cancel
Showing results for 
Search instead for 
Did you mean: 

The Pimax 8K MEGA Thread - First Reviews Now live

Shadowmask72
Honored Visionary
I've been on the fence about this and have some reservations about build quality and performance under heavy load, but for me the positives outweigh the negatives and I guess many others feel the same way. Now I never backed the Rift during its Kickstarter campaign (unfortunately) but I assume the feeling is similar where you're helping to forge the future of VR in some way, or at least like to think so. I consider myself an enthusiast and not the majority, therefore what's another £600 to experience the latest  in VR. Sadly, having Samsung not release the Odyssey here in the UK left me with an itchy wallet finger, a void that needed filling and with GO not releasing until next year either I thought the Pimax 8K will scratch the itch and might just prove to be pretty good too. I'm excited for it.

There is always going to be skepticism and no doubt a number of people will suggest I am backing a paperweight but, that's ok it's fully understandable to have those feelings and that prediction is always a possibility. But... if someone doesn't take the risk and we all play it safe, how can VR move forwards - I guess we've already taken risks with past VR investments in some way or another including with the Rift. I've never claimed any loyalty to anyone having owned The Gear VR (still have that knocking around somewhere) DK2, Vive, Rift, PSVR and come next year Pimax will be added to the list. However, I am interested like many of you agreed in the poll we did here a while back that certain things were important to improve on what we have now and it seems the Pimax 8K is offering some of those things on paper. 

I think once you can look past the silly marketing name (8K) and the hammerhead shark design and focus on what's inside, this HMD offers something many claimed they would like in the past but are not prepared to trust Pimax to deliver. I think a number of people would rather wait for a more established company (like Oculus) to offer a similar experience further down the line. All very understandable. But I am impatient, and whilst I use my Rift quite often I want more of what VR can offer sooner rather than later. 

So, (and the main reason for this post) is come Feb 2018 some point in 2018 I will gladly post impressions and comparisons without hyperbole and answer questions any of you might have. That said, I expect many others at the time will post their impressions also so there should be plenty of opinions going around from various sources.


System Specs: MSI NVIDIA RTX 4090 , i5 13700K CPU, 32GB DDR 4 RAM, Win 11 64 Bit OS.
1,860 REPLIES 1,860

Shadowmask72
Honored Visionary
Despite some negative press including an article from one publication that wasn't even a hands-on getting tossed into the mix, some impressions from the CES show floor. Although always be wary of these hand picked impressions alongside the validity and experience of those commenting.

https://www.youtube.com/watch?v=0n42dJBESLM


System Specs: MSI NVIDIA RTX 4090 , i5 13700K CPU, 32GB DDR 4 RAM, Win 11 64 Bit OS.

zboson
Superstar
I'm still trying to understand what hardware design is optimal.

As far as I can tell the focus of the eye drops rapidly within a small window in degrees.
vxkayg61kppu.png
https://en.wikipedia.org/wiki/Visual_acuity

Another problem is that displays are usually flat. This means that at angles further from zero that the arc intersected per degree increases. In other words the pixels per angle increases further from zero. This is the exact opposite of what would be optimal due to the rapid drop in focus away from zero. That's probably why some want to have curved screens so that the pixels per degree is constant. I think I read something about Microsoft putting LEDS at higher angles which produced horrible resolution but they eye could not tell the difference. So I think having a flat 8k screen to large angles is just a huge waste. It's the brute force solution. I wonder what could be achieved with piecewise displays  which drop in steps from the regions that the eye could focus.

hoppingbunny123
Rising Star
look at the side of your face towards your nose there's a ratio, the comfortable eye position as the eye points to the nose and the extreme turn of the eye that's slightly painful to look towards your nose, thats on a curve between the two values, find that curve radius value and map it across the entire eye to find the curve the eye uses when it looks at the nose to see. The small value is seeing number, this curve is probably the maximum allowable curve the eye sees comfortably in.

nosys70
Expert Protege

probably oculus will jump on it, easy way to fill the gap with pimax, and after all they are already hand to hand with samsung...(something like your hardware/our software)

Pimax being chinese and samsung korean, i do not see a match here.

Chinese would likely wait to get the product locally for cheaper.

CaptainInfamous
Protege

kojack said:

The rift has a native resolution of 2160x1200 (or 1080x1200x2, both have the same pixel count).

Pixel density isn't a direct multiple of this. 1.0 doesn't mean 2160x1200 and 2.0 doesn't mean 4320x2400.
Pixel density is the desired ratio of texels to pixels in the centre of each eye after the post processing distortion phase. Due to the barrel distortion done in the distortion phase, you need to have a higher resolution (centre pixels are bulged larger than outer pixels).

To achieve 1:1 ratio of rendered texels to panel pixels (this is pixel density 1.0), you need to render at 1.23 times the native res. (DK2 lenses needed 1.5 times, DK1 was around 1.7 I think)
So setting the density to 1.0 means your game is really rendering at 2664x1586 (the panel is of course still 2160x1200).
Pixel density 2.0 means two full texels (per axis) per panel pixel in the centre of your eye, which means the game is rendering at 5328x3172.  Get out your calculator and you'll see that 5328x3172 is 16,900,416.

So when it comes to performance, the Pimax 8K X sounds like it will require extreme hardware. But anybody who can run a rift game at density 2.0 is already doing a bigger GPU hit.

Potential vertex pipeline hit due to wide fov, however, is a discussion for another time (I'm at work).



Not only that, but given that sli vr rendering is still in its infancy, even with a huge advance in hardware there aren't any GPUs out there now that can do 8k VR resolution on one gpu.  This may have been discussed in another post, but since I don't have time to read 31 pages I'll just assume it hasn't. lol.  When you render VR, it's not just simple pixel multiplication that has to be taken into account.  You also have to account for off-camera rendering that the gpu does in order to prepare itself for when you move your field of view, and it has to do all these calculations twice since the viewpoints aren't identical, but have different depth, slightly different perspective, etc. that must be rendered independently to make the result stereoscopic - not to mention it has to keep a 90hz refresh rate to keep us all from barfing.  So it's not just the simple 16.9M pixels that it has to render, its more like trying to run two different games at 4k simultaneously on different screens in order to  give you 8k resolution, and doing both at 90fps.  A little compute time is saved on overhead obviously by running the same exe code, but I think this is a good way of visualizing the power required to generate that pixel depth in vr.  There may be some rendering power efficiency to be gained by using foveated rendering once eye-tracking is added to HMDs, and that could take the load off the GPU allowing higher res areas at your focal point while dithering areas that are out of focus more akin to how your eyes actually work.  To get 8K VR seems like a REALLY tall order without some serious R&D backing to make it happen - and to do all that for only 250 units seems a little insane to me honestly lol.  

At any rate, I'm mad curious to see how their project turns out and what new tech they come up with to make 8k a reality - even if it doesn't quite hit the mark, it's likely to open up a whole new range of VR tech.

pyroth309
Visionary
My concern with the Pimax, and maybe someone here who is more familiar on the Dev side knows, is how many games really allow a FoV past 110? Most games cap between 90-110 and don't allow wider for competitive reasons. Seems to me like stretching a 110 fov to 180+ would look bad. Is there a solution for that? 

zboson
Superstar
I found the article about Microsoft using LEDs at large FOV. They also found that it reduced motion sickness. That's very interesting because one method to reduce motion sickness is to reduce the FOV so a technique that reduces motion sickness by using a higher FOV is highly desired.

https://arstechnica.com/gaming/2016/05/how-side-mounted-leds-can-help-fix-vrs-tunnel-vision-and-naus...

https://youtu.be/o9LVwl8cmc0

Something like this seems to me to be a much better solution. Upgrade the resolution in the region they eyes can focus. I am not sure exactly that that is but I think probably the Rift and Vive got that covered. Somewhere around 100 degrees. But then add low resolution displays at larger degrees (apparently even some LEDs may be sufficient). This would probably save costs and require less rendering performance/power.

BTW, I never thought I would say this but my impression of Microsoft has improved a lot at least relative to other big tech companies.

Iriodus
Explorer
I've been following all of the Pimax "8K" threads that have popped up on the forums pretty much since they started opening up, and my 2 cents on the matter:

I'm going to hold my overall judgement of the Pimax 8K until after the final product is in the hands of reviewer, and ideally I'd hold my "final" judgement until after I have had a chance to try it myself. I maybe know of one person that may jump for this headset once all the backers get their headsets, but that's not a for certain thing.

Ultimately, I'd be more excited for a headset that has better FOV and resolution, but to the extent that the increased resolution, FOV, etc., is offset by any High Tech Sorcery (hardware or software) that has been done to improve performance (Foveated Rendering). The ideal upgrade for the Rift, for me, would be said High Tech Sorcery offsetting the improved specifications to the point where you can use the same specifications as the current Rift (More or less). I think that this is a better target to shoot for than making an enthusiast headset that only a Niche Of People³ would buy.

Phil007
Protege
Im really interested to see how these new headsets turn out. I'm sure Oculus won't leave it too long until they announce a successor. 

Again, about the 8k x resolution... I read that rendering at 2x pixel density is roughly the same as the Pimax. However, i feel that even a 1080ti cannot reach this at 90 for most current more graphically intense games. Actually quite far off. Also we need to consider, that future games over the next couple of years will be even more gpu intensive. Negating somewhat, the boost new gpus will give. A gpu in 1yr that is 40% quicker isnt going to run the games released in 1yr at 40% fps boost.

Iriodus
Explorer

Phil007 said:

Im really interested to see how these new headsets turn out. I'm sure Oculus won't leave it too long until they announce a successor. 

Again, about the 8k x resolution... I read that rendering at 2x pixel density is roughly the same as the Pimax. However, i feel that even a 1080ti cannot reach this at 90 for most current more graphically intense games. Actually quite far off. Also we need to consider, that future games over the next couple of years will be even more gpu intensive. Negating somewhat, the boost new gpus will give. A gpu in 1yr that is 40% quicker isnt going to run the games released in 1yr at 40% fps boost.


Yeah, I'd rather any game I play (VR or otherwise), play at a stable FPS, and having my FPS be as high and stable as possible (in general) is more important to me than high resolutions. It is for this reason that I have no intention of moving away from 1080p at anytime in the forseeable future, my 1080p144hz display is sufficient for my purposes, and would only replace it if I fail. Were it to fail, I would see if there are any 1080p high refresh rate gaming monitors still in production and get that, but if not I would do the same for 1440p and so on.