Brainwarp Software Progress?


#41

they’ve said in another thread that they don’t have it working yet.


#42

That is more than a bit concerning. That’s like saying your new prototype electric vehicle has a new system called bubbleworks which extends your mileage using dish soap. LOL.

That’s of course an comical exaggeration but hopefully you get my meaning. If they don’t have it working they probably should have stayed hushed about it. I was sold either way, but now I’m like “so about that brainwarp thing you mentioned earlier…”.


#43

True but brainwarp should still make it feel like your fps is 180 instead of 90. Hopefully.


#44

That’s the idea but not sure how well it will work since the displays can only do 90hz. When the PSVR up Hz from 60 to 120hz it’s so effective because the panels are actually doing 120hz.


#45

As far as I know, this is SMP and Nvidia said it should save about up to 30% GPU load. It seems though, that it’s not just an easy task to implement this in games… no game developer (I know of) makes use of this or stated that they will implement it!
At the moment this is kind like Brainwarp, just a note of a company, that hasn’t proven that it’s a step forward for VR… sadly, I have to say…


#46

Whi;e a native panel of 180nz would be really cool, its also probably true that in those upper ranges of frame rates and also refresh rates that as it becomes harder to discern what the actual frame rate is for your brain, your brain likely isnt discerning which eye is giving the most current motion data, as there is much consistency between each eyes visual information. I would guess also, that the longer your use brainwarp the more effective it is.


#47

let’s say for argument sake that Brainwarp works 100%, now the question I have is… What happens when i’m playing Arizona Sunshine, Brookhaven, Organ Quarter, Dead Effect 2, Pavlov, Onward and the countless other shooter games when I close 1 eye to aim down the gun sight?


#48

same thing that will happen if you use it while only having one eye. you will see the refresh of only one screen . Good catch btw. What i would say as a counter to that is gun pros will often tell you not to shoot with one eye closed irl.


#49

I understand irl practice but in most of these games it’s hard to see the sights on the weapons with both eyes open. That being said, maybe the inherent increase of resolution will make that task easier on the Pimax.
When you say “you will see the refresh of one screen” will it be at the full frame Hz or will there be flickering ?


#50

You see a normal 90fps


#51

No flickering. They use brainwarp and also reprojection.


#52

Both can be used simultaneously?


#53

Yeah, they do totally different th8ngs


#54

Honestly, I just want the damn HMD. My brain is just flooded with specs at the moment. Space warp, time warp, brainwarp…I’m exhausted.


#55

welllll, somewhere between 82 and 90, will see. XD


#56

So the way I see it from their diagram is it sends one screen a picture and the other screen is blank. So what’s the difference between sending two pictures to two screens than sending one picture and one blank? If the screen has to go blank for a cycle is that just not the same as the screen refreshing?


#57

My guess is the manufacturer of the panel is contributing some of that manpower.


#58

Each panels redraws every 11 ms (I know that 90Hz aren’t guaranteed, but just imagine it). Now video card renders both images and sends them to the HMD. In the classic situation both panels are synchonized and are redrawn simultaneously. With Brainwarp that is not true, there is a shift between panels redraws, which is less than 11 ms, probably half of that, so the frames are evenly spaced out in time.


#59

Brainwarp is not something panel manufacturer cares about, panel refresh rate is same, chip that is controlling the signal (ANX7530) to LCD and some logic before I guess, maybe it’s part of the chip, to enable displays to refresh with some time in between, not simultaneously.
And I was wrong about rendering at higher FPS(120-180) monoscopic, chip used doesn’t support that mode, so it will be for example 5120x1440@90Hz/FPS no 2560x1440@180, so no need for change in rendering pipeline, GPU works just the same. It is similar to technique used by PSVR, there is case where console render 60FPS but it is doubled to 120 to match 120Hz possible with PSVR HMD(OLED). Pimax will not double 90FPS to 180, but actually show half of image half of the time(T=(1/RefreshRate)/2 s), so it’s simulated double Hz/FPS, interesting but not proven concept yet.
Good news is that Brainwarp will probably work with any VR game, bad news, no monoscopic trick to reduce GPU load… And probably in my idea problem is CPU has to enable game to run @120-180FPS so would double CPU workload, not wise at all. :slight_smile:
Example:
T0 = 0ms(start time) : GPU provide image 5120x1440 to HMD(buffer), left eye part sent to L display(LD), RD is off
T1 = 5.5ms(1/90hz/2): HMD already has R eye image(in buffer), display 2560x1440 on RD, LD backlight is off
T2 = 11ms: GPU provides new 5120x1440 image, left half sent to LD
T3 = 16.5ms: right half of image sent to RD .
Note: This is example for 90Hz so GPU has to render or at least provide 90 frames(some duplicated) to HMD. Also when I write sent to display(left or right), I think backlight is turned ON at that point, I assume crystals are actually set before that, so while we can’t see this, crystals on display close to eye that is not seeing image are getting the part of the image from buffer. Hope someone understood what I meant :slight_smile: .


#60

If it’s true that Brainwarp just delays the image to one of the eyes then it’s not very impressive.

Brainwarp boosts refresh rate, reduces latency and decreases GPU pressure for a smoother VR experience.

How does adding latency reduce latency?