Pimax 8K Progress [Updated: Feb 6, 2018]


#397

More a good news for miners…


#398

Where did you see it? Every one talks about 12nm for Ampère.


#399

This info is speculation from forums
10 nm is available
12 nm is a progress too little from 16 nm after 2 years for Nvidia
With 12 nm you have a refresh Not a jump in graphic performance Because you can’t make for gaming a big chip


#400

Then no, it’s 12nm, like the nVidia Titan V (V from Volta) just released.
Well, it could be 10nm. Even 7nm too. But, for now, it’s a rumor.
Update: at the end, it seems that it’s 14nm.


#401

#402

Pause at 43 seconds, looks like a different lens to the last one…


#403

Picture quality is insane,Love it !!!


#404

I’m liking it. What surprises me though is after all these videos they are still hand holding the camera. You would think they could make some sort of clamping system that would hold it in perfect position over the lens while shooting. I guess they don’t get MacGyver in China.


#405

I’m gonna SS the shit out of this game! :joy:


#406

With what rig, @xunshu?


#407

Given how jerkily things move in the video, I wouldn’t expect too much room to spare for additional supersampling… :stuck_out_tongue:

I too have been struck with the thought they should 3D-print a mount for a high resolution, fixed focus camera (EDIT: maybe even vblank synchronised), for these videos. For that matter; At some point they are going to have to build a specialised rig for the calibration of each individual unit built (assuming quality assurance does reach as far as one might expect).

Not to be a dark cloud imposing itself on festivities, but I am unsure how we can proclaim picture quality from seriously blurry youtube videos, which (with the exception of the occasional nearsighted projection screen user) we view at significantly higher resolutions, in terms of Pixels Per Degree, on our monitors (maybe not even full screen - just the quarter-window stamp), than we’ll end up with, wearing the actual headsets. :confused:

(EDIT2: I suppose it may be prudent to explain the PPD matter a bit closer, in case somebody not familiar with VR see reason to object… Consider that what we see in a monitor window from a certain distance (normal use, that is), fills a really small part of our field of view. A door whose width from a given position is drawn across, say, a decimeter of monitor real estate, will appear its full meter to us, when in the HMD, using the same rendered pixels. (EDIT3: …and this goes even if this had been raw screen capture data, instead of poor cellphone footage))


#408

I wouldn’t proclaim picture quality based on their “through the lens captures” on any of the videos. I can say that it doesn’t look any worse than my Rift at about the same distance from the lens on close ups. The background does appear somewhat sharper as far as it can be assessed. The smoothness is a “little” off when looking at the people running down the street. considering it’s V5 with all the baggage that was CES and Fallout 4 VR which is a bitch to run on my Rift with it’s mature software, I am pretty happy with what this prototype is showing so far. Personally, I am still excited about what’s going on. Fallout 4 VR is an after thought as VR titles go so I am inclined not to let it’s performance bother me. I have a lot of titles that added VR after the fact -DCS, Elite Dangerous, Battle of Stalingrad, Iracing, Assetto Corsa.etc. They all behave at different levels depending on so many factors and especially based on how well they took to being adapted for VR.
This is true of other hmds and will be true in the Pimax 8K. I feel made for VR titles will run well and those that have added it later will perform relative to when and how well it was added after the fact. Things run better on the Rift now than they did 2 years ago. No reason to think what we see in a prototype is the end result.


#409

As long as they keep shooting these videos at 25 FPS I cannot really take them seriously.


#410

I thought we were due another update this week.


#411

Since the FoV is so large no through the lens will do it justice. This simply gives a rough idea on its potenial.


#412

and we should take you serious? did you even backed them or not?


#413

No-worse-than-my-Rift is very reasonable. There is continually a lot of people setting themselves up for disappointment, however - not just here - so now and then I can’t help but feel a reality check is in order. :7

A significant part of optimising a game amounts to limiting the amount of stuff (and how complex that stuff is to render) that is thrown onto the screen at any one time, and with a scant few exceptions, most of which are very rarely implemented, that is essentially the only thing that differs, performance wise, between a made-for-VR title, and one that has been retrofitted for VR: It uses simpler graphics (clever art direction can as always reduce the perceptual impact of this), in order to hit 90fps with two viewports, where a game targeting monitor/TV does not only only need to render once per frame, but can also often get away with framerates down to the thirties, and even dipping into the twenties.

Developers have already used every trick in the book to coax performance at a given “bling level” out of their games without VR – there is no magic spout that can be opened, in order to make the drawing of a given picture suddenly go twice as fast, only because it goes to a different display unit.

I actually had an argument, last year, with a fellow who was absolutely convinced Fallout4 VR would run like Usain Bolt, due to optimisations for VR, whatever he may have imagined those would entail.

On the contrary, as it happens; There are a lot of dirty little optimisation tricks that you can easily overlook when playing on a flat screen, but which fall down when seen in stereo, and just to make sure every road is an uphill one, all that eyecandy that one carefully rations, for performance, one really really do want in VR; Texturing that holds up to rolling one’s detatched eyeballs across? -Absolutely! Realtime dynamic lighting and unlimited highres shadows, as well as perspective correct reflections? -Yes please. Actual dense geometry, or at least parallax mapping, so that a rough stone wall does not look like it is painted onto drywall? -Well that goes without saying, as does distant land LOD that looks somewhat like its close-up counterpart. Physics on everything, yaay! :stuck_out_tongue:

Hopefully sooner, rather than later, foveated rendering will help with much of this. :7

Given the respective FOVs and render targets of the devices, I expect a tiiiny bit better resolution out of the 8k, than the Rift CV1 (actual rendered stuff - not physical display pixels, and quite possibly more of an improvement horizontally than vertically, due to screen utilisation and video scaling matters). This means that if you are supersampling today, in order to make things look acceptable, it is quite likely you are going to want to do it with the 8k as well. This on top of the fact that there is 2.84 times as many pixels to render, to begin with.

As for things running better on Rift today than at launch; I don’t know whether this is in reference to developers having learned their optimising, over time, or changes on the Oculus software stack side, but in the case of the latter; The compositor is what it is, and was always tight and with little overhead; The big difference lies in mitigation strategies for when rendering can’t keep up, but if one get to the point where more than a handful of frames have to be synthetsised per minute, one have already failed, as far as I am concerned. Some people claim that fake frames look close enough like real frames to them, that they don’t really notice… Those people are blind, I suppose - they also claim that they can “forget” and not notice the screen door effect and “god rays” in the fresnel lenses, not to mention binocular rivalry…


#414

Pimax. You Need to get some of these units in the hands of your early adapters As Soon As Possible. Let them know where you are at, and even if late, get some of them in Early Into The Process. This will go a tremendous way towards keeping the buzz going and making sure your aim is on before you fire up production and make a huge mistake. Your core group of users will let everyone know you are working on this even if they can’t talk, even if you have to cross ship multiple prototypes. Let someone behind the curtain now and let them say the show is still on and it’s looking good and to be patient while the curtain is still closed before you lose your audience. You’ll have more time that way. And you don’t want to come out with your fly down.

I’m about to build a new 5.2ghz and faster system just for VR. And I’m hoping the Pimax can be the jewel in the crown. I prefer my gems closer to flawless rather than large and gaudy.


#415

I’m also a little bit concerned if going the fresnel lens path. As I’ve seen both now, and the Pimax 4k (BE) lense is just in a class of it’s own, with no Godrays, so much more desirable, as compared to a Samsung Odyssey at same resolution display. Brightness is also a very important issue, although I think it may have something that I stare at 3 bright screens all day, and then stare at the Texas sun when I crawl out of my hole. Maybe I should just turn them down.

I am really sore at breaking my BE model. I have to say, that headset was More Comfortable than the Samsung Odyssey and I think the screen door was less visible even though the same resolution. And it had a real lens instead of the Samsungs fresnel crap. If you could get tracking like the Odyssey on the BE model it would be a contender against the Odyssey and Vive Pro. I’m sorry but the Pimax 4k regular is just too dark for me.
And I took my BE model apart because it was too dark too, and broke it. Only now I see that even the Odyssey isn’t any better at all. And it feels like I have an epilepsy helmet on my head to wear compared to BE.

This is why I hope and think the funding is going to bring great things. They did all that with no real backing. Amazing.


#416

Since around 3.5 months, we’ve not heard a single update about the progress of 8K-X. To my knowledge, there are around 400 8K-X backers.

The time that has passed with no information seems a little excessive in my opinion.
Can we please have some news on 8K-X, even if limited?

Thank you.