The Problem with Nvidia's 3D Surround.

[spigzone]

INHERENT INSTABILITY

Here is the problem I see with current 3D technology. Human eyes are evolved
to work in perfect spacial syncronicity. They move in unison. They focus in
unison. They unfocus in unison. They ACT AS ONE. Depth perception is
accomplished by each eye mechanically changing it's focal length in syncronicity
with the other eye.

But what happens with 'simulated' 3D? What appears on the monitor screen is
two offset images. When one puts the 3D glasses on each eye only sees one of
those two images. They are no longer focusing as one, they are each focusing
on a slightly different image. This is an unnatural state and creates an inherent
instability in the eye/brain visual system. The brain resolving that instability is
what simulates the perception of depth where none physically exists. But there is
a cost to this, the reported queasiness, headaches and fatigue, worsening over
time.

But what actually physically happens? The eyes normally change their focal
length to perceive depth, but a monitor screen is flat, it has no depth (in practical
terms). If the eyes actually changed their physical focal length to resolve the
simulated 3D image the entire screen would go out of focus. Therefore 'focusing'
while looking at the simulated 3D image must be accomplished by other means.

The eyes are reflexively trying to bring those disseparate images together while
the brain is trying to make sense of it all and fit it into existing visual paradigms.
This problem is inherent to any 3D system using offset images and any such
system will always generate negative word of mouth over time and a percentage
of the population will be unable to tolerate it at all.

INHERENT LAG

Here are the problems I see with 3D multiple monitor gaming. When looking at a
monitor in 2D the eyes are able to transverse the screen and access the
information contained therein very fast ... in biomechanical time. Hardwired fast
as it were. With 3D there is an added 'software' layer added in. Accessing
information now requires the eye/brain system to FIRST percieve and then
'focus' at the depth information is at (otherwise it will be out of 'focus') and
THEN perceive and process the information pertinent to the gaming action. It
introduces a time lag to information retrieval and processing. In a twitchy shooter
that inherent time lag is likely to be fatal. The gaming penalty would, of course,
be proportional to the reflex speed required to successfully play a game.

With multiple monitors, the situation worsens. The inherent advantage of wing
monitors, peripheral vision heads up and the ability to glance over to access
additional information, is largely negated. There is a physial impediment to
peripheral vision, the glasses themselves curtail the ability to 'glance' over at the
screens without moving the head, it is necessary to physically move the head
and then go through the focusing software layer to access the information
contained in the side screens, and there is an attention impediment, that artificial
focusing software layer is creating a tunnel vision effect, absorbing available
attention units as it were. When one is looking at the main screen continually
sorting through the depth layers to see vital information, information coming in
from the periphery, from the side screens, is likely to get short shrift. And always
the headache/fatigue factor, increasing over time, inherent in 3D itself and
exacerbated by the NEED to continually focus in to access gaming information.
In a fast moving 3D game that NEED will drive the eye/brain system to work very
very hard, in an unnatural way, without a break and for lengthy periods of time.

ACTIVE vs PASSIVE

When watching movies, the effect is minimized as it is a relatively passive experience (although even then the effect can be very unpleasant). But in a game one is actively participating and interacting, there are immediate and unpleasant consequences if one is not focusing and gathering information one NEEDS to.

In terms of eyestrain, when watching a movie one has the luxury of allowing the eyes to relax, wander about, focus on whatever, albiet at the penalty of missing some of the 'action', but in a game there is a constant press to keep focusing, keep looking, there is no time for the eyes to relax, unfocus.This exacerbates the already inherent problems of 3D viewing.

HARDCORE vs CASUAL vs COST

What is the demographic composition of those willing to put out the money it
would take to create a 3D multiple monitor system? Three 3D capable screen
plus what is likely 2 Fermi cards in SLI in a high end gaming system = $4000
and up.

'Hardcore' gamers, those willing to buy cutting edge systems and graphic cards
because they provide an advantage in the games they play, are those least likely
to gain a gaming advantage from a 3D multi-monitor setup for the reasons
delineated above. Bragging about your new $6000 3D multimonitor setup and
then having your ass repeatedly handed to you by your peers, followed by their
hoots and derision, isn't going to be a pleasant experience. The casual gamers
might like the 3D surround experience, and would probably most benefit from it,
but how many will pay the price to have it in their homes?

So what exactly is the demographic for a very pricey 3D multi-monitor setup?

There's always that tiny sliver of the market with money to burn and who want
the cutting edge experience, no matter what, but that ain't gonna pay the bills.

 

[Spare-Flair]

Well, your critcisms regarding syncronicity and natural focus isn't an nvidia specific problem, it is a problem with any 3D technology.

This technology is going to have a relatively low adoption because of the prohibitive cost (triple 120Hz monitors, Fermi SLI, etc.) and will be more for high-end single user experience. I don't think anybody concerned with playing multiplayer competitively would be using something like this so the advantage/disadvantage in gaming is irrelevant.

Really for Nvidia, this is more or less a no-brainer and at no real cost to them nomatter how popular it ultimately becomes since they've already developed the technology. They are simply tacking this onto their version of Eyefinity to trump AMD on paper.

 

[WorldExclusive]

Interesting read. The best part was:

Bragging about your new $6000 3D multimonitor setup and
then having your ass repeatedly handed to you by your peers, followed by their
hoots and derision, isn't going to be a pleasant experience.

Really for Nvidia, this is more or less a no-brainer and at no real cost to them nomatter how popular it ultimately becomes since they've already developed the technology. They are simply tacking this onto their version of Eyefinity to trump AMD on paper.

ATI 3D Eyefinty. The playing field is even, now it comes down to drivers and if Nvidia will lock ATI out of certain games.
http://www.hardforum.com/showthread.php?t=1484959

 

[spigzone]

Seems to me CES just exacerbated existing power/performance/running hot concerns about Fermi's with the extra power wires controversy and using SLI and watercooling to drive the demonstratons.

An air cooled single Fermi board driving the demos would have been effective to quiet those qualms. Obvious conclusion is a single air cooled board couldn't run the demos.

Meanwhile over at AMD every demo was with a single air cooled board.

 

[WorldExclusive]

Their lower level card, the GF100, which is probably equal to ATI 5850 is using 6/8pin connectors. That points to the issues of the card being too hot. You an only imagine the GF104 part(5870 equivalent), two 8pin connectors maybe.

 

[spigzone]

Well, your critcisms regarding syncronicity and natural focus isn't an nvidia specific problem, it is a problem with any 3D technology.

This technology is going to have a relatively low adoption because of the prohibitive cost (triple 120Hz monitors, Fermi SLI, etc.) and will be more for high-end single user experience. I don't think anybody concerned with playing multiplayer competitively would be using something like this so the advantage/disadvantage in gaming is irrelevant.

Really for Nvidia, this is more or less a no-brainer and at no real cost to them nomatter how popular it ultimately becomes since they've already developed the technology. They are simply tacking this onto their version of Eyefinity to trump AMD on paper.

Maybe true, but what percentage of that 'high-end single user experience' is hard core gamers willing to buy pricey cutting edge processors and video cards to beat games at their highest difficulty setting? As I understand the hard core gamers, single and multiplayer, are the biggest market for high end video cards.

It may be a no brainer, but in an opposite way. If the 3D Surround idea falls flat on it's face, which is highly likely, it is not going to reflect favorably on Nvidia. It will be considered a failure, a failure associated primarily with Nvidia as they introduced it, and adding to the accumulating PR burden.

It may not be a major cost, but Nvidia can ill afford any additional unnecessary such associations at this point.

 

[XMAN245]

INHERENT INSTABILITY

Here is the problem I see with current 3D technology. Human eyes are evolved
to work in perfect spacial syncronicity. They move in unison. They focus in
unison. They unfocus in unison. They ACT AS ONE. Depth perception is
accomplished by each eye mechanically changing it's focal length in syncronicity
with the other eye.

But what happens with 'simulated' 3D? What appears on the monitor screen is
two offset images. When one puts the 3D glasses on each eye only sees one of
those two images. They are no longer focusing as one, they are each focusing
on a slightly different image. This is an unnatural state and creates an inherent
instability in the eye/brain visual system. The brain resolving that instability is
what simulates the perception of depth where none physically exists. But there is
a cost to this, the reported queasiness, headaches and fatigue, worsening over
time.

But what actually physically happens? The eyes normally change their focal
length to perceive depth, but a monitor screen is flat, it has no depth (in practical
terms). If the eyes actually changed their physical focal length to resolve the
simulated 3D image the entire screen would go out of focus. Therefore 'focusing'
while looking at the simulated 3D image must be accomplished by other means.

The eyes are reflexively trying to bring those disseparate images together while
the brain is trying to make sense of it all and fit it into existing visual paradigms.
This problem is inherent to any 3D system using offset images and any such
system will always generate negative word of mouth over time and a percentage
of the population will be unable to tolerate it at all.

INHERENT LAG

Here are the problems I see with 3D multiple monitor gaming. When looking at a
monitor in 2D the eyes are able to transverse the screen and access the
information contained therein very fast ... in biomechanical time. Hardwired fast
as it were. With 3D there is an added 'software' layer added in. Accessing
information now requires the eye/brain system to FIRST percieve and then
'focus' at the depth information is at (otherwise it will be out of 'focus') and
THEN perceive and process the information pertinent to the gaming action. It
introduces a time lag to information retrieval and processing. In a twitchy shooter
that inherent time lag is likely to be fatal. The gaming penalty would, of course,
be proportional to the reflex speed required to successfully play a game.

With multiple monitors, the situation worsens. The inherent advantage of wing
monitors, peripheral vision heads up and the ability to glance over to access
additional information, is largely negated. There is a physial impediment to
peripheral vision, the glasses themselves curtail the ability to 'glance' over at the
screens without moving the head, it is necessary to physically move the head
and then go through the focusing software layer to access the information
contained in the side screens, and there is an attention impediment, that artificial
focusing software layer is creating a tunnel vision effect, absorbing available
attention units as it were. When one is looking at the main screen continually
sorting through the depth layers to see vital information, information coming in
from the periphery, from the side screens, is likely to get short shrift. And always
the headache/fatigue factor, increasing over time, inherent in 3D itself and
exacerbated by the NEED to continually focus in to access gaming information.
In a fast moving 3D game that NEED will drive the eye/brain system to work very
very hard, in an unnatural way, without a break and for lengthy periods of time.

HARDCORE vs CASUAL vs COST

What is the demographic composition of those willing to put out the money it
would take to create a 3D multiple monitor system? Three 3D capable screen
plus what is likely 2 Fermi cards in SLI in a high end gaming system = $4000
and up.

'Hardcore' gamers, those willing to buy cutting edge systems and graphic cards
because they provide an advantage in the games they play, are those least likely
to gain a gaming advantage from a 3D multi-monitor setup for the reasons
delineated above. Bragging about your new $6000 3D multimonitor setup and
then having your ass repeatedly handed to you by your peers, followed by their
hoots and derision, isn't going to be a pleasant experience. The casual gamers
might like the 3D surround experience, and would probably most benefit from it,
but how many will pay the price to have it in their homes?

So what exactly is the demographic for a very pricey 3D multi-monitor setup?

There's always that tiny sliver of the market with money to burn and who want
the cutting edge experience, no matter what, but that ain't gonna pay the bills.

Jesus Christ man, do you ever stop with wild, out of left field assuptions concerning Nvidia? You are doing the same shit at Beyond3D.com and have been called out on it by several people so now you come here to do more of the same.

 

[XMAN245]

Seems to me CES just exacerbated existing power/performance/running hot concerns about Fermi's with the extra power wires controversy and using SLI and watercooling to drive the demonstratons.

An air cooled single Fermi board driving the demos would have been effective to quiet those qualms. Obvious conclusion is a single air cooled board couldn't run the demos.

Meanwhile over at AMD every demo was with a single air cooled board.

Seriously, you need to stop with your assumptions about Fermi power draw and TDP, its getting stupid. You've had you ass handed to you at B3D so now you come here to try your antics on another group of people.

 

[digitalwanderer]

I don't say this a lot here, but people ignore the troll!

http://forum.beyond3d.com/showthread.php?t=56088

 

[akotlar]

I don't say this a lot here, but people ignore the troll!

http://forum.beyond3d.com/showthread.php?t=56088

I read your response over at B3D. I don't see the response here being much better. I think you're reacting more to the haughty tone implied by the length of the post, and by the fact that he isn't a well-respected poster, rather than some problem with what he's written. I don't see this as much different than the 135 page Fermi thread on B3D, which can be succinctly written as "Fermi is hot, needs 2 8W + 1 6W, adds compute features of questionable value to most of the standard consumer GPU target market..." Doesn't seem like much less of a "doom and gloom" analysis than what this guy is doing, outside of the fact that B3D's "doom & gloom" posts are written by better educated posters, or by posters who are popular on the boards. I don't see a reason to get on his back. The B3D replies were frankly rude. I'm sure they'll be similar here. I'm not on his side, but I do think that you should spend a little more time tearing apart his swiss-cheese post, and a bit less attacking his character and intentions.

 

[digitalwanderer]

Did you also see a similarity in the 44 posts he's put up at B3D mostly in the thread you refer to? They all seem to follow a similar theme, and some of his posts have been flat-out bonkers.

I don't judge by postcount but by context, and that's what I based my comment on and I stand by it.

 

[phide]

The playing field is even, now it comes down to drivers and if Nvidia will lock ATI out of certain games.

At this time, I don't believe it's even theoretically possible for NVIDIA to do this, Kyle's commentary notwithstanding.

 

[kuyaglen]

The multiple times I've played on Nvidia's 3D demo systems, I've never once gotten any hint of any "lag" or "instability". IMO the op is all "sound and fury, signifying nothing".

 

[XMAN245]

I read your response over at B3D. I don't see the response here being much better. I think you're reacting more to the haughty tone implied by the length of the post, and by the fact that he isn't a well-respected poster, rather than some problem with what he's written. I don't see this as much different than the 135 page Fermi thread on B3D, which can be succinctly written as "Fermi is hot, needs 2 8W + 1 6W, adds compute features of questionable value to most of the standard consumer GPU target market..." Doesn't seem like much less of a "doom and gloom" analysis than what this guy is doing, outside of the fact that B3D's "doom & gloom" posts are written by better educated posters, or by posters who are popular on the boards. I don't see a reason to get on his back. The B3D replies were frankly rude. I'm sure they'll be similar here. I'm not on his side, but I do think that you should spend a little more time tearing apart his swiss-cheese post, and a bit less attacking his character and intentions.

I believe he got banned at B3D for doing what the MOD believed to be trolling and nothing more.

 

[akotlar]

Did you also see a similarity in the 44 posts he's put up at B3D mostly in the thread you refer to? They all seem to follow a similar theme, and some of his posts have been flat-out bonkers.

I don't judge by postcount but by context, and that's what I based my comment on and I stand by it.

No I didn't, I haven't followed him at all. I'd actually seen only one post of his, something about "think logarithmic", which struck me as a math abuse.

I just checked some of his posts at B3D (he's banned already), and stuff like:

"It simply means Nvidia doesn't have a magic wand to wave away the laws of physics.
Fermi being ~ 50% bigger than Cypress should roughly translate to 50% higher tdp - 188w x 150% = 282w ... as a rough baseline and assuming they need to put all of the chip to work. That's physics 101.
Fermi 'straddling' GPU and GPGPU means a large chunk of that real estate is non-optimized for pure GPU tasks.
Fermi competing with Cypress on cost performance means they have to wring the last bit of GPU specific performance they can from the chip ... maximum realizable clock speeds, drivers optimized to wring as much GPU performance from the GPGPU real estate as possible ... do they even have the option of doing otherwise, they absolutely HAVE to exceed the 5870 specs by some appreciable margin.
Add in it's a new and very complicated chip design on a new process and the time element, having to go with the earliest usable (non-optimized) silicon spin possible, why wouldn't Fermi be pushing the 300 watt envelope? Or even at 300 watts without the clock speeds being where they need to be? "

doesn't stink all that bad; the figures & assumptions he's using are ok, and while I can see that he reeks like an nvidia hater, and most of the conclusions are not his own, it really isn't massive complaint worthy. Again I think the biggest issue is that he's writing like he knows something, but the reality of the post doesn't quite match up.

I think you guys are seeing the "Uncanny Valley". I was going to make a reference to pseudo-Ben Affleck in Final Fantasy, then thought of Ben Affleck being a pseudo-intellectual as a stand in for Will in the job interview scene, and then couldn't decide which one worked better.

 

[akotlar]

I believe he got banned at B3D for doing what the MOD believed to be trolling and nothing more.

Right, and I understand that. I just don't see a massive difference in the actual content of his posts and that of some of the other bottom feeders on B3D's forums, which is why I think some people their are reacting to what they perceive as haughty posting habits. He's clearly stepping on some toes. I don't see anything ban-worthy there. I do see a lot that people could disagree with, but forums shouldn't be regulated so closely that those who choose a disparate stance, or those who argue less eloquently should be banned by what is essentially mod's disagreement. Doesn't do much for the spirit of free speech. His posts aren't the most intelligent, but he's not being belligerent to other posters, and that is what should be constitute ban-criteria.

As for his post, I'd have to say that this idea that "it takes your eyes too much time to focus on 3D images, hence lag and eye fatigue" is ridiculous. How do you manage to go through a whole day looking around with your nearly 180 degrees of peripheral vision and not just wither in agony ? What eye patch do you recommend?

Also I don't quite see why it is such a problem for nVidia to release 3D support for multi-monitor gaming based on the argument of rendering power. It is clear that nVidia is doing this mostly as a marketing move, as I doubt many would be able to modern games in 3D at 6MP resolutions, but it is just a feature. It will also certainly be useful going forward. I would agree with Kyle's editorial in that the biggest issue is their likely push for proprietary exclusion.

A real question: How well do rotations work with active-shutter devices? Active-shutter glasses are basically vertically polarized right? Not that they are at all polarized, but the left & right eyes composite vertical strips, so the image is not orthogonal, and will not react well with rotation? So shouldn't the same artifacts that you have with something like Imax 3D LP glasses be present here? That might cause some problems when looking around.

 

[spigzone]

Jesus Christ man, do you ever stop with wild, out of left field assuptions concerning Nvidia? You are doing the same shit at Beyond3D.com and have been called out on it by several people so now you come here to do more of the same.

Do you ever reply based on the the merits of the physics and logic presented?

Or do you just fling wild accusations like a zoo chimpanzee flings it's own shit?

 

[XMAN245]

Do you ever reply based on the the merits of the physics and logic presented?

Or do you just fling wild accusations like a zoo chimpanzee flings it's own shit?

I would if what you posted wasn't wild assed assumptions based on ill gotten logic in of itself. You are nothing more than a troll, your banning from B3D only cements that fact. Go back to SA where you more than likely originated from.

And look in the mirror there dude, You are the one tossing shit around.

 

[spigzone]

Right, and I understand that. I just don't see a massive difference in the actual content of his posts and that of some of the other bottom feeders on B3D's forums

While I appreciate the sentiment of fairplay, I would pose the quality of my thread post rises far above anything a 'bottom feeder' comes up with.

I consider it a well thought out, comprehensive, well reasoned, scientifically grounded and quite unique look at some potential problems with 3D multi-monitor gaming and the ramifications thereof.

What I have noticed is no one has argued the science, physics or logic of my article. Here or on Beyond 3D.

 

[spigzone]

I would if what you posted wasn't wild assed assumptions based on ill gotten logic in of itself. You are nothing more than a troll, your banning from B3D only cements that fact. Go back to SA where you more than likely originated from.

And look in the mirror there dude, You are the one tossing shit around.

Do you ever reply based on the the merits of the physics and logic presented?

 

[akotlar]

While I appreciate the sentiment of fairplay, I would pose the quality of my thread post rises far above anything a 'bottom feeder' comes up with.

I consider it a well thought out, comprehensive, well reasoned, scientifically grounded and quite unique look at some potential problems with 3D multi-monitor gaming and the ramifications thereof.

What I have noticed is no one has argued the science, physics or logic of my article. Here or on Beyond 3D.

Yeah, that was a bit harsh. I don't think your posts are as well presented as you seem to think, but at least you're attempting an examination. What you're doing however isn't science or physics, but is reminiscent of scientific musing.

You're completely wrong on framerate though. Only refresh rate matters. You could have a 20 FPS game running in 3D just fine, since framerate is not linked to the number of refreshes the screen does, and the latter is what the LCD's in the glasses sync to.

You need at least 120hz, and that would probably not be enough for long viewing. While many may assume that, since most LCD's are 60hz, 120hz/2 will work well, they are not taking into account that active-shutter glasses effectively introduce 60 frames of black into the mix. So you are much closer to getting the viewing experience of a 60hz CRT which, in my opinion, is not sufficient.

 

[akotlar]

I would if what you posted wasn't wild assed assumptions based on ill gotten logic in of itself. You are nothing more than a troll, your banning from B3D only cements that fact. Go back to SA where you more than likely originated from.

And look in the mirror there dude, You are the one tossing shit around.

Honestly, this has no place here. I'm not sure how personal attacks are better than ill-though out positions.

I still don't see a compelling reason why 3D support shouldn't be released. It seems like a value-added feature, and certainly is something you don't need to use (as Dave Baumann pointed out in response to one of your posts). It is clearly not going to be practical this generation for anything other than 2*Fermi, which at >600W real world use is going to be a tough sell (I'm not buying it at least). And there are certainly potential issues with peripheral vision & image-splitting, unless I'm not understanding active-shutter tech well.

 

[Nenu]

Do you ever reply based on the the merits of the physics and logic presented?

Logic is one thing but you need facts for it to be useful.
What you posted isnt fact, its what you are thinking and I dont agree.

 

[XMAN245]

Honestly, this has no place here. I'm not sure how personal attacks are better than ill-though out positions.

I still don't see a compelling reason why 3D support shouldn't be released. It seems like a value-added feature, and certainly is something you don't need to use (as Dave Baumann pointed out in response to one of your posts). It is clearly not going to be practical this generation for anything other than 2*Fermi, which at >600W real world use is going to be a tough sell (I'm not buying it at least). And there are certainly potential issues with peripheral vision & image-splitting, unless I'm not understanding active-shutter tech well.

I see you are apply his same, out of nowhere power used method for what a single Fermi consumer level card will consume. Noone know because Nvidia hasn't said anything yet.

 

[spigzone]

I still don't see a compelling reason why 3D support shouldn't be released. It seems like a value-added feature, and certainly is something you don't need to use (as Dave Baumann pointed out in response to one of your posts). It is clearly not going to be practical this generation for anything other than 2*Fermi, which at >600W real world use is going to be a tough sell (I'm not buying it at least). And there are certainly potential issues with peripheral vision & image-splitting, unless I'm not understanding active-shutter tech well.

Value added is subjective to most. If cost is not a consideration, and using the 3D function in a multi-monitor setup proves untenable and one goes back to 2D, no big deal. If one scrimped and saved and worked hard to put together a 3D capable multi-monitor setup and the video power to run it and it doesn't provide a better gaming experience, or even worsens the gaming experience, one can indeed simply go back to 2d gaming, but the money down the drain is going to rankle. It's only a value added feature if it, on balance, adds value. If it TAKES value, it's a net negative. In either case the word of mouth will be negative.

If that net negative becomes the rule and not the exception, word will spread and 3D multi-monitor gaming will accrue a bad reputation. If Nvidia is the company that comes to mind when 3D multi-monitor gaming is brought up, and right now it owns that association, and the subject triggers a negative response, Nvidia will share in that association. The entire science of positioning in advertising is based on that phenomenon. Associate your product with what evokes pleasant or positive responses.

It's not major, but Nvidia is presently facing strong PR headwinds and there is no letup is sight. Why set yourself up for an additional negative association in the minds of either the gamers or the investors?

On the other hand, if I have my head up my ass on my logic and reasoning, then Nvidia comes out of this shining.

But until someone can provide sound specific convincing arguments to challenge my reasoning and logic, I consider that will not be the case.

 

[spigzone]

Logic is one thing but you need facts for it to be useful.
What you posted isnt fact, its what you are thinking and I dont agree.

Then provide sound logical merit based specific arguments that refute what you disagree with.

 

[akotlar]

I see you are apply his same, out of nowhere power used method for what a single Fermi consumer level card will consume. Noone know because Nvidia hasn't said anything yet.

I'm not using any assumptions here. Unless you can give me another convincing argument for why the device would need 2 8pin + 1 6pin power adapters? This is what I see here: http://www.techtree.com/India/News/CES_2010_Nvidia_FERMI_Out_in_the_Wild/551-108605-581.html

Here's a bit of logic: Tesla with 448/512 processors, at what is most likely a lower than consumer part clock speed, barely fits within the 225W envelope. The next power envelope step is 300W (and limit of 2.0). Hence, if you're already need to exceed 225W, and the extra cost associated with requiring 2 8pin power (lost sales, and power circuitry expenses), it makes sense for your part to max out 300W envelope, since this will give you the most competitive card without added cost. This is assuming they can clock high enough with 512 cores to max out the power envelope before reaching other engineering limitations.

Finally, as for why they need that 2+1 config, my guess is that the 2 8pin connectors are simply feeding the same 8pin male connector, and having 2 power feeds gives Fermi a more stable power draw. The 6 pin is needed to bring it up to 300W. So Fermi only has 1 8pin + 1 6 pin, but needs the 2nd 8pin PSU feed to give it clean power at its 3D clock. This indicates that Fermi can potentially exceed 300W, or runs within such tight tolerances that a dirty feed gives it stability problems.

This isn't "uneducated speculation". It's guesswork based on announced Fermi-based cards, the pictures we have of 2 8pin + 1 6pin Geforce, and logic. And clean up your grammar, it is embarrassing.

 

[Nenu]

Then provide sound logical merit based specific arguments that refute what you disagree with.

Provide a sound logical argument in the first place.
Each eye perceives a flat plane in the same way that a screen provides in 3D mode (as long as you dont move much), so your arguments dont make sense.

Unless you have proof of extra lag, I dont buy it.
I've used 3D many many times and have a lot of old 3D hardware. I have never experienced this lag you speak of.

From the current standpoint it looks like you are making things up.

 

[akotlar]

Provide a sound logical argument in the first place.
Each eye perceives a flat plane in the same way that a screen provides in 3D mode (as long as you dont move much), so your arguments dont make sense.

Unless you have proof of extra lag, I dont buy it.
I've used 3D many many times and have a lot of old 3D hardware. I have never experienced this lag you speak of.

From the current standpoint it looks like you are making things up.

The extra lag is nonsense, beyond rendering time. But the "as long as you don't move much" is something worth discussing in my eyes. I don't know how much rotation affects active-shutter systems, and that's important, since that's what we're using until polarized screens come along, or Dolby-3D can be implemented (I'm reading that consumer displays aren't good enough on the color accuracy side).

Regardless I think we can all agree that its always nice to have the potential to use a compelling feature, even if it isn't at all realistic for most of us.

 

[spigzone]

Provide a sound logical argument in the first place.
Each eye perceives a flat plane in the same way that a screen provides in 3D mode (as long as you dont move much), so your arguments dont make sense.

Unless you have proof of extra lag, I dont buy it.
I've used 3D many many times and have a lot of old 3D hardware. I have never experienced this lag you speak of.

From the current standpoint it looks like you are making things up.

"Each eye perceives a flat plane in the same way that a screen provides in 3D mode (as long as you dont move much)"

I do not find that sentence comprehensible. Please expand that thought in a more structured cogent manner that an honest and stimulating debate might ensue.

 

[spigzone]

The extra lag is nonsense, beyond rendering time.

Perhaps you could provide specific refutation of the logic and reasoning I used to establish the reason why lag occurs?

 

[Nenu]

The extra lag is nonsense, beyond rendering time. But the "as long as you don't move much" is something worth discussing in my eyes. I don't know how much rotation affects active-shutter systems, and that's important, since that's what we're using until polarized screens come along, or Dolby-3D can be implemented (I'm reading that consumer displays aren't good enough on the color accuracy side).

Regardless I think we can all agree that its always nice to have the potential to use a compelling feature, even if it isn't at all realistic for most of us.

Rotation doesnt figure because the 3D pictures are from a static viewpoint.
...unless you use head tracking, but that doesnt seem to cause any extra issues so I believe it to be a moot point.

Colour accuracy detracts from 3D only when it is very bad, like with red/blue glasses.
Shutter glasses need you to turn the contrast/brightness up as each eye receives 1/4 of the light they get in 2D, perhaps this is what you meant?
ie each frame is present for 1/2 the time in each eye and LCD polarising reduces light by a further 50%.

Absolutely I agree that 3D is cool, I'm buying a projector and soooo want to get a 120Hz 1080p model for 3D but they are a fortune.
For now I'm getting a 60Hz version and will get 3D later

 

[Nenu]

"Each eye perceives a flat plane in the same way that a screen provides in 3D mode (as long as you dont move much)"

I do not find that sentence comprehensible. Please expand that thought in a more structured cogent manner that an honest and stimulating debate might ensue.

lol, its not hard to understand, sit still and look round your room with one eye.
If you dont move, you see a flat plane.

 

[Nenu]

Perhaps you could provide specific refutation of the logic and reasoning I used to establish the reason why lag occurs?

covered in post #28

 

[akotlar]

Perhaps you could provide specific refutation of the logic and reasoning I used to establish the reason why lag occurs?

Sure. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC535985/

It takes the eye 200ms to focus on a new stimulus. So you're dealing with a ~ 200ms latency as a base, even with 2D games.

Smooth motion, vergence & divergence, and vestibular movements all occur within 100-160ms or so. There is no difference in terms of latency when tracking in 2D or 3D. This is why in every day life, it seems that there is no visual delay introduced by swerving your head, or tracking an object into the background.

Tracking a single object moving from the background to the foreground (although I only read the abstract, I'm sure there is a mention of focal distance there) takes 640ms, but unless the object you're tracking moves faster than a very small fraction of that time between the far background and the closest near-ground, you will not notice any additional tracking latency. Again, this greatly depends on the focal point, as unless the change in depth is great enough, your lens will not have to make a maximal adjustment (think LCD state changes). Think about putting a book right up to your face, until you can just barely read the words, and then having someone jump back with the book 3 feet, while you try to keep a focus on the words. It's not a problem you'll encounter in most games, and that is only to completely focus on the object. You'll still see it, it will just be somewhat out of focus. Which again means that unelss you're trying to read fine print as it is rushing towards you, this isn't a problem. This is the only case where gaming in 3D could "put you at a disadvantage", and if you think about it carefully, you'll see that it really is not a problem (and is offset by the advantage you have from motion-prediction based on depth).

 

[spigzone]

Sure. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC535985/

It takes the eye 200ms to focus on a new stimulus. So you're dealing with a ~ 200ms latency as a base, even with 2D games.

Smooth motion, vergence & divergence, and vestibular movements all occur within 100-160ms or so. There is no difference in terms of latency when tracking in 2D or 3D. This is why in every day life, it seems that there is no visual delay introduced by swerving your head, or tracking an object into the background.

Tracking a single object moving from the background to the foreground (although I only read the abstract, I'm sure there is a mention of focal distance there) takes 640ms, but unless the object you're tracking moves faster than a very small fraction of that time between the far background and the closest near-ground, you will not notice any additional tracking latency. Again, this greatly depends on the focal point, as unless the change in depth is great enough, your lens will not have to make a maximal adjustment (think LCD state changes). Think about putting a book right up to your face, until you can just barely read the words, and then having someone jump back with the book 3 feet, while you try to keep a focus on the words. It's not a problem you'll encounter in most games, and that is only to completely focus on the object. You'll still see it, it will just be somewhat out of focus. Which again means that unelss you're trying to read fine print as it is rushing towards you, this isn't a problem. This is the only case where gaming in 3D could "put you at a disadvantage", and if you think about it carefully, you'll see that it really is not a problem (and is offset by the advantage you have from motion-prediction based on depth).

This addresses factors in the natural 3D physical space the eye/brain system evolved to operate in. My point specifically deals with the un-natural 3D world of a flat plane containing offset images, each particular to one eye only, from which the eye/brain system then has to construct a facimile of 3D space where none physically exists. The human visual system did not evolve to operate in that enviroment. There will be an inherent and inevitable lag involved in making the additional computations and constructs within the brains circuitry as there is no existing neuronic circuitry hardwired to do this. Perforce it need be done by a make-around, utilizing more general 'software' mechanisms in the brain, using additional visual processing machinery and resources and leaving less available for other functions. This is the reasoning behind the "With 3D there is an added 'software' layer added in." and the " and there is an attention impediment, that artificial focusing software layer is creating a tunnel vision effect, absorbing available attention units as it were." sentences.

It is that which introduces an additional lag into the action of perceiving, assimilating, processing and then acting upon information contained in the changing pixels of a monitor.

While I appreciate your taking the time to run down and present information relating to the subject, an argument of your own reasoning based on the data presented would be nice.

 

[akotlar]

This addresses factors in the natural 3D physical space the eye/brain system evolved to operate in. My point specifically deals with the un-natural 3D world of a flat plane containing offset images, each particular to one eye only, from which the eye/brain system then has to construct a facimile of 3D space where none physically exists. The human visual system did not evolve to operate in that enviroment. There will be an inherent and inevitable lag involved in making the additional computations and constructs within the brains circuitry as there is no existing neuronic circuitry hardwired to do this.

While I appreciate your taking the time to run down and present information relating to the subject, an argument of your own reasoning based on the data presented would be nice.

I don't think that is true though. If anything the latency will be smaller on average, as your eyes are already forced on the lines of convergence, and because the maximum degree of perceived depth is held within tight bounds, whereas there is an evaluative period in holographic 3D space, and real depth causes distance-tied re-evaluation. If the convergence breaks, there will be added latency and that will be from the time it takes your lenses to adapt (the same thing happens with Magic Eye illustrations).

It might be worthwhile conducting a study on this. I'm not sure if there is merit to what you're saying, but it isn't unreasonable to think that there would be an added latency in corner cases (cases where convergence is lost) with this kind of technology.

However, I think that once convergence is made, latency should now be within that 100-160ms zone. The reason for this is that the depth is recorded (or simulated) by lenses offset by the same dimensions the average human eyes are. You then force the offset images to converge using the glasses, and the stereo image is formed. Basically what I mean by the first 2 sentences is that the parallax is already "baked" in, and therefore your eyes are always focused at the correct depth to observe all objects in focus (not in the "depth of field blur"). There may be some limited potential to focus on more distant objects in the frame, but this range is far smaller than in holographic space, and therefore over the entire range average latency should be smaller than in real life. And as I showed with that previous study, this average latency is really no greater than that for sarcaadic(sp?) motion tracking.

One thing that COULD have an effect (and is mentioned in that study I posted), is that fatigue greatly increases this latency. It is possible that the strain of forced convergence causes reduced responsiveness. I imagine though that there is an offsetting increase in responsiveness due to your eyes having fewer degrees of freedom within which they can operate & still keep convergence (In the statistical sense not physical...since you are operating within a 2D plane. This means that your eyes only have to make assumptions for 1 power less information...although this could be the reason behind the limited depth of these pseudo 3D images and therefore confer no added opportunity). If your brain operates heuristically, it seems that it could quickly "learn" that the 3D projection has what is essentially infinite focus - that depth relationships between moving objects can be handled by less complex rules - and can spend less time changing the convexity of the lens.

I've watched Avatar 3D, and I had some ability to re-focus my eyes to different depth slices. However, in general it was not necessary. I certainly did not notice any greater latency to focus than I do with 2D images. Also, what you do not take into account is that there is a ventral (concious, slow) evaluation of 2D images for depth based on relative movement and relative position (in front of, behind, and parallax requiring a 2nd object in 2D space). This is, I imagine, much slower than depth evaluation of objects in 3 space dimensions, as you have a 3 points of reference on a single object, and therefore do not require a reference object to extrapolate depth. Just like extrapolating the motion of 3 particle systems by recording the interactions of a 2 particle system and using an extrapolating calculation to derive the 3rd is slower than recording the interaction of 3 particles.

Hopefully this answered your request. Just wanted to say, you absolutely have a point. However, you should have framed this as a discussion of the viability of competitive 3D gaming, and not in the frame of "Nvidia is wasting your time".

"that this most excellent canopy, the air, look you! this brave o'erhanging firmament, this majestic roof fretted with golden fire, why it is no other thing to me than a foul and pestilent congregation of vapours."

that's you and 3D

 

[spigzone]

lol, its not hard to understand, sit still and look round your room with one eye.
If you dont move, you see a flat plane.

I stand corrected. I now comprehend what you said, but:

In context ... "Each eye perceives a flat plane in the same way that a screen provides in 3D mode (as long as you dont move much), so your arguments dont make sense."

You have not provided a reason WHY each eye perceiving a flat plane in the same way a screen provides in 3D refutes my point of why lag occurs.

 

[AndroidVageta]

To me honestly the whole 3D thing is just a fad. And not a very practical one at that...

In my eyes 3D movies work so well because the film is created specifically for 3D, meaning that some things will pop out more and other things will not. In 3D gaming everything is the same depth and such (because its the same frame rendered twice) making the 3D more a visual add-on then an actual "virtual reality" type 3D that is experienced in movies. In games the character right in front of your face has the same 3D properties as a character 100 feet in front of you, at least that what it looks like in my eyes when using 3D from nVidia (tried it at Frys).

Plus to me a game is more realistic based on the graphics and physics then being "3D". I'd rather a game have ultra realistic physics, animations, textures, etc then it being 3D. To me this makes a game more involving and realistic than anything 3D could do.

This is all just my opinion really, and I couldnt quite word it as well as it could be, but I think my point is made to those that understand it.

Basically...drop this 3D stuff and focus on what actually matters in a game, such as physics and more complex/realistic animations and movement as well as overall graphics...idk...

 

[Nenu]

I stand corrected. I now comprehend what you said, but:

In context ... "Each eye perceives a flat plane in the same way that a screen provides in 3D mode (as long as you dont move much), so your arguments dont make sense."

You have not provided a reason WHY each eye perceiving a flat plane in the same way a screen provides in 3D refutes my point of why lag occurs.

Thats because its not related to lag, you made that up.