Jump to content
chiliwili69

Pimax5K+ vs. Oculus Rift: Spot, ID & Gauges through lens

Recommended Posts

Posted (edited)

Well, finally it has arrived the time to objectively measure how good is the resolution of the Pimax5K+ in IL-2 VR, comparing it to our well known friend: the Rift.

I have already seen some "through the lens" pictures posted on the web, but here I want to make a fair comparison specific for IL-2 VR.

What is fair? well, just use the more practical settings that I will use in the real game for every device. It will not be fair to use SteamVR SS at 100% for both.

For the Rift I am using the IL-2 VR benchmark settings but with 180% SteamVR SS, this is about 7.7 million pixels. My 1080Ti can handle more SS, but it is not worth the extra load.

For the Pimax5K+ I am using the same benchmark settings but with Normal FOV, RQ=1 in PiTool and 66% SteamVR SS, this is about 16.7 million pixels, which are handled relatively well by my 1080Ti in Kuban.

These are the calcs:

1452564190_Pixelscalcs.png.2a70087506e5c4cfcedf19cc1ed9949f.png

 

I tested three aspects of the game: Spot, ID and Gauges. For that, I created 7  short track records (3 for spot, 3 for ID, 1 for gauges).

The tracks are available if anyone is interested to make their own pictures: https://drive.google.com/open?id=11pZiEim7nu2KKz193BAk9G0nD_HKoBrj

(BTW, It would be good to have a single track record containing elements for ID, spot and gauges. A kind of common image pattern for quality comparisons. But this is another matter.)

 

I used a Panasonic Lumix DMC-ZX3 compact camera with 14Mp resolution. I shot about 6 or 7 for each scene and took the best.

All the native images can be found in this link: https://drive.google.com/open?id=1CJp3XFXDTIvcibT470UP7RrT3aqExIOW

It is better to download the images and visualize with 1:1 image to monitor pixel ratio.

Here I upload the compositions in jpg format because size is limited. But you should click on image to see the differences better.

Please, don´t take into account the chromatic aberration in the Rift, it is a strange effect for this camera, but to the human eye there is not that aberration.

 

1. SPOTTING: It is bit better, but just only a little bit. With less SS in the Rift it will improve but then we have less detail in ID or gauges.

The size of the letter of the indicators are bigger in the Pimax.

1656872520_1-Spot-Pimaxvs_Rift.thumb.jpg.4741c586711abf013103a08147d739eb.jpg

 

2. ID: Here also it is a bit better, but not a huge difference. Maybe a more clear picture and better contrast with the Pimax but without additional details.

In the bomber in the bottom you can see the propellers with the Rift, but not with the Pimax.

1701104649_2-ID-Pimaxvs_Rift.thumb.jpg.91b654ef4c39d05bb336eebbfa2cb8f1.jpg

 

3. GAUGES: Here n is a very little bit better the Pimax, but not a huge difference. The text for the warning flaps can be read better in the Rift (this is because is 180%SS).

2057210680_3-Gauges-Pimaxvs_rift.thumb.jpg.83bbe7a6e7322b86eba6a0ff0b400bc7.jpg

 

CONCLUSION:

With the three images above, one would conclude that it is not worth to buy the Pimax (700$ only headset) having already the Rift. This would be true if we only look at the increased resolution. 

But the good thing of the Pimax is not the resolution, it is the increased FOV. And this is a very good thing of the Pimax.

 

It could be surprising that having a display with almost 3 times more pixels (2.6Million vs 7.4Million), the images are only a little bit better. But the reason is that the pimax resolution is spread (or wasted) in a very large FOV,  unnecessarily large. (Low=120º, Normal=150º, Large=170º). I still have not tested the large FOV, but for IL-2 the normal FOV is really more than enough. When the normal or small FOV is used a portion of the display is not used

 

It´s good to have a larger FOV than the Rift, but it is a pitty that all those extra pixels of the Pimax are wasted in rendering details at the periferia. The human vision doesn´t need too much details at the periferia, all the focused detailed vision happen in the 100º infront of my face.

Ideally the supersampling routines should supersample more the center of the view and less the periferia, this will save a lot of GPU load. (I really don´t know if SteamVR do that)

 

EDIT:

I have also added a comparison with the Pimax at 66%, 100%, 180% and the Rift at the same 180%.

The calcs are:

1806369985_ratiostest2.jpg.59f898b2406ace2ff89877afa365aae4.jpg

 

In the Rift, with 180% we render 7.7Million pixels which is 3 times the pixels of the physical display (the visual detail we get from that SS is minimal)

In the Pimax, with 66% we render 16.7Million pixels which is 2.3 times the pixels of the physical display (So I will not call it subsampling really)

with 100% we render 25.3 Million pixels which is 3.4 times the pixels of the physical display (this can not be handled well by my 180Ti, perhaps a 2080Ti or higher)

with 180% we render 45.6 Million pixels which is 6.2 times the pixels of the physical display (there is almost no visual gain from 100%)

 

All the native images of this second test are here: https://drive.google.com/open?id=1MLMfDHctYOIQzgF4wUJTjTofvn9G8f4t

 

ID: There is a gain going from 66% to 100%, not huge but noticeable. But going from 100% to 180% is not worth for ID.

202293302_2-Pimax(66-100-180)vsRift.thumb.jpg.93498a9b1f16679215bae0c865e754a8.jpg

 

GAUGES: Here we also see a small but noticeable gain going to 100%, specially in the text. 

491136320_3-Pimax(66-100-180)vsRift.thumb.jpg.384e4540ac90cb5f8d45d9c4cf6c3526.jpg

 

In my case it only matters to me the 66% column since this is the value which keeps my fps in the 70 to 90 fps.  If in the future Pimax can incorporate eye tracking and foveat rendering then we could have high SS ratios where you are looking to and less around, decreasing the load to the GPU. So, yes, it is interesting to see how it looks like with 180% since this could be the ratios applied to the area you are looking to.

 

 

Edited by chiliwili69
  • Like 2
  • Thanks 4
  • Upvote 6

Share this post


Link to post
Share on other sites

 

 

The Pimax looks enough better than the Rift to me that I would try the Pimax, but my computer probably couldn't handle it. Someday...

13 hours ago, chiliwili69 said:

Please, don´t take into account the chromatic aberration in the Rift, it is a strange effect for this camera, but to the human eye there is not that aberration.

When I tried the Rift, the chromatic aberration was far worse than in your pictures. One of the reasons I returned it.

  • Upvote 1

Share this post


Link to post
Share on other sites
13 hours ago, chiliwili69 said:

 

3. GAUGES: Here n is a very little bit better the Pimax, but not a huge difference. The text for the warning flaps can be read better in the Rift (this is because is 180%SS).

2057210680_3-Gauges-Pimaxvs_rift.thumb.jpg.83bbe7a6e7322b86eba6a0ff0b400bc7.jpg

 

It´s good to have a larger FOV than the Rift, but it is a pitty that all those extra pixels of the Pimax are wasted in rendering details at the periferia. The human vision doesn´t need too much details at the periferia, all the focused detailed vision happen in the 100º infront of my face.

Ideally the supersampling routines should supersample more the center of the view and less the periferia, this will save a lot of GPU load. (I really don´t know if SteamVR do that)

 

 

Thanks for the continuing review/comparison, I personally appreciate it very much!

 

Imho, with 66% SS the undersample rate is extreme. I guess you're doing it for the purpose of high fps, but it doesn't speak truthfully to hmd's capabilities. Can you make these shots using 100% SS?

 

Also, using my rift to vivepro switch as an example, the bump in resolution was extremely noticeable for me, even though on paper it was somewhere between 40 to 60%. My brain&eyes inside the HMD perceive the resolution bump a lot different than looking at the static pictures like these. The screen brightness plays a huge role in all of this as well. For example, I returned the Odyssey+ HMD because it looked like some dark & muddy shit to my brain&eyes despite the screen resolution being on par with Vivepro.

Share this post


Link to post
Share on other sites

Thanks for all the testing. These threads are very useful.

Share this post


Link to post
Share on other sites

Very interesting, but could you vive us the gauges picture with 100% SS ?

Share this post


Link to post
Share on other sites
19 hours ago, chiliwili69 said:

Well, finally it has arrived the time to objectively measure how good is the resolution of the Pimax5K+ in IL-2 VR, comparing it to our well known friend: the Rift.

I have already seen some "through the lens" pictures posted on the web, but here I want to make a fair comparison specific for IL-2 VR.

What is fair? well, just use the more practical settings that I will use in the real game for every device. It will not be fair to use SteamVR SS at 100% for both.

For the Rift I am using the IL-2 VR benchmark settings but with 180% SteamVR SS, this is about 7.7 million pixels. My 1080Ti can handle more SS, but it is not worth the extra load.

For the Pimax5K+ I am using the same benchmark settings but with Normal FOV, RQ=1 in PiTool and 66% SteamVR SS, this is about 16.7 million pixels, which are handled relatively well by my 1080Ti in Kuban.

These are the calcs:

1452564190_Pixelscalcs.png.2a70087506e5c4cfcedf19cc1ed9949f.png

 

I tested three aspects of the game: Spot, ID and Gauges. For that, I created 7  short track records (3 for spot, 3 for ID, 1 for gauges). I can make the tracks available if anyone is interested.

I used a Panasonic Lumix DMC-ZX3 compact camera with 14Mp resolution. I shot about 6 or 7 for each scene and took the best.

All the native images can be found in this link: https://drive.google.com/open?id=1CJp3XFXDTIvcibT470UP7RrT3aqExIOW

It is better to download the images and visualize with 1:1 image to monitor pixel ratio.

Here I upload the compositions in jpg format because size is limited. But you should click on image to see the differences better.

Please, don´t take into account the chromatic aberration in the Rift, it is a strange effect for this camera, but to the human eye there is not that aberration.

 

1. SPOTTING: It is bit better, but just only a little bit. With less SS in the Rift it will improve but then we have less detail in ID or gauges.

The size of the letter of the indicators are bigger in the Pimax.

1656872520_1-Spot-Pimaxvs_Rift.thumb.jpg.4741c586711abf013103a08147d739eb.jpg

 

2. ID: Here also it is a bit better, but not a huge difference. Maybe a more clear picture and better contrast with the Pimax but without additional details.

In the bomber in the bottom you can see the propellers with the Rift, but not with the Pimax.

1701104649_2-ID-Pimaxvs_Rift.thumb.jpg.91b654ef4c39d05bb336eebbfa2cb8f1.jpg

 

3. GAUGES: Here n is a very little bit better the Pimax, but not a huge difference. The text for the warning flaps can be read better in the Rift (this is because is 180%SS).

2057210680_3-Gauges-Pimaxvs_rift.thumb.jpg.83bbe7a6e7322b86eba6a0ff0b400bc7.jpg

 

CONCLUSION:

With the three images above, one would conclude that it is not worth to buy the Pimax (700$ only headset) having already the Rift. This would be true if we only look at the increased resolution. 

But the good thing of the Pimax is not the resolution, it is the increased FOV. And this is a very good thing of the Pimax.

 

It could be surprising that having a display with almost 3 times more pixels (2.6Million vs 7.4Million), the images are only a little bit better. But the reason is that the pimax resolution is spread (or wasted) in a very large FOV,  unnecessarily large. (Low=120º, Normal=150º, Large=170º). I still have not tested the large FOV, but for IL-2 the normal FOV is really more than enough. When the normal or small FOV is used a portion of the display is not used

 

It´s good to have a larger FOV than the Rift, but it is a pitty that all those extra pixels of the Pimax are wasted in rendering details at the periferia. The human vision doesn´t need too much details at the periferia, all the focused detailed vision happen in the 100º infront of my face.

Ideally the supersampling routines should supersample more the center of the view and less the periferia, this will save a lot of GPU load. (I really don´t know if SteamVR do that)

 

 

 

 

 

Your idea about ss in the center only is a great idea!

Share this post


Link to post
Share on other sites
Posted (edited)

SS = 180% on the Rift reduces contrast, since the backend information leads to merging colors/pixels when the picture is downscaled for output in the Rift again.

 

The higher the SS above native, the less contrast, the less blacks. The more planes merge with background. It's only natural and that way in every headset, the higher you go away from the native res and the smaller the object/line/border of regard is. It also seems to increase the feeling of "chromatic abberation" (which it isn't). I still have my old Rift, it doesn't seem to be a lense-problem but just appear like it was CA due to a too high SS setting.

 

TLDR; You may want to test both devices with their native resolutions. 😁👍

Edited by SCG_Fenris_Wolf

Share this post


Link to post
Share on other sites

I'm sorry Chill but while I appreciate the effort your methods are wrong.  You are deliberately hobbling the Pimax 5K with negative supersampling because that's what you play at but that is a purely subjective test.  I used this analogy before and it still holds true, would you compare a 1080p monitor against a 4K monitor and run the 4K at 1080p for "fairness" because FPS at 4K was poor?  Would your conclusion in such a poor test be valid?

 

Your test was specifically aimed at spotting, yet where in this review has Screen Door effect been mentioned?

 

I have a Pimax 5K+ and at no level is it only marginally better than a CV1 for spotting.  It is a massive improvement in clarity, SDE and overall IQ.  The only area the CV1 wins is colours and contrast.

Share this post


Link to post
Share on other sites
9 minutes ago, ICDP said:

I'm sorry Chill but while I appreciate the effort your methods are wrong.  You are deliberately hobbling the Pimax 5K with negative supersampling because that's what you play at but that is a purely subjective test.  I used this analogy before and it still holds true, would you compare a 1080p monitor against a 4K monitor and run the 4K at 1080p for "fairness" because FPS at 4K was poor?  Would your conclusion in such a poor test be valid?

 

Your test was specifically aimed at spotting, yet where in this review has Screen Door effect been mentioned?

 

I have a Pimax 5K+ and at no level is it only marginally better than a CV1 for spotting.  It is a massive improvement in clarity, SDE and overall IQ.  The only area the CV1 wins is colours and contrast.

How does your setup perform? Is it capable of 90fps?

I would be interested in a pimax, but more for the larger FOV, than resolution.

i see no point in getting a HMD that todays top tier systems cant really handle.

Share this post


Link to post
Share on other sites
Posted (edited)
58 minutes ago, ironk79 said:

How does your setup perform? Is it capable of 90fps?

I would be interested in a pimax, but more for the larger FOV, than resolution.

i see no point in getting a HMD that todays top tier systems cant really handle.

 

Well if 90fps is your goal then you may stay on CV1 for a while :)

 

I have a Vega 64 and a Ryzen 2700X and I am getting average of 45 FPS in Il-2 at medium/high settings with HDR and SSAO off, 100% SS in Steam VR and 1 in PiTools.  It gives me images as sharp as I had in my Odyssey HMD but with improved SDE and of course the wider FOV.  This is in an 8v8 furball using Kuban map.

 

There is some warping in the peripheral view but it is acceptable for the trade off in wider FOV.

 

I am lucky in that I can handle FPS down to 40-45 FPS in VR without any ill effects.  It is not ideal and of course the more the better but I count myself fortunate that I could always handle lower FPS than most.

Edited by ICDP

Share this post


Link to post
Share on other sites

90 fps ??? this is funny I have 1080 ti i7 @4.8Ghz and never saw 90 fps in il-2 VR . Game looks horrible and performance is bad.  Just my opinion .

Share this post


Link to post
Share on other sites
Posted (edited)

That's the point, we can't measure subjective opinions because yours is different than mine and neither are wrong.  I found the Oculus CV1 pretty much useless for anything other than sight seeing no matter how much super-sampling I threw at it.  Despite the many subjective opinions about "once you try VR you will never go back" I actually found myself using my 4K monitor with the CV1 sitting unused for months at a time.

 

Chilli's test is focusing on what spotting is like at what he deems to be playable on both HMDs.  The entire test is based on his subjective preferences and gives almost nothing meaningful from a technological standpoint.  Apologies but I don't mean to make this a personal attack on Chilli, just his methods  For empirical objective tests to mean something they need to be at exact setting per HMD so we can determine what HMD gives the best spotting at identical settings.

Edited by ICDP
  • Upvote 2

Share this post


Link to post
Share on other sites

Note that SS values are meaningless by themselves, since they are relative to a resolution which is arbitrarily set to something higher than the panel resolution. For this reason it is not necessarily fair to compare with SS=1 in both cases. Even with SS=66%, the resolution is way higher than panel resolution of the 5k

  • Upvote 1

Share this post


Link to post
Share on other sites
Posted (edited)
4 hours ago, 307_Tomcat said:

90 fps ??? this is funny I have 1080 ti i7 @4.8Ghz and never saw 90 fps in il-2 VR . Game looks horrible and performance is bad.  Just my opinion .

After the last updates, i get 90fps are almost constant, at least in QMB, PWCG - career mode is a special topic, it stutters for god knows what reasons.

Settings are Balanced+ (Mod with balanced presets + tree draw distance from high preset), Ultra Shadows, Grass Ultra, Clouds high etc.

Looks fantastic (esp. since 3.008, possible to use me to use ultra shadows) and runs smooth.

 

Edit: OTT SS 1.2 @1600ish x1900ish

Edited by ironk79

Share this post


Link to post
Share on other sites
Posted (edited)
3 hours ago, coconut said:

Note that SS values are meaningless by themselves, since they are relative to a resolution which is arbitrarily set to something higher than the panel resolution. For this reason it is not necessarily fair to compare with SS=1 in both cases. Even with SS=66%, the resolution is way higher than panel resolution of the 5k

 

I agree but this is because SS is random per HMD and GPU type because Steam give recommendations based best FPS, not IQ.  Never the less, massively bumping SS on a Rift CV1, while doing the opposite for the Pimax 5K+ is not a fair way to do it.  Set the CV1 at 66% SS and it will look like muck in comparison, or set the Pimax 5K+ to 180% SS and it will look more than slightly superior.

 

Just as a note:  I would have set the Pimax to be run at narrow FoV which would allow higher SS settings to compare against the CV1 at the same SS.  FPS would not have been a major factor in determining IQ for spotting.  Get the baseline SS and playable settings on the Pimax and then use exact SS and graphical settings on both HMDs.  Then a more fair comparison at similar settings would be possible.

Edited by ICDP
  • Upvote 1

Share this post


Link to post
Share on other sites
2 hours ago, ironk79 said:

After the last updates, i get 90fps are almost constant, at least in QMB, PWCG - career mode is a special topic, it stutters for god knows what reasons.

Settings are Balanced+ (Mod with balanced presets + tree draw distance from high preset), Ultra Shadows, Grass Ultra, Clouds high etc.

Looks fantastic (esp. since 3.008, possible to use me to use ultra shadows) and runs smooth.

 

Edit: OTT SS 1.2 @1600ish x1900ish

 Please what rig and hmd do you use ?

Share this post


Link to post
Share on other sites
32 minutes ago, 307_Tomcat said:

 Please what rig and hmd do you use ?

Intel i7 8086k (stock), Gainward 1080Ti (OC), 8GB DDR 3200, Samsung 970 ssd 250GB, Oculus Rift CV1

  • Thanks 1

Share this post


Link to post
Share on other sites

The Rift works very well on mid-range systems. I have a i5-4670k@4.2GHZ/Vega 56. I get a steady 45 fps in QMBs/career which goes up to 90 fps once over around 1 km. My settings are balanced pre-set, high shadows/clouds, 110% SS/2xAA. Image quality is fine, gauges are a bit blurry, but the 2xzoom solves that when I have to read one.

 

No doubt the Pimak is better, but I would have to upgrade most of my system to get decent performance, so that will have to wait.

Share this post


Link to post
Share on other sites
Posted (edited)
11 hours ago, 307_Tomcat said:

90 fps ??? this is funny I have 1080 ti i7 @4.8Ghz and never saw 90 fps in il-2 VR . Game looks horrible and performance is bad.  Just my opinion .

I have original Odyssey, i7 6800k and 1080Ti. In 1v1 QMB at 50% SS I am getting constant 90FPS in-air. For spotting I am using black dot mod. So far as I am concerned image detail is easily acceptable. The image is nice enough at 50% SS that I have put my Odyssey+ in the cupboard. Main problem with the "+" is miniscule sweet spot with rest of image a blurry mess. My Rift was relegated to cupboard last June when I got the Odyssey and has been there ever since. To best of my recollection colors looked washed out compared to the Odyssey. Frankly I see the Rift as obsolete and the original Odyssey as best value at this time.

Edited by Dagwoodyt

Share this post


Link to post
Share on other sites
On 1/8/2019 at 3:04 PM, moosya said:

Imho, with 66% SS the undersample rate is extreme. I guess you're doing it for the purpose of high fps, but it doesn't speak truthfully to hmd's capabilities. Can you make these shots using 100% SS?

 

66% is below 100%, but it is rendering 2.3 more pixels than the physical display of the Pimax. The absolute SS is 2.3. (or 230%).

In any case, you are right. I have done also pictures at 100% and also 180% to see the capabilities of the display. I have edited the OP.

On 1/8/2019 at 7:12 PM, lefuneste said:

Very interesting, but could you vive us the gauges picture with 100% SS ?

Sure, Edited first post.

On 1/8/2019 at 8:28 PM, TunaEatsLion said:

Your idea about ss in the center only is a great idea!

Well in fact this is called "Fixed Foveat Rendering" (FFR). Basically render with better quality in the center and less in the edges (where the lenses work worse and your eye have less rods and cones). I think is something Oculus used, but not sure if it is used in SteamVR.

 

When eye tracking will be available then we will use "Dynamic Foveat Rendering"(DFR) so it would be even better.

On 1/8/2019 at 8:51 PM, SCG_Fenris_Wolf said:

SS = 180% on the Rift reduces contrast, since the backend information leads to merging colors/pixels when the picture is downscaled for output in the Rift again.

 

The higher the SS above native, the less contrast, the less blacks.

 

It would be interesting to do some tests with pictures through lenses to see how strong is this effect. Perhaps you have some example to show.

You should know that when  we use 100% the headset also apply a nice internal SS. That´s why I calculate the absolute SS column in the calcs.

  • Upvote 1

Share this post


Link to post
Share on other sites
17 hours ago, ICDP said:

You are deliberately hobbling the Pimax 5K with negative supersampling because that's what you play at but that is a purely subjective test.  I used this analogy before and it still holds true, would you compare a 1080p monitor against a 4K monitor and run the 4K at 1080p for "fairness" because FPS at 4K was poor?  Would your conclusion in such a poor test be valid?

 

Your test was specifically aimed at spotting, yet where in this review has Screen Door effect been mentioned?

 

I have a Pimax 5K+ and at no level is it only marginally better than a CV1 for spotting.  It is a massive improvement in clarity, SDE and overall IQ.  The only area the CV1 wins is colours and contrast.

 

As said, I used 66% not because I want it but because a 1080Ti can not give more in the IL-2 VR. Perhaps in other game it behaves better but I made the test for me and for IL-2 VR. I wanted to share the results.  It is true that there could be people with 2080Ti or Titan RTX which can go further and exploit better the capabilities of the headset. So, I also made pictures with 100% and 180% with the Pimax.

 

The SDE is another matter, I just wanted to share the pictures, not make a full review. But, yes, the SDE is less noticeable in the Pimax.

 

With the images, I just wanted to not be subjective. Sometimes when we all buy something new we tend to see it with better eyes. And subconsciously tend to believe that is better.

Here a camera has not subjectivities and I gives you what you will see in the lense.

But don´t interpret me wrong. I love the Pimax guys a lot. What they have achieved is simply great.

They have the guts to do what the PC VR community want: Clarity, resolution and FOV. Not what the money making Facebook are doing.

BTW, two questions:

What Settings and SS are you using for the Pimax5K+?

How would you prove that Pimax is much superior than Rift for spotting?

16 hours ago, ironk79 said:

How does your setup perform? Is it capable of 90fps?

 

Well, your CPU and RAM is a bit better than mine. So it should help as well a bit.

Still not used a lot since I am putting a new 18mm cushion which corrects a distortion at edges.

I have purchased fpsVR app and I am doing some testing to see when GPU or CPU or both at the bottleneck.

Just initial impressions with CPU OC at 4.8GHz and SS=66% and the settings of the benchmark using the spitfire in Kuban I am most of time in the 60-90 fps depending altitude and where I look at. When heavy combat then I can go to 50s. (this is when CPU is demanding).

Edited by chiliwili69

Share this post


Link to post
Share on other sites

Oculus has not used foveated rendering yet but are working on it.

Share this post


Link to post
Share on other sites
7 hours ago, dburne said:

Oculus has not used foveated rendering yet but are working on it.

 

If you refer to use foveat rendering coupled with eye tracking, then you are right.

 

But Oculus already support fixed foveat rendering techniques to decrease the load of the GPU.

 

You don´t need eye tracking to do fixed foveat rendering:

https://www.vrfocus.com/2018/06/foveated-rendering-doesnt-need-eye-tracking-hardware-thanks-to-some-oculusunreal-engine-wizardry/

 

In the Rift it is Mask Based Foveat Rendering:

https://developer.oculus.com/documentation/unreal/latest/concepts/unreal-mbfr/

 

In the Oculus Go it is Fixed Foveat Rendering:

https://developer.oculus.com/documentation/unreal/latest/concepts/unreal-ffr/

Share this post


Link to post
Share on other sites
3 hours ago, chiliwili69 said:

 

If you refer to use foveat rendering coupled with eye tracking, then you are right.

 

But Oculus already support fixed foveat rendering techniques to decrease the load of the GPU.

 

 

 

Interesting, but really does not apply to what we do. Guess that is why I have not seen really any discussion on this .

Hopefully they will continue working on this to where it starts to become a little more mainstream for VR with future devices.

Share this post


Link to post
Share on other sites

Thank you for all the hard work doing this Chiliwili. :)

Any chance of showing some comparisons with the Pimax and Rift at just their native resolutions (zero SS etc)? I don't use SS on my Rift, I get a clearer image for spotting without it. I'm very interested to see how much better the Pimax would be. :)

Share this post


Link to post
Share on other sites
On 2/9/2019 at 7:04 AM, Uffz-Prien said:

Any chance of showing some comparisons with the Pimax and Rift at just their native resolutions (zero SS etc)? I don't use SS on my Rift, I get a clearer image for spotting without it. I'm very interested to see how much better the Pimax would be

 

When you say "zero SS" I suppose you mean just 100% in SteamVR or 1.0 in Oculus Tray tool .

I already made samples with Pimax at different SS including 100%, below and above in this thread.

 

I also made some comparison for gauges details just with the Rift: Look here.

 

I already sold my Rift, so no more experiments with Rift.

Putting no SS (or even using less than 100%) would allow you to get a more defined pixel for ID (the plane dot is not diffuse with the sky around), but on the other hand, you will have less defined gauges and cockpit details and worse ID.

Share this post


Link to post
Share on other sites

No, I meant actual zero (in OTT, I don't use the steam VR). :)

The difference between zero and any amount of SS vs FPS is so negligible as to make me wonder why anyone bothers with it. Maybe it's a bit panaceatic. Lol.

Share this post


Link to post
Share on other sites
5 hours ago, Agathos_Deimon said:

 

Prien, dont forget that Zero (which gives same Pixeldensity as setting it to 1 or setting steamVR at 100%) in OTT still gives the RIFT some SS, as the RIFT on default does not render the world at its native resolution. To actually have the RIFT render VR in ins native resolution of 1080x1200 per display you have to ajust the SS below 1 .

 

And i find there is a definite difference between no SS or a value of something like 1.1 -1.4 in IL2 AA will result in a very similar effect, while in DCS or WT i find SS the better way to improve quality.

How is zero not "adjust the SS below 1" ? 

Share this post


Link to post
Share on other sites

Plz guys, how with the Pimax 5K can you see your keyboard ? Is there a frontal dedicated camera in it ? Can it slide ? Flip ?

Share this post


Link to post
Share on other sites
22 hours ago, Solmyr said:

Plz guys, how with the Pimax 5K can you see your keyboard ? Is there a frontal dedicated camera in it ? Can it slide ? Flip ?

Use the Force, Luke!

Share this post


Link to post
Share on other sites
On 2/11/2019 at 10:15 AM, chiliwili69 said:

 

When you say "zero SS" I suppose you mean just 100% in SteamVR or 1.0 in Oculus Tray tool .

I already made samples with Pimax at different SS including 100%, below and above in this thread.

 

I also made some comparison for gauges details just with the Rift: Look here.

 

I already sold my Rift, so no more experiments with Rift.

Putting no SS (or even using less than 100%) would allow you to get a more defined pixel for ID (the plane dot is not diffuse with the sky around), but on the other hand, you will have less defined gauges and cockpit details and worse ID.

 

Chili - try testing PiTools render at 2

SteamVR Video Render at 20% and then SteamVR application setting render at Pimax 5K native resolution or a factor of x2 there of.  Start with 1440 as the vertical render res and then 2880 followed by 5760

 

This is something I am testing at the moment on my 8K but I am not sure if I stick with a factor of 1440 base vertical res or try 2160.  The 1440 would make sense as that is what is transmitted to the headset to the 8K scaler chip but I also try the 8K panels native res 2160 on the vertical as well ans see what comes of it.

 

I think the 8K down the track with new cabling and removal of the scaler chip but with eye tracking and foveated rendering based on eye tracking or even fixed but with the center being at native 4K, that would be great.  There is much potential with these headsets and the wide (150+)  FOV is great for flying in.

Share this post


Link to post
Share on other sites
On 2/14/2019 at 6:20 PM, Solmyr said:

Plz guys, how with the Pimax 5K can you see your keyboard ? Is there a frontal dedicated camera in it ? Can it slide ? Flip ?

 

Touch type.

Share this post


Link to post
Share on other sites
15 hours ago, BlueHeron said:

 

Touch type.

 

? Do you mean you're blind once it's on your head ?

 

Or is this an unguided joke ? (Don't see any condenscending there, but asking because of your rig in signature.. Or is it outdated ?)

Share this post


Link to post
Share on other sites
6 hours ago, Solmyr said:

 

? Do you mean you're blind once it's on your head ?

 

Or is this an unguided joke ? (Don't see any condenscending there, but asking because of your rig in signature.. Or is it outdated ?)

 

D'oh!   yes, I totally forgot to update my signature, I'm running an I7 and 1080Ti now.  :)

 

Correct, I can't see my keyboard at all but for the few keys I use I'm able to find them by touch.  It still breaks immersion a little so I try to map most functions to my HOTAS.

Share this post


Link to post
Share on other sites
44 minutes ago, BlueHeron said:

 

D'oh!   yes, I totally forgot to update my signature, I'm running an I7 and 1080Ti now.  :)

 

Correct, I can't see my keyboard at all but for the few keys I use I'm able to find them by touch.  It still breaks immersion a little so I try to map most functions to my HOTAS.

 

Haha ok, I better understand ! ;)

 

Anyway, the Pimax 5K would have been way too much demanding for my GC, and when I change it I'll not spend 800-1200 euros, more likely in the 400-500.. I finally ordered a O+ and am waiting for it to arrive.

Share this post


Link to post
Share on other sites
On 1/8/2019 at 1:07 AM, chiliwili69 said:

With the three images above, one would conclude that it is not worth to buy the Pimax (700$ only headset) having already the Rift

 

I received my PiMax 5k+ yesterday (backer 6XXX) and have been experimenting in DCS and IL-2.

 

So far, I think it's unusable in DCS, we need reprojection or that DCS is made less CPU and GPU-hungry. I'm using the latest PiTool version, non-beta (so no brainwarp).

 

In IL-2, I think it needs above 2800 lines to give a nice picture ideally more, but then framerates suffer a lot. I'm using an RTX2080, and with the PiMax I'm GPU-limited. To get decent framerates 75-80 I have to stay at or below 2500 lines. If I run the rift with the same resolution, I get comparable, maybe even slightly better clarity than in the PiMax.

 

As I have no base stations, the PiMax is 3 dof only for me. This prevents me from seeing my six, even though the FOV is wider.

 

My conclusion is pretty much the same as chiliwilli69. It's an interesting headset, but for now I do not recommend getting it if you already have the rift.

 

Moreoever, PiMax is having a meltdown when it comes to support requests, they have closed their support and their mail, and many people have problems with their headset. No problem for me, other than a velcro that has started coming loose, but many have problems with their cables, which apparently are insufficiently shielded, or with the screens (dead pixels, black dots...). I think they'll eventually get that fixed, if their reputation isn't ruined by the time they manage to catch up with all the issues that need attention. This is another reason why you should not buy now. Maybe things are better 6 months from now.

Share this post


Link to post
Share on other sites
3 hours ago, coconut said:

 

I received my PiMax 5k+ yesterday (backer 6XXX) and have been experimenting in DCS and IL-2.

 

So far, I think it's unusable in DCS, we need reprojection or that DCS is made less CPU and GPU-hungry. I'm using the latest PiTool version, non-beta (so no brainwarp).

 

In IL-2, I think it needs above 2800 lines to give a nice picture ideally more, but then framerates suffer a lot. I'm using an RTX2080, and with the PiMax I'm GPU-limited. To get decent framerates 75-80 I have to stay at or below 2500 lines. If I run the rift with the same resolution, I get comparable, maybe even slightly better clarity than in the PiMax.

 

As I have no base stations, the PiMax is 3 dof only for me. This prevents me from seeing my six, even though the FOV is wider.

 

My conclusion is pretty much the same as chiliwilli69. It's an interesting headset, but for now I do not recommend getting it if you already have the rift.

 

Moreoever, PiMax is having a meltdown when it comes to support requests, they have closed their support and their mail, and many people have problems with their headset. No problem for me, other than a velcro that has started coming loose, but many have problems with their cables, which apparently are insufficiently shielded, or with the screens (dead pixels, black dots...). I think they'll eventually get that fixed, if their reputation isn't ruined by the time they manage to catch up with all the issues that need attention. This is another reason why you should not buy now. Maybe things are better 6 months from now.

Interesting observation Coconut.

 

I would agree with you with the Render Target Resolution needing to be high (in the vertical as the horizontal changes pending on the FOV setting you have dialed in).

I have the Pimax 8K and have been testing it extensively and again when I put together my eGPU based on a 2080.

 

At first I was thinking dial in vertical res the same as that of my HMD's native resolution so 2160

Then it was a thought of maybe dialing a resolution based on the resolution that gets sent upline to the HMD's scaler chip so 1440 then 2880 and then 5760

After I tried 4320 and then tried also returning back to SteamVR Video Auto resolution which gave me a target of 2755 with the RTX2080.

 

Problem I found was PiTool and SteamVR are not very condusive to dialing in set target resolutions - I am using the SteamVR Beta and PiTool Beta by the way.

 

For the 2880 and 5760 resolutions I had to use PiTool setting of 1.25  with SteamVR SV26% and SA134%

 

For 2160 and 4320 I had to use PiTool setting of 1.75 with SV20% and SA50% and 200% respectively.  Il2 on Balanced with high clouds and shadows.    I experimented with AA at off, x2, x4 and with Nvidia AA settings at x8 augmenting ingame AA.  Still wondering about that.  Also was trying just the 100% setting on SA with PiTool at 1.75 and SV at 20% - it gave quite a nice view.

 

You will not get 90fps out of the HMD at decent Target Render Resolutions but even with frames down to the 40's Il2 is still flyable on the Pimax HMD's and for me at least, I have no issues flying campaigns.  Checking 6, well - yes, do the "Crazy Ivan" and check it that way.  You still have reasonable rear view and not having to do a 180 head turn helps (joking - thinking of The Exorcist here).  What does bug me with the no show (thanks to Valve/Vive as they are not releasing V2 of the lighthouses to 3rd parties) of the lighthouses and 6DOF is lining up targets in the gun sight. At the moment I have have repositioned my center view forward slightly but with the Soviet Sights - it still sucks.

 

Looking forward to effective forward foveated rendering and also what their reprojection/BrainWarp will bring to the table.  So far in the Beta PiTool BrainWarp was Warping good but not they way it was intended.

 

As for Pimax Support - I have found through the forum and forum PM system their support has been good.

 

Pity you are coming from an Oculus Setup as you would have been able to use your lighthouses and controllers if it was previously a Vive System.

 

You are correct with your observations with Support though - it should be much better down the track especially as they clear out the early poorer Quality Control HMD's that were shipped to backers.  By years end or so, I would be very keen to see if they have a 8K-X (native res supported 8K headset).  Very keen as even on my current setup - I believe the scaler chip is a bit of a pest - either 4K native x2 or using foreated rendering - bypassing the scaler chip on the current 8K HMD's

 

 

Share this post


Link to post
Share on other sites

There is just something about that term "Brainwarp" that bothers me...

I think they could have gone with a little better name lol.

:wacko:

Share this post


Link to post
Share on other sites
5 minutes ago, dburne said:

There is just something about that term "Brainwarp" that bothers me...

I think they could have gone with a little better name lol.

:wacko:

Maybe it is something to do with their Social Credit System - those sneaky Chinese they will have us all melting down via BrainWarp if we do not comply 😁

 

Least someone is pushing VR boundaries and getting the big players off their $$$ Asses.

Share this post


Link to post
Share on other sites

Coconut thanks for your observations.

Since I am very happy with the Odyssey + ,  after reading how is working and many others, I just cancelled my Pre order for the Pimax 5K

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×
×
  • Create New...