One could believe that this is their plan - but the Rift will not attract new customers for much longer if others release truly better headsets within a reasonable price point, i.e. a fair bit better than say the Vive Pro in a couple of aspects like the lenses but at a consumer price. So early 2020 I cannot see anybody who is willing to dish out a couple of hundred dollars to buy the good ol‘ Rift just because of its nice platform software. So why go through all of the pain & investment of building this platform, if you‘d lose all the acquired customers between 2020-2022/3 until your new killer headset is released ? The sane thing to do is release at least a facelift with Vive Pro resolution & Go lenses.
3 Likes
OK so when and in what version of Pitool was the compatibility option added?
1 Like
The one backer/dev at the Berlin Meet touched on an issue with Curved displays & rendering. This is likely why Oculus is going slow to figure a workaround.
Long ago during Testing.
1 Like
Like the initial testing from over a year ago when the controllers will still wired or when the units when out to the testers?
Testing as in M1 testing. The demos (v prototypes) ran off of commanline (like dos).
OK so June. That isn’t that long ago, I would still call that new personally.
But it’s compatability mode. The new render however is not selectable it just works if you have a 20series card.
What I mean is the only honest way to describe what is behind this “parallel projection” stuff (compatibility mode) is to say any HMD using angled displays (like the pimax we are discussing here) will suffer a big handicap when playing games that are not based on unity or on UE >4.18 or oculus >1.17.
With all those games (it will of course get better over time but for now I guess they are many, probably the majority of existing VR games) my understanding is a lot of the GPU power is wasted just to make them compatible with HMDs using angled displays.
For me this information has just been hidden so far by pimax. I wonder for how long pimax has been aware of this angled displays handicap with many games. And even now they are disclosing it they don’t do it in a honest way, instead of acknowledging a handicap (= wasted GPU power) with many games for HMDs using angled displays, they try to present it as a “performance gain” for those games able to manage angled displays natively.
I have just made the connection with this “performance improvement” thread after reading today the post I quoted from sgallouet. That’s what has made things clear for me.
At the time of the KS I was making extrapolations based on the SS level I was able to reach on the rift with my GPU (my benchmark being 90fps stable without ASW) and the input resolution for the pimax.
Then came all the testing from the youtubers and I found their fps were lower than what I expected based on my previous extrapolations. I was initially explaining this difference by the pitool software possibly being less efficient than oculus software (even without ASW, I have used my rift a long time under win7 which can’t benefit from ASW), but if all those games need a compatibility trick wasting GPU power because they cannot handle angled display natively then this performance lower than I expected makes full sense now.
But this has never been explained that straightforward by pimax. Even now in this thread. And that was the same when they were asked to explain why the 5K was performing better than the 8K with a 1070 GPU (late July update). Now we know why, but not because pimax explained it but because many of us together were able to guess it through the informations supplied by the youtubers.
I’m angered at pimax for never being straightforward with us. They don’t lie but they don’t tell the truth straigthforward neither. That was the case for the 8K refresh rate, for 8K vs 5K+, for the parallel projections discussed here, for the coupon restrictions, for the availability of base stations outside bundles, for the bakers shipments and ramp up of production, for brainwarp, and I could continue that list.
This is boring, we always have to investigate to discover what is hidden behind their twisted way to communicate with us. People working for Pimax need to learn to be straightforward with their customers (and de facto with us bakers too). I believe the hit from bad news is always less when directly telling the truth, the pure facts, than when trying to hide the real facts that will sooner or later be discovered by the customers.
2 Likes
But compatibility mode is 30% SLOWER, pimax are claiming a “new” rendering mode that is 30% FASTER
1 Like
The Angled displays has been known since OpenVR(Valve added support for it). Not to sound condescending but any following this project would have some basic idea on this.
Many games didn’t use OpenVR or the like & implemented VR using their own workarounds. Now to be fair as I have tried to point out to others we should never expect ppl to know things just because they back things like this.
Ie lay user whom wants pickup & play, buys the top hardware so they don’t need to tweak a bunch of settings & learn new things.
Once we ditched DOS & promoted plug n play & user friendly. We opened up the pc to the masses.
Right and I would tend to agree that their communication skills is probably the worst thing they have going for them. For instance we have no idea on how units have actually shipped or why many of the headsets that were supposed to already have shipped have no tracking numbers.
It does make sense as even with the new benchmark video from Sweviver, the 2080ti should be seeing a larger jump in performance than what was shown. The compatibility mode is likely affecting the frame rate.
1 Like
Gotcha apparently I’m slow today as I’ve been putting out fires at work all day.
On 20 series Carda.
So 2080 ti 4 example 30% faster on pimax with new render algorithm.
-
Parallel Projection (-30%) = 0 (but 30% than sayb1080 ti)
-
Without. +30% No PP +30% new algorithm= +60%approx gain vs say 1080 ti.
Ah you work for a firefighting company vs being an actual firefighter. Lol
1 Like
The IT firefight… It never ends. Today is downed email server and Backup project planning day, with a sprinkle of Xerox coming in a messing things up at another clients.
1 Like
The best inbound tech support involved flying a fellow I knew to help them turn a switch on. Lol
Oh there’s lots of that, I have to laugh because my company bills me out at almost $140/hr.
1 Like
I suppose VirtualLink will be a must have for all “VR Ready” gpu.
As for VRS, sure it’s probably the closest answer we have to fovated rendering at the moment, but being part of VR Works (which is strange since it could also benefit for non VR games), it is probably proprietary like most of their techs.
AMD will probably aim for a more generic implementation in DX.
I perfectly knew that, my post was just already long enough to not add unecessary details like that.
My point was this thread was to make look like a gain what in reality is a handicap (performance hit).
I will always promote objective and transparent facts above marketing speech only leading to confusion.
Btw, mixing that handicap from compatibility mode with the gain from 2080ti over 1080ti is another marketing trick to try to confuse people and dilute the bad news (handicap from compatibility mode) into “good” news (gain from nextgen GPU).
Looking at this the other way I could also make such statements (let call that the anti-marketing speech): the gain from a 2080ti over a 1080ti will just compensate for the hit from compatibility mode, and that will cost you 1500bucks. Or: a 1080ti without compatibility mode will run the same as a 2080ti with compatibility mode, for much less money.
That sounds much less appealing than “About 60-70% performance improved on RTX2080Ti” doesn’t it ?
Better to stick with objective unbiased facts:
-
2080ti is faster than 1080ti (and with current games probably most of the time “not that much faster considering the price difference”, see sweviver video for an insight about that in some VR games and make your own opinion).
-
compatibility mode should imply “about 30% performance hit”, whatever the GPU you use (this is based on the information from OP)
-
no compatibility mode is just normal situation (no GPU power “wasted” on the sole purpose of compatibility with non-natively supported angled displays)
-
as an exemple, most of the existing racing sims and flight sims will require compatibility mode. They are already GPU demanding games in VR (need SS to improve clarity at far distance), add the performance hit from compatibility mode on top of that.
-
one may wonder if the performance hit from compatibility mode would differ for CPU-limited games (like many existing flight sims for example)
1 Like