I can.

    The gyros in the Rift do not introduce significant lag, they are actually one of the fastest things in the pipeline. We sample internally at 8000hz and report the smoothed samples over USB at 1000hz, so there is only 1ms between updates. The total latency of the tracker from motion to PC is about 2ms. It is definitely not accurate to say that CastAR has almost zero lag when compared to the Rift, the optical system they are using is innately 8 times slower than the Rift tracker, not including the time it will take to do image processing (which is minimal, since they are doing that in hardware). Full 6DOF tracking is definitely important, though, it does a lot to reduce simulator sickness! You are also partially wrong about being able to use CastAR without setting up a mat, you still need to set up markers for the optical system in whatever direction you plan on looking. Looking around 360x360 needs visible markers on all the walls, ceiling, and floor. I think they mentioned that they might build a gyro into the final product and use it in tandem with the optical system, which is a great way to do things.

    I am excited to see the final CastAR, I backed the Kickstarter for $10k. Can't wait to see what they end up with!
    I'm glad you're cool about it all. I think we both be interesting in seeing Star Citizen supporting both Oculus Rift and castAR.

    I'm glad to here the ms sampling is down to good level. I think some further refinements but to get the AR experience with OR might introduce more problems that can be problematic... at least now. Palmer... The mat referred to is the RR sheet. They themselves are NOT the IR LEDs themselves. The IR LED is on a separate piece of material with the IR LEDs modulated that you would place near or around the sheet when using the projection mode AR. However, when using the AR/VR clip on units, the RR sheets themselves are just 1m x 1m OR 1m x 2m sheets cut from 1m wide x _____m long rolls of retroreflective fabric. I probably have an idea who and make and model(s) of the fabric but that is largely irrelevant.

    Disregard any error I may have stated about the OR in terms of anything timing. I haven't read through the entire thread but it is great to hear your input as well. I don't intend to diss Oculus Rift in any intended bad light or meaning. It is a good system and I believe it is a good VR system. castAR is a good AR/VR system and there is simple solutions to resolve immersion through clip on side shields to increase immersion feel by blocking the sides along the peripheral.

    The IR LEDs are infrared so they aren't necessarily going to emit visible light and is more like what you'll see on TV remotes.

    The modulation would be different but head tracking only going to really need maybe 8 modulated IR LEDs which you can mount on little objects in the room so it'll work out. Only need 4 to 5 IR LEDs within something like a 110 degree FOV.

    Just wanted to clarify a few points for your better understanding since you are putting a nice chunk of change out for it so, hey, why not. In this world, I think there is room for both castAR and Oculus Rift.

    Just like there is room for Windows and Mac.

    It's all good competition (not enemies but fair competitors striving to make improvements in AR & VR). I know the current OR available broadly... is neck killer over long time of game play... though better than the ones in the 90s that I've seen.

    However, I think you plan on improving things a bit.

    No human head can look in 360 degrees around all X,Y & Z axis and would have to do alot to do that. However, not actually something most average gamers are going to do... considering there are plenty that just aren't super athletic. However, it is possible to have plenty of IR LEDs around to track. Keep in mind that the modulation and things happening to operate and track the IR LED is operating at a lot faster than 120 Hz but there is other things that goes on. However, the head tracking at 120 Hz is more than enough but it isn't necessarily the rate of tracking that is important but the accuracy of each report. Since the image refresh rate is at 120 Hz per eye, it is faster than the eye and ultimately the brain is going to perceive change since what you see is not what the eyes see but what the brain does to process what the eyes collect. Since the projected rendering on each eye is being down at 120 Hz, it would be largely computational waste to go faster than the same timing base for movement refresh. The movement can be collected in lock step. How much distance can a human move in 1/120th of a second? Not very much. So precise distance change in that is like what...?

    Lets take a human at full out run at 20 MPH... (about as fast as an olympic runner can run)

    That computes to 105,600 feet per hour. Lets divide that down to 1/120th of a second. That's 0.244444 ft. Lets convert that to inches... 2.933333 inches and then convert it to millimeters... 74.506666666 mm. Say... 74.5mm... might push to 75mm per second. With a collection of IR LED markers that can be modulated and read rapid... is not much different than surveying techniques. You can gauge the distance through the modulation pulse and timing. Since IR works on a frequency in the 300 GHz to several TeraHertz and probably some few KHz wide channels for the modulation of pulses and processed by a processor component that I would say is alot faster than 120 Hz and can work in between the sampling rates and then compute back a coordinates results all within the window between each read. The trigonometry being computed is surveyor accuracy and at this short distance... computed accuracy would be extremely accurate. In the conditions, issues of environmental accuracy would be minimal. Atmospheric conditions and PPM corrections will be meaningless in these short range. Accuracy of a survey would take known distances between several readings and trigonometry is used to compute accuracy. Exactly how that works fully, has not been explained.

    I have some ideas how she is achieving the IR tracking or at least inspirational source techniques that inspired the particular technique but 3-d trigonometry comes to mind and something about the signal from each IR LED besides being modulated.




  • WavestarInteractive

    Civilian

    Posts: 24

    Posted:
    Edited:

    Civilian

    Posted:
    Edited:
    0ni | 0ni said:

    @Dimididales & @HydraShock
    I'm well aware that moving a screen closer to your face doesn't make magic pixels appear. This isn't my first rodeo as it were. :p
    I've in the past had a triple projector setup at just over 180° surround and I can tell you that FoV does more for immersion then pixel count. Obviously if the resolution is too low it's a bummer. I would never go back to gaming at 1024x768 for instance. Though filling your view is infinitely more important for immersion then pixel density, at least for me. Hence why I said I'll probably go with whatever has the widest FoV, because that's simply what matters to me. I always go with what makes the most sense for me, why should I buy something because lots of people think it's great when a different product has the features that matter to me? Barring of course some huge failure in another area. If the oculus came out with a 200° FoV but 12ms latency obviously I'd skip it, but as it's been apparently officially proclaimed to be 2 that really narrows down the margin for error,

    Price is also a factor, if anyone charges a crazy/unjust premium I vote that shit down with my wallet. So at the end of the day, if castAR can't beat Oculus for FoV I'm skipping it unless some hard facts come to light that prove it significantly better in other categories.

    I don't know why but y'all make me want to avoid the product for some reason. This thread is a real turn off for me, which is weird because I was really interested in the product before. I guess I still am, I dunno. Thread title still sounds like "castAR: I am right everyone else is wrong". Maybe I'm just getting the wrong vibe.

    At 120 FPS tracking and refresh rate, we are talking 8.333333 ms between each update. IR is a spectrum at the speed of light. Processing the IR tracking information is all entirely processed within the 8.3333333 ms time window and probably pipelined. If you designed a 6502 core then you know about pipelining. Do yourself a little thinking. A guy has a rifle with a bullet that will travel 5000 feet per second. He is 45-ft. from you. He pulls the trigger... do you think you'll detect the moment the bullet exits the barrel or even see the bullet at all?

    Probably not at all. The primary reason you won't detect is the conditions you'll be in during game play. You will almost never be in perfect darkness but there is an afterimage residual because the images are considerably bright and the darkness would be eaten up. The after-image will persist in the eye for a period of time. You'll see this with your monitors and TV. Especially the 100 Hz.

    The detection is one thing. To identify and process and determine if the change is enemy, friend, or anything of particular importance including identifying patterns... all that takes much more time. This is where things get more difficult. In 1/120th of a second, you aren't going to detect and identify that you have a monster approaching and attacking you. In that moment, you might have just the fraction of a notice that something is approaching. Then it comes down to skills. 3d sound is probably going to cue your attention quicker in a particular direction before you would even notice on peripheral. You sure isn't going to bead on and target the monster until you see it in front of your eyes. At least in and game play you experienced.

    However, the latency in the tracker is 2ms as Mr. Luckey said. The latency is irrelevant for the most part. It is the quality of the results. What is the tracking resolution? I can spit out garbage numbers 1 quadrillions times a second on a fast super computer. This is meaningless to the end user. castAR is calculating absolute position. Not a mixture of actual and hypothetical predictive guessing which can be problematic. Why waste computational time and resource to predict something that might in reality not quite occur according to the computer theory? Why waste computational resource on above the screen refresh rate. After all, the player isn't moving very far in 1/120th of a second not to mentioned, each sampling and calculations is absolute position calculations at 1/120th of a second time period. It would detect your position at an accuracy of around 1/10th the diameter of a typical 0.7mm mechanical pencil lead... That is 0.07mm accuracy. That is computed mathematically. It isn't speed that necessarily matters. Although to a point... it won't matter so much when you aren't going to visually detect and respond to changes in a typical video game at that speed.

    Heck, most people can't even outpace a 24fps movie camera. Only person known to do that was Bruce Lee and maybe Jet Li. Perhaps Ip Man if he was ever filmed in his younger years in fighting. I doubt the average not too physically athletic gamer is going to be that quick and nimble.

    Good chances.. nope. Maybe only on twitch thumb action but that's about it.

    To add a few points, typical computer graphics may use a number of techniques such as frame tweening or motion blurring techniques and antialiased rendering or any number of techniques that might slightly soften the edge and so forth and get the fluid feel. You name it. We are only capable of discerning 10-12 individual frames per seconds as individual frames. We might detect flicker or other some 'change' at faster rate but for game play... you don't optically perceive patterns and characters and things like that on a screen until 1/10th to 1/12th of a second at best. For ascertaining and making a decision and what not, we are talking maybe a response turn-around time of maybe 1/8th of a second to about a 1/4 second. For most, anything reaction of game pads above 1/4 seconds is just players slapping every darn button they can like they might do in Street Fighter. Maybe 2/3 of the buttons pressed were not consciously chosen but just wildly pressing buttons as fast as you physically can. While maybe a 1/3 of them or less is actually thoughtfully chosen sequences. Most people just act on twitch and not on thoughtful decision. This twitch-play is largely promoted by Unreal Tournament where you just shoot anything that moves on twitch reaction to movement.


  • JeriE

    Civilian

    Posts: 1

    Posted:

    Civilian

    Posted:
    Hi All,
    I'm Jeri Ellsworth one of two co-founder of Technical Illusions. Rick and I are both industry veterans with over 30 years of combined experience shipping quality products. We proudly stand behind our product and have shown castAR publicly to thousands of people since May. We'd love for you to give it a test drive and see for yourself at GDC Next November 5 - 7th.
  • Dimididales

    Posts: 357

    Posted:
    Posted:
    A DnD castAR try-it-yourself setup would be *the* attraction at spielwiesn - a really large board game exhibition in Munich this weekend where thousands of gamers can try out old and new board games themselves.
    Perhaps you can get in touch with them, lending them some demo devices?
    NoamLoop2.png
  • Woopate

    Posts: 2573

    Posted:
    Posted:

    Hi All,
    I'm Jeri Ellsworth one of two co-founder of Technical Illusions. Rick and I are both industry veterans with over 30 years of combined experience shipping quality products. We proudly stand behind our product and have shown castAR publicly to thousands of people since May. We'd love for you to give it a test drive and see for yourself at GDC Next November 5 - 7th.

    Crank up the awesome in this thread another notch, why don't ya?

    I think the biggest hurdle IS getting people to test these systems themselves. Most people (myself included), have no frame of reference to envision the effects of various limitations and strengths for various AR/VR systems. The strengths/weaknesses calculation CANNOT accurately take place until people have put one at least 1 (preferably 2 or more) competing systems. Crap, the first and only VR system I have ever tried was a silly parachute game in an arcade in the 90s. It was boring, the visor was heavy, and they blew a fan in my face to get me to think I was falling.
    Ztj7fnR.png
  • Dimididales

    Posts: 357

    Posted:
    Edited:
    Posted:
    Edited:
    0ni | 0ni said:

    I don't know why but y'all make me want to avoid the product for some reason.

    @Oni sorry, it was never my intention to glorify any product by dissing others - hope you didn't get this impression. People are searching and relying on opinions - and so many interesting technologies didn't make it because fans of existing things made negative propaganda against it to keep it from being successful - and it works unfortunately - even - or especially if "counter fans" try to counter it with the same approach.
    As I can't stand this kind of FUD wars and thought castAR doesn't deserve this, I tried to contribute what I am sure enough I understood right to shift the propaganda wars to a more informed, fact based discussion.
    NoamLoop2.png
Sign In or Register to comment.