1 hour ago, robert young said:

Well I have just been delivered an 8700k with a 1080ti and if that can't run things half decently I might as well give up! I sometimes use up sampling with a lowly i2700 and 1050ti and if I'm sensible even that runs pretty well. Here's hoping.

It is an ideal setup for VR in P3D 4.4 it will be perfect for Rift, Vive, but may not be enough for the Pimax 5K+. No-one yet has really tried out that headset that I know of in P3D 4.4. The 5K+ would be ideal for flightsim with the 200deg field of view.

I have been using 1.6 super-sampling on my Rift with great performance, which totals to 3.456.000 pixels per eye being rendered (if I am not mistaken). The Pimax 5K has about 3.686.400 pixels per eye, so it should not be dramatically slower.

1 hour ago, GCBraun said:

I have been using 1.6 super-sampling on my Rift with great performance, which totals to 3.456.000 pixels per eye being rendered (if I am not mistaken). The Pimax 5K has about 3.686.400 pixels per eye, so it should not be dramatically slower.

Thanks hadn't thought of it like that but you have got a point.

Could you back up my maths?

Rift CV1 = 1080 x 1200 pixels per eye = 1,296,000 pixels

Supersampling 1.7 is pretty comfortable for a 2080 GTX so:

1,296,000 x 1.7^2 = 3,745,440 pixels per eye rendered by the GPU internally.

Pimax 5K+ = 2560 x 1440 pixels per eye = 3,686,400 pixels

So the question is:

Rift CV1 on supersampling 1.7 approximately = Pimax 5K?

I think one issue is that the Pimax 5K needs super sampling as well because the pixels are spread out over a wider area.

From what I remember, reviewers of the Pimax 5K were adding super sampling.

So super sampling a Pimax 5K could potentially stress out a 2080ti.

13 hours ago, robert young said:

Has anyone else tried using the native 3d VR images in P3d as a means to use eyes only?

I think this is doomed to fail. The reason is clear if you try to view stereo videos on YouTube with the naked eye, which is only possible if you show them very small. As far as I can tell the images need to no further apart than your interpupillary distance, in which case the images will merge when you learn to relax your eye muscles while still focusing. With the naked eye, this means the images would overlap if they were wider than your IPD. By contrast, in a HMD such as the Rift, the images are fed independently to each eye by lenses, which means they can be much larger without overlapping. 

In principle it is possible to view stereo images cross-eyed, in which case you can use somewhat larger images, but not very much larger. It's also probably not safe (and certainly not comfortable) to do this for very long. To do this without a HMD would need to find some other way to feed separate images to each eye.

I think this is doomed to fail. 

I think I more or less conceded that after opening the discussion, which was really an exercise in speculation rather than a theory. However, as discussed, there is no evidence that parallel or "crossed eyed" viewing does any harm, any more than exercising other muscles does much harm, within reason.

3 hours ago, robert young said:

I think this is doomed to fail. 

I think I more or less conceded that after opening the discussion, which was really an exercise in speculation rather than a theory. However, as discussed, there is no evidence that parallel or "crossed eyed" viewing does any harm, any more than exercising other muscles does much harm, within reason.

Not sure what this has to do with what I said.

Edited January 8, 20197 yr by MarkDH

If my maths is right, a 2080 will handle a Pimax5K very well without any supersampling.

But a 2080ti would only be able to handle a Pimax5K at about 1.2 supersampling before it too would start hitting its limits.

Edited January 8, 20197 yr by glider1