I’m very confused by Pimax’ most recent update, saying that the 5K performs better than the 8K.
Worth to mention that, with GTX 1070, 5K performs better than 8K. For those who are not planning to upgrade the graphics cards, we will provide you options to exchange your 8K for 5K + accessories.
If anything I would have expected the 8K to have better performance than the 5K.
Both headsets have a 2560×1440 input resolution.
Both headsets have a 200° FOV.
The 8K will have 80hz (or 85hz?), while the 5K will have 90hz (as far as I know).
So is there is something very wrong with the scalar? And even then I can’t imagine what GPU performance has to do with it.
Am I missing something here?
I´m also wondering about this statement.
The only difference I know ist the display.
5k = OLED / 8k = LCD
Am I right?
I’m not aware of that.
I assumed both used LCD panels, probably IPS based on reports of backlight-bleeding. But this wouldn’t make a difference for performance in terms of FPS either.
I am indeed very concerned about this as well. It makes no sense at all. The only reason why it would be like this is that the upscaler chip does in fact not work and the 8K is in fact what they marketed as 8K-X.
We lack the technical facts meaning it is all speculation but with the facts that we - the kickstarter backers have this is not possible.
I am not all doom and gloom. I got a Titan X pascal overclocked and watercooled. If that is enough to play most games and some wont - I wont be moaning.
I am however sick and tired of the pimax announcements. They are all vague, misleading or downright false information.
I am glad we got beta testers working hard and I am sure pimax works their ass off delivering a viable product. The way of communication with us is starting to get the better of me though.
IIRC, they had issues with getting the scaler to scale to 90 fps. I don’t know if that qualifies as very wrong, but in general you can scale fast or well (doing both is hard / expensive, also I’ve hated upscaling since it started showing up in DVD players years ago). It’ll certainly add some latency if nothing else. I’ll venture a guess that the scaler drops framerate by 5 or so. So if you’re sensitive to less than 90 fps then you’d be better off with a 5K?
If that would be the case though then the 8K would tax the GPU less than the 5k because the render target would be 82 FPS (82 is what they considered stable). The 8K being harder on the GPU than the 5k makes no sense at all unless there is no upscaling at work and the GPU renderes the full resolution (wich would be very hard on any GPU available).
Based on what we know the 5k should be more taxing. The headsets are meant to be pretty similar in terms of transferred resolution, but the 8k has a lower frame rate due to the scalar, but I am beginning to question whether this is actually really what is going on or if PiMax is actually doing something else and hasn’t explained things properly. All this statement really tells us is there is something clearly missing technically in the explanation that PiMax hasn’t detailed and they need to. Part of the problem is we are speculating on speculation in the absence of real facts.
I believe when using the 8K your PC renders at 6K or 8K downscales it to 5K sends it to the headset and then upscales it back to 8K on the headset.
Interesting and should give nicer image but why don’t they render the 5K at a higher resolution as well and downscale it to 5K to send over DP1.4 ?
This is the reply given in the update thread:
Please explain why the 5K would have better performance than the 8K with GTX1070?
A: rendering at 6K or 8K, then downscale to 5K, and finally 5K upscale to 8K.
B: rendering at 5K, then 5K upscale to 8K.
The visual quality with the rendering path A looks much better than B.
that’s why the supersampling works well & with 1070 graphics cards, 5K performs better than 8K.
Yeah, that still doesn’t make much sense. Doesn’t answer the question “why does the 5k perform better than the 8k”. In scenario A, you could do the same thing with the 5K minus the final upscale to 8K, which is done by the internal chip in the 8K, so unless that is slowing things down, I don’t know why it would give different performance. Same thing with scenario B, if you’re rendering both at 5K, why would the 5K perform any differently?
Even if I don’t understand why you need to downscale before upscaling for a better picture I am however very pleased that an answer was given.
The downscale (which is done by the GPU) is needed to be able to send via cable, as the bandwidth of the cable is too low.
It’s not really clear why it’s rendered at a higher resolution for the 8K than the 5K.
Yes, the only thing that doesn’t make sense to me is why the performance is different between the two. They’re essentially saying you can supersample which is then downsampled to a resolution that can be sent over the cable, but what doesn’t make sense is why the final upscale on the 8K would lower performance. That shouldn’t happen normally with a well designed scaler.
Apparently, the scaler chip (which is spec’ed to run at 90 Hz) cannot actually do that, once the “real world” is factored in.
It sounds like Pimax changes the supersample level, depending on whether you are using a 5K or an 8K, but that doesn’t make much sense. You should be able to configure your SS level to whatever level your GPU can handle. Personally, I want the highest SS I can get (with a good framerate), since Elite Dangerous has bad aliasing artifacts.
I think maybe its because of super sampling before ts down scaled
I wonder why we can’t choose our own SS value for the 8K, for example if we wanted to upscale from 5K what the issue would be? Obv higher SS will look better in the end, but already we know to SS according to our graphics card ability.
If I had to throw a guess, probably something to do with the variable Hz that will result and I still don’t quite understand that situation.
Both headsets are using clpl screens.
While I can’t say for sure on performance. My best guess.
5k No scaler(1 less step in getting pic to screen) & performs at 90hz
8k Has scaler (extra step) has same input res but only stable at 80hz. The scaler could be why not able to hit 90hz & as such will have less performance seen potenially.
80fps vs 90fps. A guess anyway.
The 8k-X as it is like the 5k (no scaler) should be able to do stable 90hz if your gpu setup csn support it.
Less gpu load doesn’t necessarily equal better performance. As tge scaler adds a step to the render process as mentioned potenially adds latency.
The backer meets & such should give more info.