Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post Reply
User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

Hi guys,

We are all Stereo3D fans here, and we all want to get the best out of our setups. Unfortunately, a shared nemesis has always been the performance cost we needed to pay.

In this thread, I hope we can come together here to discuss how to decrease performance penalties. I think this will become more important as the next gen consoles are released with powerful 8c16t Zen cores, while we may well be left behind due to the '3D Vision 3core CPU bug'. https://www.nvidia.com/en-us/geforce/fo ... ation-thr/

Even in the past years, we have had games scaling past 4, 6, and even 8 cores such with Horizon Zero Dawn etc.
Image

Even our old friend GTA5 will scale decently up to 10c20T, and likely beyond:
https://youtu.be/c_Gcg-tFfu0?t=435

So let's begin: to start off, I did some preliminary benchmarks with the only modern game I currently have installed on my system: We Happy Few, so please take with a grain of salt - these are my personal opinions in this specific scenario.

System:
2080 Ti OC
7700k @ 5.1GHz all core.
32GB 4000MHz CL16 memory
Optane XPoint SSD + RAM disk
Full specs in Sig at bottom.


2D = 248 fps of 2D perfromance
Image

3DVision = 128 fps
Relative performance vs 2D = 128/248*100 = 52% of 2D perfromance
Image

HelixVision = 96 fps (uncapped)
Relative performance vs 2D = 96/248*100 = 39% of 2D perfromance
Image

VorpX = Playable ~55 fps @ GPU sync SAFE (recommended/default), ASW off
-- Not really playable = ~69 fps @ GPU sync FAST (stuttery, not recommended)
Relative performance vs 2D = 55/248*100 = 22% of 2D perfromance
-- Relative performance vs 2D = 69/248*100 = 28% of 2D perfromance
Image

This bar chart shows the comparative difference. Solid VorpX shows recommended default setting. Pattern shows max fps but stuttery/unplayable:
Image


I am disheartened that the effect of S3D is so large, especially in VR, and am very interested in improving things.

-- Bo3b has said that for HelixVision, a performance impact is caused by the background tasks needed for things to be converted to VR. The above calculations show that the cost seems to be ~11% in this particular scenario.

-- After Ralf banned me for "2 weeks" from his VorpX forums for daring to point out his toxicity; as many predicted, he has now extended that 2 week ban to permanent status after I told him I wanted to discuss performance and how to improve it, with other VorpX users - apparently he will now only allow posts to technical support, and only after he has screened and approved what I have to say in his 'irregularly checked moderating queue' :roll: :lol:

Indeed, this is an interesting but not unexpected turn of events, after the last time he threatened legal action against me when I reported performance figures:

"Your lift has been banned" he now exclaims... whatever that means ¯\_(ツ)_/¯

Be that as it may, what we really care about is getting better performance, over such banality - and towards that end, there are plenty of friendly places to discuss such things openly without censorship. Since I can't discuss how to improve performance on the VorpX forums, as there is a large crossover between S3D and VorpX users, perhaps we can glean some insights here.

Performance improvements:

HelixVision: :woot
-- Bo3b recommends to exit the SteamVR window before entering game to free up resources.
-- Uncap the newly installed 45fps cap in Fix Manager settings for the game.
-- Use a custom resolution of an aspect ration which better fills the HMD FOV. For example for the Rift CV1 which has a circular viewpoer, I use an almost 1:1 aspect ratio of 1600x1400. This can be accomplished by CRU.
-- Use sharpening to improve image quality

VorpX: :3D
-- Set game VSync to OFF if not already set
-- Disable Asynchronous Warp / reprojection to uncap the 45fps frame cap. Disabling reprojection is tricky with the latest beta, but still do-able.
-- Disable fluid sync
-- Use GPU sync as FAST if playable, or SAFE.
-- Use sharpening to improve image quality
-- Use custom resolution as above.

I would absolutely love to see 3DV/HelixVision/VorpX perform better - everybody would win. Again, maybe we could all learn from each other - at least that is a hope of mine.

All suggestions welcome, including Ralf's, who shall be undoubtedly making an appearance shortly ;-)

Edit: added links etc.
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

WZZ
One Eyed Hopeful
Posts: 12
Joined: Wed Aug 05, 2020 3:04 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by WZZ »

"exit the SteamVR window before entering game"
is this means quiting steamVR just after the steamvr launched?

i wonder if decrease desktop mirror resolution can improve fps?

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

@RD: There is nothing left to say to someone unscrupulous enough to even spindoctor an olive branch into an action against him. You probably believe you are good at rethorics. As far as I'm concerned, I won't jump through your hoops anymore. Sorry. I neither have the time nor any interest in continuing these childish quarrels.

Have a nice life, I wish you all the best (really). Relax a bit occasionally if you can.

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

@WZZ, I don't know mate - I have an Oculus Rift so can't comment how this would affect other headsets. I would take Bo3b's statement as true at face value though I have read multiple times that the screen mirror doesn't take performance away from VR. Something is probably different in regards to HelixVision however.

@Ralf: It's difficult to reply to your posts because you keep editing them. Your earlier post was nicer but since then they have progressively become edited worse, however, thanks for the chuckles ;-)

Indeed, your "olive branches" of banning me, threatening legal action, and attempted censoring of my performance findings - twice, are not needed. If you have any technical insights to add, we are all ears - increasing VorpX performance is in all our best interest.
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

User avatar
helifax
Sharp Eyed Eagle!
Posts: 396
Joined: Tue Jan 12, 2010 5:09 pm
Which stereoscopic 3D solution do you primarily use?: Head Mounted Display (HMD)

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by helifax »

One thing to add here is that, in SteamVR, you can never push hundreds of frames. VR simply works differently than 2D.
There is greater emphasis put on delivering frames at EVEN pacing and not how many frames you can push. As a result it will always try to get to the 90/100/120 Hz of your display and not above. Your GPU will not constantly run at 90+%. If it detects that you cannot maintain stable 90 FPS it will then drop to 50% of your Hz and add re-projection on top.
I believe you already know all of this, but I wanted to say, that you cannot really compare 2D/3D with VR because of the way they work :)

If I made any mistake, please correct me :)

Cheers!
http://3dsurroundgaming.com:
- Home of OGL3DVision wrapper & Vk3DVision - Play your favourite OpenGL & Vulkan games in Stereoscopic 3D using Nvidia 3D Vision OR Virtual Reality using HelixVision(https://store.steampowered.com/app/1127310/HelixVision/)
- Home of some of my UHD Surround/Eyefinity/21:9/32:9 Fixes. (Or you can always check http://pcgamingwiki.com/)

Want to contribute to the development of Vk3DVision? You can check my Patreon Page for the project: https://www.patreon.com/Vk3DVision. If you still like my project but don't want to contribute monthly you can always send me a PayPal: tavyhomeppal@hotmail.com.

You can always follow me on Twitter: @OctavianVasilov

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

@RD: Predictably as always you left out everything in your original post that doesn't fit what you want to bring across. I specifically told you that I don't trust you anymore since our very first exchange, but that with a little bit of goodwill from both sides the restrictions on your account don't have to stay forever. What are you doing with that? Making a fuss as always, mysteriously leaving out everything that doesn't support your point... Precisely the god awful spindoctoring that made me distrust you in the first place. For me that just verified the decision.

Believe that or not: I really wish you all the best and that you can relax a little occsionally.

I actually edited the post since I feared the original version would likely trigger you again. I think the main bit I removed was: 'BTW: You'll always have a place in my heart for the funniest oxymoron ever: 'giving a lesson in humiltity' will be hard to beat.'. If you consider that nicer, fine, I didn't.

User avatar
DJ-RK
Binocular Vision CONFIRMED!
Posts: 268
Joined: Thu Sep 19, 2019 8:13 pm

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by DJ-RK »

I love you guys... mainly because you help me feel like I'm not the *only* person to be an a-hole around here, sometimes (just the biggest!). :lol:

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

I sincerely and honestly apologize for wasting anyone's time with this childish nonsense in the first place. Hope it won't happen again.

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

Since it would probably end up here anyway in a carefully edited fashion, this time I'll rather post a mail to RageDemon myself:

Sorry again for wasting anyone's time.

---

Hello [removed for privacy],

Final message you will receive from me.

I'm not sure what made you feel the need to transform a discussion about a feature in vorpX into something personal posted online a year ago or so and several times afterwards. What I do know however is that I won't waste more time on this.

If you want another reason to make a fuss: you are now indeed banned indefinitely. Just ask yourself whether you would let someone into your house, who - intentionally or not - caused a mess each time you dealt with them. I also won't reply to any of your inquieries anymore which would likely just end in another online outburst somewhere if I'm not careful enough to give you precisely the answer you want to hear.

I'm more than sorry that it had to come this. As always you will probably consider yourself absolutely innocent, so any further discussion of the matter is pointless.

Kind regards,
[removed for privacy]

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

@ Ralf: I Am Jack's Complete Lack of Surprise :roll:
"Grovel and never dare show my own foolishness up again, and I might consider un-banning you from my precious forum sometime in the future if and when I feel like it" - yeah; thanks but I'll pass on this amazing opportunity of a lifetime :lol:

Forget me Ralf; just focus on the technical aspects. We can all work together - people don't have to like each other to work well together. The info you provide would be beneficial to everyone including yourself, and you might learn from others too - there are extremely knowledgeable people on these forums who I am constantly humbled by.

Anyway, moving on:

@Helifax, (and everyone): absolutely right mate. Both HelixVision and VorpX find it difficult to maintain 90fps so they both default to 45 - HelixVision permanently, but VorpX goes to 90fps when it can, which is rare in a modern game.

To unlock the 45fps caps:

We need to disable Asynchronous Warp in Oculus software (Oculus Tray Tool), or Reprojection in SteamVR (Shift-A on the mirror screen which needs to be done each time).

-- For HelixVision, this is a recent change via nVidia's frame capping to 45fps by default. However, disabling the 45fps frame cap will indeed get you ~100% GPU utilisation and over 90fps on the screen mirror (96fps as photographed above). This would be absolutely mandatory for anyone with a powerful system, lest one wants to always be stuck at 45fps.

I don't see any tearing inside the HMD so I think internally the HMD is limiting to 90fps. Of course, the actual limit ought to be the limit of your HMD e.g. 90 in my case. Bo3b recommends allowing a little spare GPU power for processing of additional tasks for best performance so I would say that if a powerful GPU is running ~95% to generate capped 90fps, it would be ideal.

-- For VorpX, there are additional steps:
From memory:

Fluid Sync: I think this might be VorpX's version of reprojection. Enabling this limits FPS to 45 if it runs below 90 - If I recall correctly; it has been a while. Maybe Ralf can tell us more...

GPU sync: This syncs the GPU to VorpX's display. It has 3 options: Fast / Safe / Safest - default.

- "Fast" gives the highest FPS and the GPU is about 100%. Unfortunately, this introduces stutters - 69fps in this case.

- "Safe" gives lower FPs and lower GPU usage however there are no noticeable stutters - this is playable at about 55fps in the above scenario. Then there is the

- "Safest" option, which decreases the fps even further - this is the default option - it gave around 50fps if I recall but I thought it unfair to include that in the benchmarks as it unfairly hamstrung VorpX when the "Safe" option worked well enough.

From initial impressions, HelixVision seems to handle sync better than VorpX - I'm not a coder, I don't know. I was hoping people could discuss such aspects and implement the best solutions. HelixVision and VorpX would do well to be collaborators rather than competitors (or both); but I can't speak for the devs - that's all up to them.

What VorpX does really well is the 6 degrees of freedom head tracking hooked into the game - that is top notch innovation.

==============

On a related note - one thing which often comes up is people erroneously believe that VorpX is VR whereas HelixVision (and maybe 3D Vision in general) is a virtual screen. They tell me that they can't be compared. Well - I can say that in my experience this is absolutely not the case:

Despite all the HooHaa about VorpX's VR mode - all it is, is a virtual screen which has been pulled close to the face. HelixVision does seemingly exactly the same thing if you move it close using motion controls. The only difference is that HelixVision doesn't optionally glue the screen to the face (yet) and enable Head Tracking.

In fact, VorpX over stretches the screen close to the face to fake the VR feel - so much so that the screen becomes blurry and low resolution because a lot of the processed pixels are now off-screen. The first thing I always do is move the screen back until I begin to see borders on the sides of the so called "VR" virtual screen to ensure I am seeing the full processed image area with all the granularity that comes with it.

This is a big reason why setting close to a 1:1 resolution is crucial for getting the best out of VorpX's "VR" mode - the screen ratios and aspect ratios available for VorpX are only the ones available for your monitor/projector/main display. For my 720p projector, this means that VorpX only allows 720p resolutions in game at a 16:9 aspect ratio.

It is easy to get a custom resolution going which is much higher, but careful note must be taken to only use an aspect ratio which your HMD displays - e.g. close to 1:1, or 4:3 if that fails. This is important because the pixels you don't see are still rendered - you are losing all that GPU processing power calculating the extra pixels on the wide 16:9/16:10 ratios but never displayed, which could be better utilised increasing the pixel resolution and HMD FOV in a 1:1 ratio.

--------------
Example: You set the game to display 1920x1080 = 2,073,600 pixels.
On an (example) Oculus rift with a circular (about ~1:1) FOV, VorpX's VR mode will maybe show you 1080x1080 = 1,166,400 pixels.
It's actually worse than that because the rounded corners are also cut and the screen isn't centred to the HMD FOV - there is less space at the top than the bottom. Best case, it will show you maybe ~1 million pixels of the 2 million that your GPU is rendering.
--------------

However, if you set a custom resolution to 1400x1400, then you get the full pixel resolution of the set resolution in the HMD at the same performance cost as the 1080x1080 image above. You end up getting double the performance for 'free' if I recall - full 2 million pixels.

Alternatively, you can simply set a manual custom resolution of 1080x1080 to only have the GPU process 1 million pixels instead of 2 million to give you double the performance.

This is important for both HelixVision and VorpX; but you can always move your head around HelixVision's the wide virtual screen so all pixels are not entirely wasted; - However, this is especially important for VorpX because the VR screen is glued to the face - you will never see those extra pixels with default wide resolutions unless you maybe zooming out and lose the HMD's FOV real estate.



Summing up my experience - when reprojection/ASW are disabled and the artificial frame caps are disabled and the GPUs are working close to 100% giving me floating fps from 75 - 90 solid, I find it far more immersive than 45 locked on both HelixVision and VorpX. I am sure I can't be the only one as the difference is night and day to me.

For 2D gamers, 60fps on PC is a minimum to aim for. 45fps to me is unacceptable especially in something in relation to VR - there is a reason most HMDs aim for 90Hz minimum - anything lower the manufacturers found to be unacceptable. Certainly I have been stuck at 60fps Stereo3D for the last 2 decades. Seeing HelixVision at 90 fps for the first time was quite stunning - Kudos to the developers.

I don't know how much of the above details are absolutely correct - they are my current understanding - I created this thread so that we can discuss and share information from all viewpoints - I want to learn from everyone and maybe some can pick up the odd thing from me. The important thing is that we discuss, and...

https://www.youtube.com/watch?v=du_iG7Veupc

:) :woot
DJ-RK wrote: Mon Aug 10, 2020 8:06 am I love you guys... mainly because you help me feel like I'm not the *only* person to be an a-hole around here, sometimes (just the biggest!). :lol:
To be fair, I have never witnessed you being an a-hole. I think being an a-hole needs to have an air of being unreasonable about it - sometimes your words are blunt, but never unreasonable :lol:
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

User avatar
DJ-RK
Binocular Vision CONFIRMED!
Posts: 268
Joined: Thu Sep 19, 2019 8:13 pm

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by DJ-RK »

RAGEdemon wrote: Mon Aug 10, 2020 1:53 pmTo be fair, I have never witnessed you being an a-hole. I think being an a-hole needs to have an air of being unreasonable about it - sometimes your words are blunt, but never unreasonable :lol:
I appreciate that. I always felt that if I backed up my bluntness with sound arguments that even if I was wrong that at least people could understand my point of view (even if it would make simple statements turn into massive paragraphs and posts), and glad that wasn't lost on everyone. Just seemingly being the only outspoken person most of the time can (and should) cause a person to question whether, "Is everyone else the a-hole, or is it just me?" and I'm at least willing to entertain the idea that since the common denominator is me that it's at least possible I can be a bit of one at times. It's either that and/or the fact that this community was just TOO good (a bit goodie two-shoes'ish actually at times, at least for me, personally), and nothing that good ever lasts.

Truth be told, I never ACTUALLY thought anyone else was being one here (or elsewhere) either, just thought a little self-depreciating humour would be fun and welcome to lighten the mood.

Also, "to be fair," I don't really understand how or why more than one individual has taken serious issue with your postings and/or character. At least from what I've seen, to me you've always seemed sensible, courteous & respectful in your debates (though perhaps a bit eager to air "dirty laundry" publicly, which some can find distasteful), always willing to try to come to a mutual understanding/turn the other cheek, and ultimately wanting what's best for the community and ACTUALLY PURSUING that through both inquiry and testing/statistics... yet here we are! :lol: Then again, a lot of your arguments use fancy terms like "ad hominem" and the like, and seem to be battles of what's read between the lines (to which I'm oblivious to), so I've just chalked all of that up as being how people probably more educated than I go to war! :lol: :lol:

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

There are a bunch of inaccuracies/mistakes in RD's assessment, I'll mainly address the most important one to keep the post short:

With fully supported profiles vorpX's FullVR mode is a lot more than 'a screen pulled close to the face'. In more than 200 games vorpX can automatically handle field of view (absolutely crucial for Full VR), 1:1 headtracking by directly accessing a game's memory, provide VR controller support and also automatically set the resolution of games (after adding custom resolutions to the graphics driver).

Most of the above has to be done by game, imagine it as additional work on top of what you know from creating stereo profiles.

Apart from the above there are a lot of VR specific features in vorpX that make it way more than a S3D screen in VR. Don't want to transform this into an ad though. If an S3D screen in VR is what matters most to you, you probably won't need vorpX.

On a sidenote: The heydays of VR conversion tools are over since several years already. Eventually they, including me (some will love to hear that), will fade away either because there is enough high quality VR content or, if the current Alyx hype proves to be insustainable, because VR will eventually take the unfortunate road of S3D into enthusiast only land.

User avatar
skyrimer
Binocular Vision CONFIRMED!
Posts: 327
Joined: Sun Sep 22, 2013 11:43 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by skyrimer »

Yeah performance is the biggest issue with Vorpx, i have a pretty solid rig, but you can get in trouble to run at playable fps on geo3d even in very old games. However at least on first person games, i'll play them in VR over 3d vision any day, the extra immersion of Vorpx can't be matched by 3d vision. I see a ton of interesting performance tips here, so thanks for those. I have to say though that fake 3d on Vorpx is amazing, Kingdom Come deliverance in fake 3d tricked me for a few moments that it was real 3d, and the performance gains were huge, so kudos for that.

About the drama, I don't know, looks to me that Ralf has huge trust issues, and always sees the bad side of whatever you say, it happened to me too in Vorpx forums, it's a bit sad that whatever you say, he always seems to understand you want to take advantage of him or you're messing with him, which many times it's not true, at least on my case. On the good side, he's improved a lot, in the early Vorpx days it was a lot harder to interact with him. We love you Ralf, i'm extremely happy that Vorpx exists, how amazing your hard work is with it and that you have found a working business model for a 3d/VR injection app, something that 3d vision is suffering of and we're going through difficult times here, I sincerely hope Vorpx keeps being successful, hopefuly Quest being so succesful and Vorpx compatible have helped your sales.
ASUS PG278QR Monitor
Windows 10 1809
AMD Ryzen 5 3600 @3.60ghz
Kingston HyperX Predator DDR4 3200 PC4-25600 16GB 2x8GB CL16
MSI B450 Gaming Plus MAX
Corsair RM650 650W 80 Plus Gold
SSD M.2 2280 500GB
NVIDIA 1080 Ti MSI AERO
VR: Pimax 8k & Index controllers

Visit http://www.gamermovil.com for a ton of mobile news and more!

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

@ Skyrimer: The case with RageDemon is something pretty special, I guess we are just not compatible. Nuff said.

The thing you have in mind is indeed a point where I have to admit I'm really eccentric, don't take it too seriously. I thought I had explained that. Wasn't about me, but about volunteer profile creators. I understand that encouraging and praising volunteers usually comes from a good place, I just see that doing so also puts a certain kind of subliminal pressure on people who are sacrificing their free time largely for others. I just feel uneasy to see too much of this pressure put on people who are already doing way more for free than most would ever do. Just as your desire to encourage and praise volunteers, my scepticism in that regard also comes from a good place.

User avatar
masterotaku
Cross Eyed!
Posts: 114
Joined: Sun Sep 22, 2019 4:43 am
Which stereoscopic 3D solution do you primarily use?: S-3D desktop monitor

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by masterotaku »

I'm an expert conflict avoider, so I've never been personally involved in stuff like this in the years I've been around the 3D Vision scene. Most of the time these things happen over misunderstandings.
I've talked to both sides on many instances (other users, not this thread's case), and it usually comes down to someone (or both) feeling insulted/underestimated when nothing real ever happened. The real insults can come after the misunderstanding happens, though. And then sometimes one person leaves for good...

Both HelixVision and VorpX are cool and have their pros and cons. That's normal and they have different goals. No need to pick sides.

However, I still don't have any VR headset to judge them personally, so my only experience is reading about them :p. I may end up making shader fixes for VorpX too, who knows (years in the future).

User avatar
Losti
Diamond Eyed Freakazoid!
Posts: 732
Joined: Tue Sep 17, 2019 5:30 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by Losti »

Yeah Masterotaku is the chilleds Person in the Web i know :-) die sure. I dont understand how you do this. Great Respekt. Btw. I like the comparison from the first Post!

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

ThinkVR wrote: Tue Aug 11, 2020 3:26 am RageDemon is something pretty special.
Speaking of taking things out of context, I thought this extract would be pretty comical ;)

Your self professed eccentricity might explain your behaviour Ralf, but it does little to excuse it - as has been abundantly apparent from a long list of people who have had less than constructive interactions with you - Even for people who like to stay quiet on open forums, your actions do not go unnoticed - I am glad to hear at least, that you're getting better...

Keep in mind that we are not your enemies - we all want you to do well, in spite of everything.

Back on topic:
Aside from the head tracking innovations in VorpX, would you say something insightful about improving performance?

For example:
-- Can you speculate why HelixVision might seem to sync better, with ~100% GPU utilisation in comparison to VorpX's safe/safest modes?

-- More to the point, can you speculate how you/we might be able to use this to improve VorpX performance where perhaps under safe/safest scenarios it can take advantage of more GPU power or modify "Fast" sync so that it doesn't cause artifacting? If the remedy involves causing tearing, some like me might live with that...

-- Do you plan to make changes to performance in the near future?


@DJ-RK, you're too kind mate. Most people here regard you as a sage, rather than anything remotely negative.

You bring up some fascinating points - for a long time, I too took the niceness of this community for granted - I thought most places would be like this. But now, when I venture into other communities on places such as Reddit, or VR forums in general, I see them filled with toxicity and friction, even trolling. I continue to be grateful that we are not one of those...

We could go into detail about your other pondering regarding present company, but I think that would derail this thread into something less constructive :D
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

You started this, noone else. We had a discussion over e-mail about a vorpX feature, for whatever reason you felt the need to post parts of that online, carefully edited in your favor, a repeated pattern ever since. Now, of course, you will say, all of that was my fault. And in a way you are right that it was my fault. My fault was to ever reply to you in the first place. Find someone else for your childish drama queen games. I'm getting too old for ridiculous internet fights over nothing.

Edit: Also I seem to remember that in this very forum you not too long ago had just another fight with the HelixVision head programmer. His assessment of your, let's call it discussion style, sounded all too familiar to me at the time. I guess he also didn't behave right in some way from your perspective.
Last edited by ThinkVR on Tue Aug 11, 2020 4:08 pm, edited 1 time in total.

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

ThinkVR wrote: Tue Aug 11, 2020 3:51 pm You started this, noone else. We had a discussion over e-mail about a vorpX feature, for whatever reason you felt the need to post parts of that online, carefully edited in your favor, a repeated pattern ever since. Now, of course, you will say, all of that was my fault. And in a way you are right that it was my fault. My fault was to ever reply to you in the first place. Find someone else for your childish drama queen games. I'm getting too old for ridiculous internet fights over nothing.
Quoted before it starts being edited again... :roll:

========= Added.

Now now Ralf; Nicely played - It would be foolish to be ensnared in such an obvious trap to derail this thread, as you so desperately desire.

It's unfortunate that you insist on making this thread personal - it contains good info and some valuable insights from/for everyone. You would have done well for everyone and VorpX by contributing knowledge.

Let's be amicable and only use references to your quotes to set records straight... ;)

1. In that heinous post you are constantly obsessed with, you will note that I gave you and VorpX a glowing review. "Hats off to Ralf and his software" is what I concluded.

2. The slightly negative part of that post was me pasting the nasty emails that you sent me where you contradict yourself - damn man, you kept confusing Convergence and Depth - at least you went on to admit you were wrong in the end.

3. Your emails were pasted pretty much word for word with all of your nastiness - and I do state clearly that I cut out some fluff to keep things readably short for a single forum post.

-- Ralf, you're only bitter because you were exposed - I quoted your nastiness word for word, and you felt humiliated for the truth being exposed; - for what you had done. This might be news to you but in society, Maybe don't be nasty if you don't want people to know you're nasty? :roll:

Missing the steps where you threatened legal action against me for speaking my personal experience with VorpX performance...

4. Later, you were nasty on your forum. I quoted you again, word for word - with hard proof on my linked thread. You felt humiliated again and quickly edited your post and deleted mine under false pretext, and then banned me in an attempt to not be exposed and humiliated again... :roll:

5. Now you say you have double banned me with cherry on top ¯\_(ツ)_/¯

You are your own worst enemy Ralf.

You feeling humiliated by someone quoting your own messages is your own undoing I'm afraid - perhaps you might consider sending yourself a nasty email threatening legal action? ;) :lol:

If you want to contribute knowledge to this thread, then please do so. Otherwise, it is often wise to stay silent ;)

Think of it from your own interest: When someone googles VorpX performance, do you really want people to see nastiness from you, or your insights into how performance may be improved? One path leads to further loss of respect and sales. The other path leads to more respect and increased sales.

The choice is yours... I hope you make the right one; I am rooting for you :)
Last edited by RAGEdemon on Tue Aug 11, 2020 5:16 pm, edited 3 times in total.
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

Don't forget the edit, just for you I marked it:

Edit: Also I seem to remember that in this very forum you not too long ago had just another fight with the HelixVision head programmer. His assessment of your, let's call it discussion style, sounded all too familiar to me at the time. I guess he also didn't behave right in some way from your perspective.

Since I'd like to let helifax's wise words ideally be the closing of this dramedy, I'll add my reply to your additions here:

You know, we have one thing in common, sometimes not knowing when to stop (just like now, both of us). There is a major difference though: you are the one who for reasons only known to yourself feels the need to drag personal arguments into public. I have zero clue what you think you can gain from that, and honestly I don't care. You seriously post private conversation online, edited in your favor, and then tell the other party it's their fault? Come on man, you're to smart to believe that yourself. You know exactly what you do. Just the why stays a mystery, maybe even for you.

I think I wasted another hour on this nonsense now today. It was the last one. Do what you think you have to do. Just do it without me.
Last edited by ThinkVR on Tue Aug 11, 2020 5:33 pm, edited 1 time in total.

User avatar
helifax
Sharp Eyed Eagle!
Posts: 396
Joined: Tue Jan 12, 2010 5:09 pm
Which stereoscopic 3D solution do you primarily use?: Head Mounted Display (HMD)

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by helifax »

Come on guys.. Really?! Do we constantly need to "fight over air"?
I mean I see both sides "hold grudges" for whatever reason! But "COME ON"! What is the benefit?
Sure! I'm an imbecile for fighting with DHR a few years back! I'm the imbecile for fighting with RAGEDemon some times ago as well :D And so on!
But the bottom line is... none of this helps with anything... If it does... please let me know so I can be move to a different planet ^_^.

Now, seriously! Both Ralf and Ragedemon, can't we all be friends?! :!:
http://3dsurroundgaming.com:
- Home of OGL3DVision wrapper & Vk3DVision - Play your favourite OpenGL & Vulkan games in Stereoscopic 3D using Nvidia 3D Vision OR Virtual Reality using HelixVision(https://store.steampowered.com/app/1127310/HelixVision/)
- Home of some of my UHD Surround/Eyefinity/21:9/32:9 Fixes. (Or you can always check http://pcgamingwiki.com/)

Want to contribute to the development of Vk3DVision? You can check my Patreon Page for the project: https://www.patreon.com/Vk3DVision. If you still like my project but don't want to contribute monthly you can always send me a PayPal: tavyhomeppal@hotmail.com.

You can always follow me on Twitter: @OctavianVasilov

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

@ Helifax, I would/have done many times mate; and being Ralf, he would triple ban me tomorrow and start complaining again ;-)

Back on Topic:
I did some researching on other 3D drivers:
This is ATi's driver with DDD/Tridef vs 3D Vision:

Image

Interestingly, 3D Vision driver seems to be far more refined, giving substantially higher comparative performance compared to 2D.

Bo3b says that if the 30XX series due to be previewed next month outright do not support 3D Vision, then he will start work on a 3D Vision-esque driver himself.
He has asked already nVidia for direct access to the driver to enable high performance, but unfortunately they have told him to "pound sand".

I have a suspicion that 3D Vision is comparatively high performing because it has secret direct access to the nVidia driver - I am not a coder so I don't know what else to call it - API?

If a person has to write a driver on their own, e.g. Tridef/iZ3D/Bo3bVision - performance might not be on par with 3D Vision even with its core bottleneck...

Thoughts?

Helifax, could it be at all possible to make Vk3DVision-like 3D driver without the need to use 3D Vision? Presumably that would be an insane amount of work? :o :)

=====================

Edit: Found a 2010 MTBS3D article comparing 3DVision vs iZ3D vs DDD:

https://www.mtbs3d.com/articles/editori ... dia.html/5

Lots of good graphs with mixed results - here is a sample:

Image
Image

Albeit it's a good read that shows in Resident Evil 5, other drivers were slightly better than 3DV.
Last edited by RAGEdemon on Wed Aug 12, 2020 2:51 am, edited 4 times in total.
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

BazzaLB
One Eyed Hopeful
Posts: 25
Joined: Mon Sep 16, 2019 8:53 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by BazzaLB »

Yes, the risk of not being able to inject 3DVision support into the latest drivers which no doubt will be necessary to support 30XX cards is of some concern. With VR you need all the horsepower you can get and I will probably get a 3080 ti card along with CPU upgrade, but I still have some titles I'd like to play using 3DVision.

User avatar
The_Nephilim
Certif-Eyed!
Posts: 540
Joined: Sat Dec 22, 2007 3:38 am
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup
Location: 3rd Stone from the Sun

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by The_Nephilim »

how to I get rid of the nvidia 45fps cap?? Is all I want to know as LOTRO and Arma 3 dont go beyond 45fps?? I found the framelimiter in Helixvision I put it on unlimited but still get ppor frames in Arma 3 and LOTRO..

I do not have Vsync or reprojection on.. I use a 9700K @ 4.8GHZ and a 1080GTX with 445 drivers with a samsung Odyssey Plus while using Helixvision..
Intel i7 9700K @ 4.8GHZ / MSI MPG Z390 Motherboard / G. SKILL 32GB 3200MHZ Ram / Asus 1080GTX / SoundBlaster Z / Samsung Odyssey + VR Headset / HOTAS Cougar / Thrustmaster MFD's

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

I don't have Odyssey, or another SteamVR headset so I can only speculate: Are you sure you don't have reprojection on? It's on by default and is pretty tricky to get rid of because the option has been hidden in the latest SteamVR. How did you disable it?

How are you measuring fps?

what fps do you get on your 3D Vision monitor?

What is your GPU usage when you are in game?

I notice that your signature says you have a GTX 660. The minimum for even old Generation 1 VR is a GTX 970. According to benchmarks, your card is ~46% the power of a GTX 970. I think that is the problem. In your specific case, I don't think it would be wise to turn reprojection off, nor lift the 45fps cap unfortunately.
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

As a little sign of goodwill and apology to everyone for this mess, a little trade secret (not really) as a productive addition to this topic, after all at some point there will have to be a successor to vorpX. Not yet, but in a few years for sure.

There are two major things 3D-Vision does not handle right for VR at all by design, both may be elusive to many of you since you are trained to handle stereoscopic 3D way above natural levels.

1. Stereo 3D for VR is not just shifting vertices to the left and the right, but also up and down since head roll has to be factored in. Head roll not being considered is one of the major causes for discomfort in traditional stereo 3D, e.g. in movies. In VR this is no problem anymore, native apps handle that right automatically, a 3D-Driver for VR would also have to consider it. edit: it's actually bit more complex than described above, catchphrase: head and neck model.

2. 3D-Vision has no concept of IPD (interpupillary distance). You can and often will encounter situations where you have parts of the scene being farther apart than your eyes, making the eyes diverge when you look at these parts of the scene. This never occurs in normal life, and it also should never occur in VR since it also causes eye strain and discomfort.

Again, both points are likely elusive to trained S3D enthusiasts, and as such may not be considered important for many of you, they are however of utmost importance for providing a comfortable VR experience.

User avatar
helifax
Sharp Eyed Eagle!
Posts: 396
Joined: Tue Jan 12, 2010 5:09 pm
Which stereoscopic 3D solution do you primarily use?: Head Mounted Display (HMD)

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by helifax »

RAGEdemon wrote: Tue Aug 11, 2020 5:39 pm
Helifax, could it be at all possible to make Vk3DVision-like 3D driver without the need to use 3D Vision? Presumably that would be an insane amount of work? :o :)
That is the plan :)
3D Vision for Vk3DVision is only used for displaying. That's it. I don't use any of the Automatic stuff. I've already done this in the OGLWrapper where you can use it with 3D TVs without even having 3D Vision in the first place.
It's actually not that hard :) Sure there is more coding and time involved, but that's it.
http://3dsurroundgaming.com:
- Home of OGL3DVision wrapper & Vk3DVision - Play your favourite OpenGL & Vulkan games in Stereoscopic 3D using Nvidia 3D Vision OR Virtual Reality using HelixVision(https://store.steampowered.com/app/1127310/HelixVision/)
- Home of some of my UHD Surround/Eyefinity/21:9/32:9 Fixes. (Or you can always check http://pcgamingwiki.com/)

Want to contribute to the development of Vk3DVision? You can check my Patreon Page for the project: https://www.patreon.com/Vk3DVision. If you still like my project but don't want to contribute monthly you can always send me a PayPal: tavyhomeppal@hotmail.com.

You can always follow me on Twitter: @OctavianVasilov

User avatar
The_Nephilim
Certif-Eyed!
Posts: 540
Joined: Sat Dec 22, 2007 3:38 am
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup
Location: 3rd Stone from the Sun

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by The_Nephilim »

RAGEdemon wrote: Wed Aug 12, 2020 2:26 am I don't have Odyssey, or another SteamVR headset so I can only speculate: Are you sure you don't have reprojection on? Yes I am sure it is off.



It's on by default and is pretty tricky to get rid of because the option has been hidden in the latest SteamVR. How did you disable it? I disabled it in the VR Settings file "motionReprojectionMode" : "disabled",

How are you measuring fps? Fraps

what fps do you get on your 3D Vision monitor? I dont have a 3d monitor. I ran Arma 3 in anaglpyh and was getting 80-85FPS.

What is your GPU usage when you are in game? I dont know Fraps does not show GPU USage. I may need to find another utility any you reccomend?

I notice that your signature says you have a GTX 660. The minimum for even old Generation 1 VR is a GTX 970. According to benchmarks, your card is ~46% the power of a GTX 970. I think that is the problem. In your specific case, I don't think it would be wise to turn reprojection off, nor lift the 45fps cap unfortunately. Umm in the bottom of my post I said what PC I was currently using but I have updated the 8 year old sig with my new specs I am more then capable for VR.
Intel i7 9700K @ 4.8GHZ / MSI MPG Z390 Motherboard / G. SKILL 32GB 3200MHZ Ram / Asus 1080GTX / SoundBlaster Z / Samsung Odyssey + VR Headset / HOTAS Cougar / Thrustmaster MFD's

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

@Helifax, fantastic news mate - bo3b said he has passed along the relevant HelixVision code to you for when you wish to implement it - no pressure! hehe :)

It's great to see that there are paths forward for both nVidia and AMD, even if future generation of cards are completely incompatible with 3DVision.

@ The_Nephilim: I use MSI Afterburner for GPU related statistics. (except FPS).

Have you tried Forcing VSync to OFF from the nVidia control panel?

As of the latest HelixVision, I too have an issue that is caused by (*I think*) HelixVision forcing VSync ON(?) - I could not get past the low refresh rate I had set with my custom resolution even at Unlimited fps even after unticking custom resolution and frame limiter checkboxes under the 3DFM settings tab.

Forcing VSync to OFF in the game profile with nvProfileInspector fixed the issue for me.


@Ralf, thanks for the contribution.

1. Your implementation of S3D even with head roll is cool and innovative - kudos.

2. 3DV locks the IPD for each display to a maximum to prevent eye strain. Everyone's IPD is different, especially kids - so nVidia used a conservative value which many of us were not happy with - a lot of us use a hack to bypass this limit. It's a fundamental adjustment which all experienced (trained) S3D users are familiar with, controllable via hotkey - certainly it's the first adjustment any of us make when starting a game.

I don't believe HelixVision uses a value referencing the HMD's set IPD - yet. VorpX does not allow adjustment to this - people like myself would find this very useful especially for custom profiles which do not always have a good setting set, if you ever have the time to implement an adjustment for advanced users.

I don't completely agree that demand for software such as VorpX/HelixVision/S3D drivers will be due to dissipate any time soon, or be succeeded completely by VR easily. Even though VR will get better, there will always be 2D games for the casual market, which people like me/present forum company will want to play in VR/S3D.

With the general public rejecting 3D every other decade it comes around (or at least rejecting its usually horrid implementation), I personally think there will be a sustainable if not growing market for such software for at least the next few decades.

I notice that VorpX has been updated yesterday to now include a "New" GPU Sync option, and an "Original" option. Preliminary testing shows that "Original" is the previous "Safest" option which gave 50fps as mentioned in OP. The "New" implementation gives 62fps in the same scenario without any stuttering - Kudos.
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

ThinkVR
One Eyed Hopeful
Posts: 26
Joined: Thu Jul 09, 2020 8:45 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by ThinkVR »

1. I see it just as doing things as they have to be done in VR, but thanks anyway. Appreciated.

2. That’s the good thing about a VR headset, everything is precisely defined (provided users dialed in the right headset IPD), making it possible to avoid diverging eyes entirely, which, except for a few lucky ones, is a *really* uncomfortable experience.

As far as the new sync feature is concerned, it can now in many cases fully utilize the GPU without any stutter/judder. There can be situations/games though were falling back to the original method makes sense. So always worth a shot to try both.

User avatar
The_Nephilim
Certif-Eyed!
Posts: 540
Joined: Sat Dec 22, 2007 3:38 am
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup
Location: 3rd Stone from the Sun

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by The_Nephilim »

@RAGEdemon, I set the Vsync in nVidia CP to vysyns off and in 2d I was getting around 70-85FPS. in 3D it weas doen to 60-70 and in VR I am at around 45FPS? I will double check the Reprojection.. I guess I need a beefier video card for HelixVision?

Hopefully I will be able to grab a 3080 when they come out..


EDIT:Earlier In the game in MP I was getting as low as 18FPS :(
Intel i7 9700K @ 4.8GHZ / MSI MPG Z390 Motherboard / G. SKILL 32GB 3200MHZ Ram / Asus 1080GTX / SoundBlaster Z / Samsung Odyssey + VR Headset / HOTAS Cougar / Thrustmaster MFD's

whyme466
Binocular Vision CONFIRMED!
Posts: 232
Joined: Fri Mar 11, 2011 1:16 pm
Which stereoscopic 3D solution do you primarily use?: S-3D HDTV

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by whyme466 »

Some time ago, I suggested that HelixVision might include a “stuck to the face” mode, in order to offer a higher performance option (see https://www.mtbs3d.com/phpbb/viewtopic ... 58#p172158).
bo3b felt HelixVision’s performance loss was lower than RAGEdemon’s testing seems to indicate - assuming same image size used in each test environment. This mode might be considered - for users who value 3D game performance over a VR environment.

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

Some good news for ease of use:
The latest steamVR beta has made available per-application reprojection/motion smoothing settings:
Image

On the Windows side of things:
The latest Windows Updated has also added extra functionality to it's per-application GPU selection.
Image
[image link: reddit]

In theory, what this *might* means is that since pretty much most people have an integrated GPU which they never use (most Intel CPUs), some apps can be offloaded to other GPUs in the system to reduce impact on the main game being run. I don't know how well this might work, or if it might even work at all, but the premise is intriguing: offloading all [ 3D driver? ] overheads to another GPU so that the main GPU can concentrate on the main game only.
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

BazzaLB
One Eyed Hopeful
Posts: 25
Joined: Mon Sep 16, 2019 8:53 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by BazzaLB »

The SteamVR per application settings have been there for donks. BTW, Motion smoothing and reprojection are not the same thing. Different things entirely.

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

Thanks for the input, however I'm not quite sure how relevant it is:

In the context of improving performance, why does it mater for how long the dialogue has been around? Why does it matter whether motion smoothing/reprojection are same or different?

They each drop the fps to 45 and need to be disabled for our testing; the above dialogue presents options for each separately as shown.

Perhaps you might like to suggest other settings which might aid in better performance in HelixVision/VorpX/3DVision?

@The_Nephilim, I believe you have disabled one via settings file editing, but not the other - Let's try the dialog to disable both.

Further, my system has 4 VR settings files under different paths - can you be certain that the correct file was edited and that is the one the system is using?

Image
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

BazzaLB
One Eyed Hopeful
Posts: 25
Joined: Mon Sep 16, 2019 8:53 am

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by BazzaLB »

Because reprojection does not drop the frame rate to 45.. motion smoothing however does. So perhaps it is relevant. But anyway, its your thread so sorry for the intrusion to clear up some misinformation.

User avatar
RAGEdemon
Sharp Eyed Eagle!
Posts: 472
Joined: Thu Mar 01, 2007 1:34 pm
Which stereoscopic 3D solution do you primarily use?: S-3D Projector Setup

Re: Performance tested: 3DVision vs HelixVision vs VorpX vs 2D

Post by RAGEdemon »

Thanks, all input is appreciated - we're trying to ascertain best settings and I'm sure there will be mistakes to iron out before we reach somewhere ideal; an unfortunate disadvantage of using trial and error methods.

Your statement that reprojecton does /not/ drop frame rate to 45 was quite surprising to me.

After some reading, it turns out that there are 2 reprojections being used on SteamVR - Legacy (interleaved) reprojection, which does drop the fps to 45; and new Async Reprojection which does NOT drop the fps to 45. Without your sharing, I would have not known about this, so cheers for that - Non-oculus users would be thankful to you.

One of the problems is that I personally have an Oculus Rift so SteamVR's implementation of both Async Reprojection and Interleaved Reprojection are not used by my headset if I recall correctly. Oculus uses its own Asynchronous Space Warp and Asynchronous Time Warp. Unfortunately this makes it very difficult for me to test settings for people running SteamVR headsets. Interestingly, this *might* also be why Bo3b reports perfectly smooth FPS at 45 with Reprojection enabled on his SteamVR headset while on an Oculus headset, 45fps looks horrid(?)

Would you gents kindly try HelixVision with reprojection in both Async and Interleaved modes? I believe older SteamVR allowed the user to enable both simultaneously or disable both altogether via checkboxes - it seems from the dialogue that now you are forced to enable one or the other (?), but you cannot have both on simultaneously, nor can you disable both simultaneously (?).

Unfortunately, I cannot do any testing on this with my headset...
Edit: Clarity | Windows 10 64-Bit | 7700K @ 5.1GHz | 2080 Ti OC | 32GB 3956MHz CL16 RAM | Optane PCIe SSD | Sound Blaster ZxR | 2x 2000W ButtKicker LFE | nVidia 3D Vision | 3D Projector @ DSR 1600p | Oculus Rift CV1

Post Reply

Return to “NVIDIA GeForce 3D Vision Driver Forums”