Quest 2 Virtual Desktop Performance Thread

Have you the router setup as an access point? That helps reduce the latency.

Melmac, thanks for the response!

The router doesn't have an AP software setting, but the manufacturer had a tutorial that showed how to manually set it up. All it said to do was to disable DHCP server on the router, and then assign it an IP address on the main router's subnet. Assuming that turns it into an access point, that's what I did.

I used it with VD and seemed to be getting decent results.

However, I also posted the diagram in the VD discord, and the developer suggested that it would be best to put the dedicated router in router mode on it's own subnet, b/c he said all of my traffic was still going through the main router the way I had it set up there.

I reconfigured it to his suggestion, and this is an updated diagram, including where I moved the CAT5e to the WAN port. It seems to be working as intended like this, and I'm waiting for his response as to whether I have it set up correctly.

I played Alyx with it in this config, and was getting 80 FPS and 40-43ms... I dont know how good that is, but it looked/played good to me. Once I get confirmation the setup is correct, I can work on tweaking the performance a bit.

https://i.imgur.com/Misms8c.png
 
Melmac, thanks for the response!

The router doesn't have an AP software setting, but the manufacturer had a tutorial that showed how to manually set it up. All it said to do was to disable DHCP server on the router, and then assign it an IP address on the main router's subnet. Assuming that turns it into an access point, that's what I did.

I used it with VD and seemed to be getting decent results.

However, I also posted the diagram in the VD discord, and the developer suggested that it would be best to put the dedicated router in router mode on it's own subnet, b/c he said all of my traffic was still going through the main router the way I had it set up there.

I reconfigured it to his suggestion, and this is an updated diagram, including where I moved the CAT5e to the WAN port. It seems to be working as intended like this, and I'm waiting for his response as to whether I have it set up correctly.

I played Alyx with it in this config, and was getting 80 FPS and 40-43ms... I dont know how good that is, but it looked/played good to me. Once I get confirmation the setup is correct, I can work on tweaking the performance a bit.

https://i.imgur.com/Misms8c.png

What was your latency in Half Life Alyx with the previous setup?
 
What was your latency in Half Life Alyx with the previous setup?

Dont quite remember, b/c I wasn't really monitoring it when I had it set up in that config. Once I get more familiar with settings in VD and Alyx/Steam, I may test it both ways to see what difference there is, if any.

I've been searching/reading past discussions on the VD discord, and it seems that there is not a consensus on whether AP mode or router mode is preferred for the dedicated router. The dev recommends router mode, but many of the prominent discord members recommend AP mode. Based on the conversations I've read, and my limited knowledge of networking, it seems that some potential drawbacks of router mode are that devices on the dedicated router may not be able to communicate with devices on the primary router, and the possibility of a double NAT (whatever that is).

If I'm interpreting some of the developers old comments properly, it seems that his reasoning for preferring router mode is because in access point mode, packets going from the PC to the Quest have to go all the way back to the primary router, and then get routed back to the Quest. In router mode, they only have to go to the dedicated router, and then get routed straight out to the Quest.

So, assuming I'm understanding everything, if you have it in router mode it's a shorter distance that packets between PC-Quest have to travel. Whether that makes a meaningful difference, I dont know.

I also saw a comment from a user that said he talked to a "networking guy", and was told that even in access point mode, the AP will remember that the Quest is connected to it, and will send data coming from PC with destination Q2 directly to the Q2, without having to send it through the primary router first. If that's true, then it seems it would make router mode unnecessary, unless there are other benefits.
 
Dont quite remember, b/c I wasn't really monitoring it when I had it set up in that config. Once I get more familiar with settings in VD and Alyx/Steam, I may test it both ways to see what difference there is, if any.

I've been searching/reading past discussions on the VD discord, and it seems that there is not a consensus on whether AP mode or router mode is preferred for the dedicated router. The dev recommends router mode, but many of the prominent discord members recommend AP mode. Based on the conversations I've read, and my limited knowledge of networking, it seems that some potential drawbacks of router mode are that devices on the dedicated router may not be able to communicate with devices on the primary router, and the possibility of a double NAT (whatever that is).

If I'm interpreting some of the developers old comments properly, it seems that his reasoning for preferring router mode is because in access point mode, packets going from the PC to the Quest have to go all the way back to the primary router, and then get routed back to the Quest. In router mode, they only have to go to the dedicated router, and then get routed straight out to the Quest.

So, assuming I'm understanding everything, if you have it in router mode it's a shorter distance that packets between PC-Quest have to travel. Whether that makes a meaningful difference, I dont know.

I also saw a comment from a user that said he talked to a "networking guy", and was told that even in access point mode, the AP will remember that the Quest is connected to it, and will send data coming from PC with destination Q2 directly to the Q2, without having to send it through the primary router first. If that's true, then it seems it would make router mode unnecessary, unless there are other benefits.

Yes, I agree with said "networking guy" I believe the Virtual Desktop developer is incorrect in this. The traffic won't have to go back to the primary router. Once the Quest 2 Access point knows the "route" to the PC, any network packets will be sent to the PC from the Quest 2 without having to go back to the main router at all. The Less the access point/router has to do, the lower the latency should be. If you have it in Router mode, it's doing more work, whereas in Access point mode it's doing just the basics.

I tested mine in both and found access point mode quicker. That's not saying yours will be better in Access point. The Virtual desktop developer is always quick to point out that you should use what works best for you.
 
I would also say the investigation is pointless until you have a frame of reference for comparison and trying both confgurations.

It took me a long time to figure out what is the optimal setup for me. I have done it a bit differently where I have my Q2 Router (in AP mode/DHCP off) connected to my Super Hub 3 router (in Router mode), and the SH3 is connected to my PC.

I did try it the other way round (i.e. similar to yours) and found no difference.

melmac, I discovered something... when my guardian boundary appears, it seems to exhibit the "sllow down/juddering" phenomenon that I experienced recently.
When the guardian goes away, it seems to be less prominent.
 
melmac, I discovered something... when my guardian boundary appears, it seems to exhibit the "sllow down/juddering" phenomenon that I experienced recently.
When the guardian goes away, it seems to be less prominent.

Since you have developer mode, try disabling the Guardian and see what happens. Be careful though!! :p

Another thing you should try out. Users with 3xxx cards are noticing a lot less stuttering when they use H.265(HEVC) mode in the Streamer App. Reduce the VR bitrate until you get the same latency as using H.264.
 
Since you have developer mode, try disabling the Guardian and see what happens. Be careful though!! :p

Another thing you should try out. Users with 3xxx cards are noticing a lot less stuttering when they use H.265(HEVC) mode in the Streamer App. Reduce the VR bitrate until you get the same latency as using H.264.

Reguarding Guardian - Yep thats what I did but of I do need the Guardian, so what I did was reduce the sensitivity so it only triggers when I get very close.

Tbh its not bad at all, its just literally looking down with the microstuttering.

Regarding HEVC - I could have sworn tis adds another 10ms to the latency. I do know that video buffering does help as well.. but again more latency.
 
A router is made for handling data packages, windows isn't.

However, I read something on Reddit, with an attached video from uploadvr (I think) showing a way to overcome a lot of the issues when using a hotspot (bit of code and playing with some settings I believe). So it may be there is a way to make it work well.
 
The Virtual desktop developer is always quick to point out that you should use what works best for you.

Yes, I should clarify that he also recommends AP mode to people, depending on their situation.

I tested mine in both and found access point mode quicker. That's not saying yours will be better in Access point.

I tested mine both ways and more or less came to the same conclusion. I didn't do extensive testing, but I found almost no difference between the two modes when using Alyx. If I had to pick one, it seemed like AP mode might have been 1-3ms better, but that's barely a difference, and again, more extensive testing might show something else. For now, I left it in AP mode.

Here are the settings I used:

NVIDIA - low latency mode on

VD PC app:
Auto adjust bitrate off
h.264

VD streaming:
VR graphics - Medium
80 FPS
65 Bitrate
Sliced encoding on

Using these settings, I was getting 80 FPS and roughly 31-35 ms playing Alyx on medium. High was about the same (maybe 1-2 ms higher at times).

I didn't test Jet Island for both setups, but in AP mode I was getting 80 FPS and 30-33 ms. I did have to put the in game sampling to .7, as 1.0 was causing low FPS.

i5-10400, RTX 2060, 16GB ram, m.2 SSD.
 
Just had another quick go in propogation, but this time via virtual desktop. I'm guessing it's quite easy on the system, but it ran beautifully, so much more immersive without a wire, and a good test to see if I can use virtual desktop in my little garden office :D.

How do I get the overlay to see the details?
 
Just had another quick go in propogation, but this time via virtual desktop. I'm guessing it's quite easy on the system, but it ran beautifully, so much more immersive without a wire, and a good test to see if I can use virtual desktop in my little garden office :D.

How do I get the overlay to see the details?

The Overlay is in the streaming section of Virtual desktop menu. Just enable it. You don't have to quit out of Virtual desktop.
 
I finally found a solution to my problem.
What did it for me was to set channel bandwidth at 40mhz on a free dfs channel (112 in my case).
It runs perfectly. To begin with, do not just use any software to see if you are stuttering or not.
Half-Life Alyx is optimized like crap, as is Boneworks. Robo recall on the other hand is amazing in this regard to do your testing.

I am still stuttering in steam VR games but it is not related to the network. It is due to poor ampere drivers. The first ones introduced micro stuttering and it is still the case to this day, even if the last driver iteration was supposed to help (it clearly didn’t for me). Nvidia acknowledged it.
My computer is extremely optimized and built around low latency as well so I know it is not the problem.
Lowest Virtual Desktop settings also cause stuttering on Steam VR games whereas it is flawless on Robo recall (even highest settings are). If you don’t have an ampere card, rollback to previous drivers.
If you do and still have issues with the latest drivers, you have to wait more, unfortunately, or buy your games on the oculus store.
 
Last edited:
Ok, to be more precise, 80 MHz works even better than 40 MHz. If I turn around very quickly I can see the image creating itself on the side of the lenses if that makes sense. As if the streaming was trying to catch up.
At 80 MHz this effect disappears.
Wifi 6 also works better than wifi 5, as expected.
What was destroying the overall performance was to set my router in AP mode. Back to router mode, the experience is excellent, except in certain games.
 
My 5ghz router (80 mhz width) only supports 433 Mbps. Would I see a benefit by upgrading my router to one that is capable of 866 Mbps?

I've seen some people say that 433 is sufficient, and others say that there would be a bitrate/latency improvement by upgrading to 866. Thoughts?
 
Just had a quick browse of this thread and altered settings to:

VR graphics - High
90 FPS
65 Bitrate
Sliced encoding on

and has my first play of HL Alyx. The odd stutter but pretty much flawless. Graphics are incredible.
 
I'm tempted by the Oculus Quest 2 and am wondering what my chances are of getting acceptable wireless PC VR performance with my existing set up, as I already own a bunch of VR games on Steam and on Oculus for the Rift CV1. My PC is aging but my GTX 1080 should hopefully still be up to the task. The PC is connected to a BT Smart Hub 2 via an ethernet cable. Anyone know whether the odds are in my favour for wireless PC VR working in a room that is a couple of metres away from the BT Smart Hub 2 please?
 
I'm tempted by the Oculus Quest 2 and am wondering what my chances are of getting acceptable wireless PC VR performance with my existing set up, as I already own a bunch of VR games on Steam and on Oculus for the Rift CV1. My PC is aging but my GTX 1080 should hopefully still be up to the task. The PC is connected to a BT Smart Hub 2 via an ethernet cable. Anyone know whether the odds are in my favour for wireless PC VR working in a room that is a couple of metres away from the BT Smart Hub 2 please?

Is the PC connected to the Router by cable or is it wireless?
 
"The PC is connected to a BT Smart Hub 2 via an ethernet cable" :D

haha, sorry, DOH!! :p

It depends on how good a signal it gets, how many walls between them etc. How good your phone connects to the wifi might give you some indication of the how the Quest 2 will perform. There is also the overhead of your Hub been the router for the house.

Some people have had no problems with the Router been in a different room, others have had a terrible time.

But, even it if is terrible, it's a solvable problem. Just buy a cheap AC router to use as a dedicated access point for the Quest 2.
 
haha, sorry, DOH!! :p

It depends on how good a signal it gets, how many walls between them etc. How good your phone connects to the wifi might give you some indication of the how the Quest 2 will perform. There is also the overhead of your Hub been the router for the house.

Some people have had no problems with the Router been in a different room, others have had a terrible time.

But, even it if is terrible, it's a solvable problem. Just buy a cheap AC router to use as a dedicated access point for the Quest 2.

Cheers for that. Inching ever closer to making an purchase...
 
Back
Top Bottom