Ultra non-scientific benchmark of beat saber

[Solaris17]

On a machine that barely qualifies as a "VR" machine.

So built myself a VR room. Mostly for friends and relatives. The unit itself is mega basic.

AMD Ryzen 2600 (non X variant)
16GB of the cheapest DDR4 money could buy
and a Radeon RX570 8GB
some asus board

Its via an HTC VIVE (non-pro)

It runs windows 10 home. The drive is a 250gb samsung 970 evo

Now while this isnt super scientific the system requirement of Beat Saber according to steam

The system itself is connected to a 55" hisense 4k TV (3840 x 2160) at 60hz

I played for about 25min ( all GPU settings in game on highest value) while logging with GPU-Z and then watched the girls play a few songs (not logged) consecutively. I used the log tool @TomWoB made to graph out my results. https://www.techpowerup.com/forums/threads/logviewer-for-gpu-z-available.185299/

The log file is attached. Nothing was overclocked.

The graph read out is below.

I am only really making this thread because I don't really "pay attention" to VR. To be completely honest I took a chance with buying all this stuff (months ago prices were a little higher.) I didnt really know what "made a VR PC". I just decided to take some time to poll it since I was curious what the stress of a 4k TV AND the offload to the HMD would do to mid range equipment.

While I know (again) that this wasnt even properly done (sorry couldn't be assed to do it like my other guides) I did want to throw up some preliminary data mixed heavily with salt from my optic nerves. Since I managed to experience both and know the TV is at 60hz I experienced no hitching.

This obviously isnt say Hellblade: Senuas Sacrifice for VR. It is however a party or "group" game. as such other "group" games have similar system requirements and while VR engines do differ, to the best of my knowledge there arent alot of different VR engines as of right now being used. It might be safe to say these games would perform "similarly" with a similar Pc configuration.

At any rate its not like this forum gets alot of activity, and I am very very interested in the performance aspects of VR games since media and marketing coverage of GPU/CPUs indicates things as "VR ready" but what is "VR ready"? and how much $$ do you need to spend to have a good experience?

So far, in my testing not alot.

 

[Deleted member 50521]

Beat saber is not very demanding TBH. Try Moss or FO4VR and you will see what I mean.

 

[Solaris17]

Beat saber is not very demanding TBH. Try Moss or FO4VR and you will see what I mean.

I mentioned that. Unfortunately those two games I would not consider group or party games. In my case VR is only used when people are over. No real drive to ever use it for a single player experience.

 

[Mindweaver]

Nice! That's great information for users with lower systems. The Quest is coming out soon and is an AIO VR HMD and you can play Beat Saber on it and maybe a better entry system for some users without your know how, buddy. I feel FO4VR would play ok on your system after applying some mods and using teleport over smooth movement. My gtx970 played it fine after a bunch of mods.. lol The texture optimization mod was the biggest improvement. I would be interested to hear about who other games perform.

My 2600k is really getting long in the tooth and I'm starting to notice it. I'm thinking about pulling the trigger on a cheap Ryzen 5 1600 right now and then dropping the new Ryzen 5 3600 when they release. I know it would probably be a lateral move, but having everything set to just drop in a new CPU would be nice. Thanks for sharing! I'm hoping this section starts to take off with all of the new HMD's being announce right now.

 

[TomWoB]

Hi Solaris17,

when you "play a little with the width" of my LogViewer, you can force "a single line time-axis", which allows to show bigger graphs:

Another tip is to put same units, e.g. temperature graphs, "together in one diagram":

Anyway ... nice test !

Regards
Tom