Updated Vireio Drivers, New Development Thread

The place for all discussion of the Oculus Rift compatible open source 3D drivers.
User avatar
Neil
3D Angel Eyes (Moderator)
Posts: 6882
Joined: Wed Dec 31, 1969 6:00 pm
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Neil »

Front page news today!

http://www.mtbs3d.com/index.php?option= ... &Itemid=73

Regards,
Neil
virror
Sharp Eyed Eagle!
Posts: 427
Joined: Fri Jan 18, 2013 7:13 am
Location: Gothenburg, Sweden

Re: Updated Vireio Drivers, New Development Thread

Post by virror »

Congratulations!
And a huge thanx to all involved for getting it this far : D
User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11406
Joined: Sat Apr 12, 2008 8:18 pm

Re: Updated Vireio Drivers, New Development Thread

Post by cybereality »

Awesome!
Matthew
One Eyed Hopeful
Posts: 41
Joined: Tue Jun 25, 2013 11:45 pm

Re: Updated Vireio Drivers, New Development Thread

Post by Matthew »

I've now looked at the source code. It uses two additional back buffers with separate depth buffers and separate stencil buffers, like I was saying.

It does not, however, do pixel averaging with the interleaved stereoscopic modes -- it just does a "mask" operation and uses the respective pixels of the full resolution left and right eye images.

Although this is, unfortunately, how most stereoscopic viewers display interleaved content, it is a terrible way of doing it. Not only does it create aliasing artifacts, the artifacts are different in the left and right eyes. (And no, the brain does not "fill in" the missing data to obtain a full resolution image.)

This is the main reason people dislike interleaved stereoscopic displays. It's not because the display technology itself is bad -- it's because software or hardware that displays interleaved content doesn't do it properly.

The proper way of displaying interleaved content is to downscale the image to the necessary lower resolution using pixel averaging before displaying it, with no variations within each pixel quad. If it's done this way, the quality will be the same as other types of stereoscopic displays would be at the lower resolution.

This would be very easy to do -- you would just have to modify the shaders. If it's not done in the official version, I'll do it myself before I play games with it. :)


baggyg wrote:Regarding autofocus this is an option in Tridef. Some users (admittedly not most) do prefer this. It is far better for developers to give the users the choice of whether we like something or not.
If a user prefers that, it's because they don't know how to do 3D right. This feature is fundamentally unnatural and unrealistic.

Mathematically, "screen shift" represents half the physical distance between the viewer's eyes in the real world (as opposed to half the virtual distance between the left and right eyes in the virtual world). This is why this shift (unlike the camera shift) is done in screen coordinate space, not world coordinate space.

This parameter never changes. Not between different games. Not between games and videos. Not even between different stereoscopic displays, in terms of the actual physical distance.

This parameter should always be set at around 3 cm (both eyes are shifted, and twice that is 6 cm, just slightly less than the distance between the viewer's eyes). It should never be set at anything else. Of course, the necessary number of pixels will vary depending on the display.

This is something about 3D that very few people realize. Extremely few people know how to truely do 3D right.

When I set the "screen shift", I actually measure it with a ruler, using something at virtually infinite distance (for example a cloud, or the moon) as reference. This may seem picky, but it really does improve the 3D.

Remember, the purpose of 3D is to create a virtual reality experience. The "autofocus" feature is fundamentally unnatural and unrealistic, and creates a weird, discomforting effect.

Including an option for such a feature is like including an option for affline, not perspective correct, texture mapping. If you include features like that, naive people will use them not knowing that they harm the experience.

Tridef Ignition even enables the "autofocus" feature by default.

On a side note, Tridef Ignition is by far the worst stereo driver or wrapper I've ever seen, in terms of design. The eDimensional wrappers trump it in that regard.



I'm looking forward to trying out the Vireio wrapper. I can't right now, because my computer (which had an nVidia GeForce 8400 GS video card I purchased last year but was 9 years old) went dead a few weeks ago, and I'll have to purchase a new one soon.
User avatar
Neil
3D Angel Eyes (Moderator)
Posts: 6882
Joined: Wed Dec 31, 1969 6:00 pm
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Neil »

Hi Mathew,

First, I'm in agreement about "auto-convergence" features. I never liked them, and in all the MTBS game reviews, I recommend turning them off because they are usually completely unnecessary and undermine performance. We definitely do NOT use auto-convergence features in Vireio Perception or have plans to.

However, I really think DDD deserves their due. They have rarely (never?) had the game developer inside track, and through their cunning and programming prowess, they regularly put out better 3D quality than what the game developers market as 3D Ready or even natively develop sometimes! Yes, even with auto-convergence turned off! :lol:

Github isn't telling you the whole story. The source code is there, but we have also been using a separate organizational site to get things moving. One of the tasks in the works is to "Reconstruct StereoInterleaveClass" (is this the same thing?). I don't know how big a priority this is, but one of the developers has had this on the list for some time.

Thank you for your continued enthusiasm! If you are interested in helping contribute to the effort, feel free to send me a PM!

Regards,
Neil
GameDev
One Eyed Hopeful
Posts: 24
Joined: Fri Apr 26, 2013 5:18 am

Re: Updated Vireio Drivers, New Development Thread

Post by GameDev »

Awesome news.

Any idea when there will be a release? Looking forward to trying it out. Is there a games list that it would now support? Hoping that it supports Outlast and possibly Amnesia!

Thanks
User avatar
Neil
3D Angel Eyes (Moderator)
Posts: 6882
Joined: Wed Dec 31, 1969 6:00 pm
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Neil »

We are aiming to have a very early build release within a week. However, it's more of a "knock,knock! We are still alive" release. Everyone has been waiting too long for an update, so it's time already.

Some big ideas will be coming out of the software as a proof of concept, and diverse game support will follow as we stabilize the code. I'm hoping after you see some things, you'll say "#*$&! That's really cool!". Even then though, there are important innovations around the bend that won't be included in this initial release. We are still inventing as we go!

Expect only a handful of games to start with a rapid acceleration to follow. I haven't been this excited about an MTBS project in a very long time, and I really think the software will be something for the community to be proud of. More than that, it's all free.

Regards,
Neil
User avatar
RescueGamer
Cross Eyed!
Posts: 108
Joined: Thu May 16, 2013 1:02 am

Re: Updated Vireio Drivers, New Development Thread

Post by RescueGamer »

That's awesome news!
Neil, I want to request something if it's possible. For the new version can you work on a good .fx file for DIY Rifts? we need something more than a side by side, I don't know if it's possible, but with the oculus rift option the image is a bit small for the diy rifts. Some DIY Rift owners tried to create a god fx file in this thread (http://www.mtbs3d.com/phpbb/viewtopic.php?f=141&t=17741), the result was good, but not the best solution while we can't see some parts of the game (I think that a side by side keeping the aspect ratio and adding wrapping would be good for games were you can change the fov). I hope you can dedicate some time to this. Really thanks for your work, with vireio I can really use mi DIY Rift :D
PD: Sorry for my english level :S
Last edited by RescueGamer on Wed Oct 02, 2013 4:23 pm, edited 1 time in total.
User avatar
Neil
3D Angel Eyes (Moderator)
Posts: 6882
Joined: Wed Dec 31, 1969 6:00 pm
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Neil »

You're talking about hardware profiles, right?

While I have no objections to this, our priority is clearing up the crash errors first. It would be a big help if there is a developer out there that would like to focus on specialized HMD outputs like this - someone who would literally keep tabs on the specialized needs of the different solutions and create display profiles. Or better yet, as the industry grows, if HMD makers want to submit their own display profiles, that might work too.

It's understandable that people don't have a lot of time in their day. However, if a chosen few can choose something they are really good at, it's easy to divide things into manageable bites.

Regards,
Neil
User avatar
RescueGamer
Cross Eyed!
Posts: 108
Joined: Thu May 16, 2013 1:02 am

Re: Updated Vireio Drivers, New Development Thread

Post by RescueGamer »

I know that it's nor the priority (you have better things to focus on), if you have time some day i will help in anything i can. I used the most common parts (7 inch 1280x720 lcd and 5cm 5x lenses), so is near to the oculus rift, but the zoom(or wathever it is called) of the lenses is lower, i think, and in my case, each center of the lens is aligned the centers of each half of the screen, don't know if this is siggnificant.
I think tha the only I can help is testing anything you need an recovering information. I have knowledges of java, html, xml, sql, and others languages, but that's all, don't know how to create hardware profile very well :S
GameDev
One Eyed Hopeful
Posts: 24
Joined: Fri Apr 26, 2013 5:18 am

Re: Updated Vireio Drivers, New Development Thread

Post by GameDev »

Neil wrote:We are aiming to have a very early build release within a week. However, it's more of a "knock,knock! We are still alive" release. Everyone has been waiting too long for an update, so it's time already.

Some big ideas will be coming out of the software as a proof of concept, and diverse game support will follow as we stabilize the code. I'm hoping after you see some things, you'll say "#*$&! That's really cool!". Even then though, there are important innovations around the bend that won't be included in this initial release. We are still inventing as we go!

Expect only a handful of games to start with a rapid acceleration to follow. I haven't been this excited about an MTBS project in a very long time, and I really think the software will be something for the community to be proud of. More than that, it's all free.

Regards,
Neil
Understood Neil, thanks a lot this is great to hear! Can not wait to try it :)
User avatar
EAPidgeon
One Eyed Hopeful
Posts: 45
Joined: Tue Mar 19, 2013 5:20 pm
Location: Sonoran Desert

Re: Updated Vireio Drivers, New Development Thread

Post by EAPidgeon »

Neil wrote:We are aiming to have a very early build release within a week. However, it's more of a "knock,knock! We are still alive" release. Everyone has been waiting too long for an update, so it's time already.

Some big ideas will be coming out of the software as a proof of concept, and diverse game support will follow as we stabilize the code. I'm hoping after you see some things, you'll say "#*$&! That's really cool!". Even then though, there are important innovations around the bend that won't be included in this initial release. We are still inventing as we go!

Expect only a handful of games to start with a rapid acceleration to follow. I haven't been this excited about an MTBS project in a very long time, and I really think the software will be something for the community to be proud of. More than that, it's all free.

Regards,
Neil
Really glad to hear this, not to knock on VorpX or TriDef, but backwards compatibility is something I feel really should be free and will be a huge benefit for the average consumer adopting the Rift. I can imagine it being more then a bit off-putting to have to lay down 40$ just to get it. I won't lie however, Neil, the previous drivers at least when I "attempted" to experiment with them were something of a mess. Granted TF2 felt like a bit of a claustrophobic mess for me using Valve's implementation (when I used to use C-Lenses).
User avatar
Neil
3D Angel Eyes (Moderator)
Posts: 6882
Joined: Wed Dec 31, 1969 6:00 pm
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Neil »

There is no crime in making a living. We are keeping the drivers free because as a media website, we want to avoid conflicts of interest with industry. Also, once you start accepting payments for service, you are directly responsible for it. We've got enough on our plate thank you very much!

Free or otherwise, I am very confident Vireio Perception will be a remarkable improvement over what you saw previously. We've got a sharp team, everyone is on the same page of where we want to go, and it's gratifying to know we are solving real problems in the process.

Thanks for your kind support.

Regards,
Neil
Matthew
One Eyed Hopeful
Posts: 41
Joined: Tue Jun 25, 2013 11:45 pm

Re: Updated Vireio Drivers, New Development Thread

Post by Matthew »

I now have a new computer -- an i5-4440 a GeForce GT 635 graphics card with 64-bit Windows 8. I will try out the Vireio wrapper soon. :)

The main criticism I have of Tridef Ignition is the terrible adjustment features.

And if they put out better 3D quality than what game developers natively develop, that's because the game developers don't know how to do 3D right.

Here's a little known fact: The main reason Oculus Rift is so much more immersive than other types of 3D displays typically are is not because of the technology, but because of the "screen separation". With Oculus Rift, by nature, the screen separation is about the distance between the viewer's eyes, like I was saying. But with other types of 3D displays, you have to set it manually, and most people don't set it that way. If you set the screen separation that way with a 3D monitor or 3D television, like I said in my earlier post, the effect will be almost as good, except of course that you can't turn your head around. There may be a little ghosting, but the effect is worth it. It feels like the screen is a "window" into the actual scene. :)


Neil wrote:One of the tasks in the works is to "Reconstruct StereoInterleaveClass" (is this the same thing?).
No. The stereoscopic format output is done entirely in the shaders. It has a different shader for each format.
User avatar
Fredz
Petrif-Eyed
Posts: 2255
Joined: Sat Jan 09, 2010 2:06 pm
Location: Perpignan, France
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Fredz »

Matthew wrote:Here's a little known fact: The main reason Oculus Rift is so much more immersive than other types of 3D displays typically are is not because of the technology, but because of the "screen separation".
It's one of the aspects, but I wouldn't say that's the most important one for immersion. Wide FOV and head tracking are a lot more important to me. And to some VR researchers (from LEEP for example), stereoscopy is not even required for immersion. I read a paper about people in a VR lab who misconfigured there software so the rendering was pseudo-stereoscopic and they were surprised that nobody noticed. It depends a lot on the content also, if most of the content in the scene is past a certain distance (~65m) stereopsis is no longer the main depth cue.
Matthew wrote:With Oculus Rift, by nature, the screen separation is about the distance between the viewer's eyes, like I was saying. But with other types of 3D displays, you have to set it manually, and most people don't set it that way. If you set the screen separation that way with a 3D monitor or 3D television, like I said in my earlier post, the effect will be almost as good, except of course that you can't turn your head around. There may be a little ghosting, but the effect is worth it. It feels like the screen is a "window" into the actual scene. :)
That's true, but the reason most people won't configure separation this way is not necessarily because they don't know about it but because it's also straining on the eyes because of the vergence-accomodation conflict. The Rift doesn't suffer from this problem since the lenses make the eyes focus at infinity and not at < 1m like on a monitor.
Matthew
One Eyed Hopeful
Posts: 41
Joined: Tue Jun 25, 2013 11:45 pm

Re: Updated Vireio Drivers, New Development Thread

Post by Matthew »

Fredz wrote:I read a paper about people in a VR lab who misconfigured there software so the rendering was pseudo-stereoscopic and they were surprised that nobody noticed.
If they didn't notice, it was because they were very naïve about 3D and/or the 3D had very little depth.

I hate it when 3D is that way. I hate it so much that whenever I set up 3D on my 3D monitor, I check first to make sure that it's not reversed, before even looking at it in 3D. I look through only the left eye until I've checked that it's correct.


Fredz wrote:That's true, but the reason most people won't configure separation this way is not necessarily because they don't know about it but because it's also straining on the eyes because of the vergence-accomodation conflict.
If you're not used to it, it can be hard on your eyes, but over time, your eyes easily learn to adapt to it.

Setting the screen separation a different way actually has far more potential to cause eyestrain, because it's unnatural, and can result in excessive depth.

The only time that 3D causes me eyestrain is when the eyes are reversed or the screen separation is improper.


Fredz wrote:It depends a lot on the content also, if most of the content in the scene is past a certain distance (~65m) stereopsis is no longer the main depth cue.
As I explained in my earlier post, unlike camera separation, screen separation is depended on the physical distance between the viewer's eyes, not on the content being displayed. Even if everything is at infinite distance, setting the screen separation that way will make it a lot more immersive -- it will make it actually appear at infinite distance.

As far as the importance of screen separation... It makes the difference between a girl appearing tiny, like a Barbie doll, and appearing of realistic size, like you could cuddle with her. It makes the difference between the 3D being a cheap gimmick that pops things up in your face, and a virtual reality experience.
User avatar
Guig2000
Binocular Vision CONFIRMED!
Posts: 257
Joined: Wed Nov 25, 2009 9:47 am
Location: Bordeaux, France

Re: Updated Vireio Drivers, New Development Thread

Post by Guig2000 »

Matthew wrote:
baggyg wrote:Regarding autofocus this is an option in Tridef. Some users (admittedly not most) do prefer this. It is far better for developers to give the users the choice of whether we like something or not.
If a user prefers that, it's because they don't know how to do 3D right. This feature is fundamentally unnatural and unrealistic.

Mathematically, "screen shift" represents half the physical distance between the viewer's eyes in the real world (as opposed to half the virtual distance between the left and right eyes in the virtual world). This is why this shift (unlike the camera shift) is done in screen coordinate space, not world coordinate space.

This parameter never changes. Not between different games. Not between games and videos. Not even between different stereoscopic displays, in terms of the actual physical distance.

This parameter should always be set at around 3 cm (both eyes are shifted, and twice that is 6 cm, just slightly less than the distance between the viewer's eyes). It should never be set at anything else. Of course, the necessary number of pixels will vary depending on the display.

This is something about 3D that very few people realize. Extremely few people know how to truely do 3D right.

When I set the "screen shift", I actually measure it with a ruler, using something at virtually infinite distance (for example a cloud, or the moon) as reference. This may seem picky, but it really does improve the 3D.

Remember, the purpose of 3D is to create a virtual reality experience. The "autofocus" feature is fundamentally unnatural and unrealistic, and creates a weird, discomforting effect.

Including an option for such a feature is like including an option for affline, not perspective correct, texture mapping. If you include features like that, naive people will use them not knowing that they harm the experience.

Tridef Ignition even enables the "autofocus" feature by default.
  1. In theory you are right, but practically, the use of an auto-focus feature can be interesting on a few cases:
  2. When you want something non-realistic, or for you use auto-convergence for an artistic reason. In example, there is the RTS R.U.S.E. (even if don't like the way they did it because they added a filtering to their auto-convergence, leading to a latency between the zoom and the convergence) where Eugen (the devs) implemented a real time modification of the convergence proportional to the zoom on the map. Their reason is simple: when you zoom out, you are like a character on the headquarter, and you see the map of the operations, with little tokens representing units, so you need to have a model effect. While when you zoom in the map, you are like in the real battle, so you need a convergence setting that place objects at a much far distance to get more real sized objects and make it more realistic.
  3. In order to compensate some S3D issues in RTS-like games (civilization, supreme commabder etc): the game is often full of 2D icons all over the screen, so if you can't 3dize them, the gap between 2D icons and the 3D world will make the game unplayable in S3D except if you use a auto-convergence feature can help to maintain objects at a depth which is near the 2D icons.
  4. When the game operate a zoom which is obtained by reducing the fov, if you do not modify convergence and separation, everything will become flat when zoom in. One of the obvious example is sports events filmed in 3D like football or soccer: when the 3D camera zoom-in, the 3D effect is totally killed.
But yes it's true that the 2 firsts cases will be very rare for a HMD user. Usually nobody play RTS using a HMD.
Image
User avatar
Fredz
Petrif-Eyed
Posts: 2255
Joined: Sat Jan 09, 2010 2:06 pm
Location: Perpignan, France
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Fredz »

Matthew wrote:
Fredz wrote:I read a paper about people in a VR lab who misconfigured there software so the rendering was pseudo-stereoscopic and they were surprised that nobody noticed.
If they didn't notice, it was because they were very naïve about 3D and/or the 3D had very little depth.
I doubt it since the test was conducted in a VR lab, but unfortunately I can't find where I read this. Anyway the conclusion of the researchers was that wide FOV and head tracking was far more immersive than stereoscopy.

I found the quote by Eric Howlett (inventor of the LEEP optics) though :
"Only recently have I realized in retrospect that this was a clear indication that a substantial FOV, at least in a display that permits turning the head, was a much more powerful agent in creating a sense of immersion than stereopsis ever could be."
Matthew wrote:If you're not used to it, it can be hard on your eyes, but over time, your eyes easily learn to adapt to it.
Maybe it was the case for you, but it may not be the case for everyone. To me it even sounds quite dangerous to tell people to force their visual system in an uncomfortable and headache inducing configuration without knowing the implications.

There is a reason why the default values for depth in stereo drivers are quite low and that moviemakers also minimize this conflict, and there is a whole body of research about this subject. At this time it's also not known if unnaturally decoupling accommodation and vergence has any irreversible long-term effects, especially for high values of the conflict.

From Stereoscopy and the Human Visual System :
"It also remains to be determined whether there are any short- or long-term effects of prolonged, repeated exposure to the unnatural stimulus presented by stereoscopic displays. In adults, accommodation-vergence coupling is quite adaptable (Schor & Kotulak, 1986), and so there is a possibility that accommodation function may take some time to return to normal following prolonged viewing of stereo 3D media. Moreover, as the stereo 3D industry continues to develop, our use of stereo media will change from an occasional activity to an everyday one.
The introduction of stereo computer games, in particular, will expose viewers to vergence-accommodation conflicts regularly for potentially long periods of time. We may need to be particularly cautious about long-term effects of vergence-accommodation conflicts on younger children because their visual systems are still developing (Rushton & Riddell, 1999). We know of no specific causes for concern at this time, but the research required to identify relevant issues has not yet been done. It is reasonable to assume that vergence-accommodation coupling exists because it is beneficial, and so one should be cautious when systematically disrupting its natural operation."

Matthew wrote:Even if everything is at infinite distance, setting the screen separation that way will make it a lot more immersive -- it will make it actually appear at infinite distance.
It will make it appear at infinite distance for vergence, but not for accommodation, and that's precisely the problem here. The arrangement of optics used in HMDs like the Oculus Rift will make far objects appear at a correct distance for both vergence and accomodation perception, without inducing eye fatigue or headaches.
virror
Sharp Eyed Eagle!
Posts: 427
Joined: Fri Jan 18, 2013 7:13 am
Location: Gothenburg, Sweden

Re: Updated Vireio Drivers, New Development Thread

Post by virror »

So, when can we expect some news? : D
User avatar
baggyg
Vireio Perception Developer
Vireio Perception Developer
Posts: 491
Joined: Sat May 19, 2012 5:20 am
Location: BB, Slovakia

Re: Updated Vireio Drivers, New Development Thread

Post by baggyg »

virror wrote:So, when can we expect some news? : D
American Thanksgiving, as per their last big update.
virror
Sharp Eyed Eagle!
Posts: 427
Joined: Fri Jan 18, 2013 7:13 am
Location: Gothenburg, Sweden

Re: Updated Vireio Drivers, New Development Thread

Post by virror »

Must have missed that : P
User avatar
drifter
Binocular Vision CONFIRMED!
Posts: 242
Joined: Tue Nov 27, 2012 9:39 am
Location: Little Britain

Re: Updated Vireio Drivers, New Development Thread

Post by drifter »

Fredz wrote:I read a paper about people in a VR lab who misconfigured there software so the rendering was pseudo-stereoscopic and they were surprised that nobody noticed.
It happened to me during 15 good minutes, one day (now I am careful ^^"), retrospectively I was amazed too see how the brain is good at creating depth if it thinks there should be depth.
User avatar
Neil
3D Angel Eyes (Moderator)
Posts: 6882
Joined: Wed Dec 31, 1969 6:00 pm
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Neil »

baggyg wrote:
virror wrote:So, when can we expect some news? : D
American Thanksgiving, as per their last big update.

Well! Just because we've publicized our public build release date, doesn't mean we won't have news leading up to that point! That wouldn't be any fun!

More to come...

Regards,
Neil
Matthew
One Eyed Hopeful
Posts: 41
Joined: Tue Jun 25, 2013 11:45 pm

Re: Updated Vireio Drivers, New Development Thread

Post by Matthew »

That paper has a number of questionable claims.


It is reasonable to assume that vergence-accommodation coupling exists because it is beneficial, and so one should be cautious when systematically disrupting its natural operation.
This is ridiculous. It does not exist because it's beneficial, it exists because it's the only way the human eye can see things clearly in the physical world.


It also remains to be determined whether there are any short- or long-term effects of prolonged, repeated exposure to the unnatural stimulus presented by stereoscopic displays.
Fredz wrote:Maybe it was the case for you, but it may not be the case for everyone. To me it even sounds quite dangerous to tell people to force their visual system in an uncomfortable and headache inducing configuration without knowing the implications.
There certainly aren't any long term effects.

One of the biggest myths about 3D is that it can harm people. Although it is well known that it can cause short-term effects in some people, those effects (at worst) go away within hours (probably minutes). There is no scientific evidence that stereoscopic viewing is capable of causing lasting harm.

The human body has evolved to withstand, and recover from, a significant amount of abuse. It's not going to be harmed by something as trivial as improper displaying of images to the left and right eyes.

Numerous scientific studies have demonstrated the human body's ability to adapt to things. Decoupling of accommodation and vergence is nothing.

It's also important to note that decoupling of accommodation and vergence is unavoidable with currently available stereoscopic viewing technology. Even with smaller amounts of depth, it will still be decoupled, just not by as large an amount. Since currently available stereoscopic displays do not display a 4D light field, everything is at the same focal distance. Technology to display a 4D light field does exist, but won't be commercially available anytime soon.


Fredz wrote:There is a reason why the default values for depth in stereo drivers are quite low and that moviemakers also minimize this conflict
And this is why people are calling 3D a fad and a gimmick.

Very few people realize how terrible the 3D in these movies has been. Avatar was hugely overrated. The three Toy Story movies were a joke. The Life of Pi was practically 2D. And even Final Destination 5 and Hugo, which were done by professional stereographers, were only good some of the time.

It's a huge mistake to "tone down" the 3D just so that people who can't view 3D comfortably can view it. This ruins the 3D for everyone else.

People that can't view 3D comfortably can always watch it in 2D.


drifter wrote:retrospectively I was amazed too see how the brain is good at creating depth if it thinks there should be depth.
Of course the brain will still see depth if the eyes are reversed, but it won't be the correct depth.


Fredz wrote:It will make it appear at infinite distance for vergence, but not for accommodation, and that's precisely the problem here.
It will still appear at infinite distance. It isn't as good as with the Oculus Rift, but it's still spectacular, and is the best that can be done with conventional 3D displays.

The Oculus Rift is certainly better, but it will be some time before such technology becomes mainstream. Until then, properly configured conventional 3D displays are the next best thing.
Last edited by Matthew on Wed Oct 23, 2013 3:26 pm, edited 2 times in total.
User avatar
cybereality
3D Angel Eyes (Moderator)
Posts: 11406
Joined: Sat Apr 12, 2008 8:18 pm

Re: Updated Vireio Drivers, New Development Thread

Post by cybereality »

@Matthew: I agree w/ you that most "3D" movies have a very shallow depth and most of the time I can't even tell it's working at all.
Croccy22
Cross Eyed!
Posts: 140
Joined: Wed Feb 04, 2009 7:51 am

Re: Updated Vireio Drivers, New Development Thread

Post by Croccy22 »

I know it was sort of mentioned a few posts up. But will the new version of Vireio still support DIY Rifts? e.g will if still have the Shared memory tracker plugin for example.

Like rescue gamer. I have a 5" screen based DIY Rift with the 5x Aspheric lenses. With a few tweaks the original Vireio drivers worked very well in DIY Rift mode, but we have small black bars top and bottom.

If I try any software that is designed for the oculus, it doesn't work properly. I will try to explain but I expect you are already aware. On the oculus, since it uses a 7 inch screen, the "eye" centers are not in the middle of a SBS view. They are moved closer to the center of the screen. Also since the oculus has higher magnification lenses the actual image is a lot smaller on the screen.

All we need is the option to move the eye center back outwards so they are at 25% and 75% on the monitor | : | : | (<=Really badly drawn picture of what I mean). And then the option to zoom in so that the image generated fills the entire half of the screen.

Do these options exists. Or can you even understand what the hell I am trying to say :P. It's been a long day ok!

Thanks, Matt.
User avatar
Neil
3D Angel Eyes (Moderator)
Posts: 6882
Joined: Wed Dec 31, 1969 6:00 pm
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by Neil »

Hi Matt,

We'll do our best - I can't promise.

Regards,
Neil
Matthew
One Eyed Hopeful
Posts: 41
Joined: Tue Jun 25, 2013 11:45 pm

Re: Updated Vireio Drivers, New Development Thread

Post by Matthew »

Another of the biggest myths about 3D is that it is new.

Stereoscopy has existed for more than 100 years.

The first time I saw a 3D movie was in May 2003, when I saw Ghosts of the Abyss in an IMAX 3D theater which used liquid crystal shutter glasses. I was 15 years old at the time.

In fact, liquid crystal shutter glasses systems for PCs, and stereo drivers for them which allow you to play DirectX and OpenGL games in stereoscopic 3D, have been commercially available since the mid 1990s. They just didn't become mainstream until recently. I've owned an eDimensional wireless shutter glasses system since early 2008.

And Zalman Trimon circularly polarized 3D monitors have been commercially available since 2007. On my computer system, I'm using a 22" Zalman Trimon monitor that I purchased in late 2009, a few months before Avatar was released. Even back then, it only cost around $300.

And Avatar was not the first movie in this 3D revolution. It was just the one that got everyone's attention. A few months before Avatar was released, another major live-action 3D movie had been released: The Final Destination. It was filmed in 3D using the same stereo camera system used to film Avatar.

When Avatar was released, I was struck by how everyone was talking about 3D as if it were new.

To the people that think Avatar's 3D is good: The Final Destination trumps it.

And Final Destination 5 is even better.

And Ghosts of the Abyss was far better still.

Avatar set a very low standard for the quality of 3D.



I'm shocked that people think The Life of Pi's 3D is good.



Here's a quote from James Cameron about The Life of Pi:
It does what good 3D is supposed to do, which is it allows you to forget you’re watching a 3D movie.
Of course it allows you to forget you're watching a 3D movie. Because it's barely 3D!
JoeReMi
Two Eyed Hopeful
Posts: 50
Joined: Thu Feb 07, 2013 3:41 pm

Re: Updated Vireio Drivers, New Development Thread

Post by JoeReMi »

Is it possible you've become desensitised to the effect? Because I thought Life of Pi had plenty of depth.
virror
Sharp Eyed Eagle!
Posts: 427
Joined: Fri Jan 18, 2013 7:13 am
Location: Gothenburg, Sweden

Re: Updated Vireio Drivers, New Development Thread

Post by virror »

So close now! : D
Alesh
One Eyed Hopeful
Posts: 40
Joined: Sat Jun 22, 2013 2:46 am

Re: Updated Vireio Drivers, New Development Thread

Post by Alesh »

Yeah, I thought it was going to be a midnight release :)
Biting my nails now..
User avatar
baggyg
Vireio Perception Developer
Vireio Perception Developer
Posts: 491
Joined: Sat May 19, 2012 5:20 am
Location: BB, Slovakia

Re: Updated Vireio Drivers, New Development Thread

Post by baggyg »

Hmmmm 11am in Europe. What time does Neil get up? I think this release is probably the most VR excited I've been since the Rift itself arrived.
User avatar
RescueGamer
Cross Eyed!
Posts: 108
Joined: Thu May 16, 2013 1:02 am

Re: Updated Vireio Drivers, New Development Thread

Post by RescueGamer »

$Hype++;
$Patience--;
kieth890
Vireio Perception Developer
Vireio Perception Developer
Posts: 6
Joined: Sat Jul 20, 2013 4:34 pm

Re: Updated Vireio Drivers, New Development Thread

Post by kieth890 »

Wow, you guys did stay up for it! It is released, there are a few outstanding issues we were trying to solve (this is Josh btw) but unfortunately it did not happen tonight. I'll detail more about these later, along with some useful tips for using Vireio, after I get some sleep. Expect a relatively (as in within a month, but hopefully much sooner) quick update to fix some minor issues. And don't hold back on your thoughts, both good and bad!
GAFBlizzard
One Eyed Hopeful
Posts: 5
Joined: Thu Mar 21, 2013 9:28 pm

Re: Updated Vireio Drivers, New Development Thread

Post by GAFBlizzard »

It seems to work at the full framerate with Mirror's Edge now, for the most part!

The one thing that I still need to figure out is crazy flickering. It looks like tearing occasionally, but there is also flickering between the two images, for example on the buildings and/or sky on the main menu.
User avatar
RescueGamer
Cross Eyed!
Posts: 108
Joined: Thu May 16, 2013 1:02 am

Re: Updated Vireio Drivers, New Development Thread

Post by RescueGamer »

GAFBlizzard wrote:It seems to work at the full framerate with Mirror's Edge now, for the most part!

The one thing that I still need to figure out is crazy flickering. It looks like tearing occasionally, but there is also flickering between the two images, for example on the buildings and/or sky on the main menu.
How you get mirrors edge working? For me the game starts, and the two images are showed but with glitches on the corners and menus, but i didn't try gaming itself.
Leahy
Cross Eyed!
Posts: 181
Joined: Thu Jun 07, 2012 1:36 pm
Location: France

Re: Updated Vireio Drivers, New Development Thread

Post by Leahy »

Thanks to everyone for all their hard work getting 2.0 out. It's like getting my Rift all over again.
User avatar
LordJuanlo
Binocular Vision CONFIRMED!
Posts: 224
Joined: Thu Jul 26, 2012 12:55 am
Location: Spain
Contact:

Re: Updated Vireio Drivers, New Development Thread

Post by LordJuanlo »

Big thanks to the developers for making Vireio 2.0 possible.
Image
Comunidad española de RV / Spanish VR Community
User avatar
baggyg
Vireio Perception Developer
Vireio Perception Developer
Posts: 491
Joined: Sat May 19, 2012 5:20 am
Location: BB, Slovakia

Re: Updated Vireio Drivers, New Development Thread

Post by baggyg »

Thanks guys for the work. The more I play with it the better it is.

Left 4 Dead in particular is flawless. Latency amazing!

Skyrim also "VR Ready" apart from a few shadow issues.
GAFBlizzard
One Eyed Hopeful
Posts: 5
Joined: Thu Mar 21, 2013 9:28 pm

Re: Updated Vireio Drivers, New Development Thread

Post by GAFBlizzard »

RescueGamer wrote:
GAFBlizzard wrote:It seems to work at the full framerate with Mirror's Edge now, for the most part!

The one thing that I still need to figure out is crazy flickering. It looks like tearing occasionally, but there is also flickering between the two images, for example on the buildings and/or sky on the main menu.
How you get mirrors edge working? For me the game starts, and the two images are showed but with glitches on the corners and menus, but i didn't try gaming itself.
I have the Steam version of Mirror's Edge, and running the Perception utility did not seem to do anything. So I closed the Perception utility, copied the three DLL's into the Mirror's Edge binary directory, and ran the game from Steam. It seemed to come up, and the Ctrl-T menu actually works, but due to the tearing and flickering I was afraid I would give myself a seizure or something. :P

I also tried running Project Cars in D3D9 mode, and the main menu actually DOES get split in half which I think is better than the old version of Perception did, but there is a popup error crash (illegal operation, break or exit, or something like that) if you try to actually start a race.

Cool progress though!
Post Reply

Return to “Development / General Discussion”