Virtual Desktop vs Cable Link (Quest), something new?

gonvise · February 2

I've read your post about VD, and really, there's nothing I'd like more than VD giving me the same (or better) quality and fluidity that the Quest Pro gives me with Cable Link. This past year I've already tried twice VD and try to stick with it, in May and October, with an Asus Wifi 6 dedicated router, 4090, Questo Pro, 7800x3D, but neither time (different versions of VD) have I managed to get close to the quality mentioned. There have been other posts about it, and the conclusion reached is that yes, VD is very good software, and with normal configurations it can improve Cable Link (perhaps, or so you say). But with SS/PD levels, like in my case, x2, it's impossible for me to reach the same quality and fluidity as with cable link.

In your comment, I see that you talk about PD at 1.1 or 1.2 (AKA Super Sampling), but I'm talking about 2.0 via Oculus Tool

With that level of SS, the Quest Pro looks almost like on a HD monitor, and with the meta ASW, I just have to keep it at 45 to have a smooth experience.

Can you tell me your experience and how you have configured VD to work better than the link cable with that super sample (2.0)?

And please, from the other post, this is not a war, I really want to be able to have the best configuration I can have, and "I want to believe" in VD ;-), but I have not achieved it (Neither I nor many other people who want to use SS levels > 1.5), using the VD "god like" and raising the bitrate, as you say, to 900, but...:

1. VD reprojection is not as good as meta's ASW.
2. No matter how much I increase the encoding speed, it is always somewhat more bluried than with cable link.
3. Stutters, and more stutters in VD.

4. Quad Views doesn't work in VD (has this changed?) in Quest Pro (with eye tracking) Quad Views are magical, and the performance increment is very significant, and allows me to raise the SS to values like 2.

All the best!

Phantom711 · February 2

vor 25 Minuten schrieb gonvise:

4. Quad Views doesn't work in VD (has this changed?) in Quest Pro (with eye tracking) Quad Views are magical, and the performance increment is very significant

QVFR definitely works with VD, and has been doing so for a while now. You have to enable eyetracking in VD as well and if you have a custom QVFR settings file, make a setting for [VirtualDesktopXR].

gonvise · February 2

12 minutes ago, Phantom711 said:

QVFR definitely works with VD, and has been doing so for a while now. You have to enable eyetracking in VD as well and if you have a custom QVFR settings file, make a setting for [VirtualDesktopXR].

Ok, and the reprojection has improved? it is comparable to ASW?

Edited February 2 by gonvise

Qcumber · February 2

5 minutes ago, gonvise said:

Ok, and the reprojection has improved? it is comparable to ASW?

Yes and no. SSW us better than it used to be but not as good as ASW. However, VD is better at managing FPS below target refresh rate than Meta (subjective opinion).

Fiztex · February 2

23 минуты назад, gonvise сказал:

Ok, and the reprojection has improved? it is comparable to ASW?

What are your CPU+GPU and what and where do you fly? I'd say if you want ASW then for sure forget VD as it's much inferior in that regard, but I would be surprised if you actuallly need repro. I.e. I really highly doubt you need reprojection now with DLSS 4 to fly 72 Hz. I don't with my 4070 Super. Unless maybe if I decide to run a Supercarrier with a full deck and shadows, but I think that would drop even a 4090 into lower frames. But that is more about ED "optimization" than settings - IMO no point targeting full frames there if we end up flying with 50% GPU load later...

If you've got a 4090 try VDXR Godlike, DCS PD 1.2, DLAA with profile J or K without DLSS. You can find my QV settings in my signature, as well as the rest of instructions. They are in Russian, but screens are in English and I bet a translator will manage to do well for those basic instructions. Or you can also contact biotech so he can make some screens of settings for you - he's flying a Quest Pro with 4090 now with DLSS 4 and VDXR through ethernet adapter:

You can save your DCS settings as a template in DCS so you won't loose yours if decide to roll back. And stuff like enable/diable QV and switch between Quest and VD is easy, so should be no problem.

Edited February 2 by Fiztex

gonvise · February 2

29 minutes ago, Fiztex said:

What are your CPU+GPU and what and where do you fly? I'd say if you want ASW then for sure forget VD as it's much inferior in that regard, but I would be surprised if you actuallly need repro. I.e. I really highly doubt you need reprojection now with DLSS 4 to fly 72 Hz. I don't with my 4070 Super. Unless maybe if I decide to run a Supercarrier with a full deck and shadows, but I think that would drop even a 4090 into lower frames. But that is more about ED "optimization" than settings - IMO no point targeting full frames there if we end up flying with 50% GPU load later...

If you've got a 4090 try VDXR Godlike, DCS PD 1.2, DLAA with profile J or K without DLSS. You can find my QV settings in my signature, as well as the rest of instructions. They are in Russian, but screens are in English and I bet a translator will manage to do well for those basic instructions. Or you can also contact biotech so he can make some screens of settings for you - he's flying a Quest Pro with 4090 now with DLSS 4 and VDXR through ethernet adapter:

You can save your DCS settings as a template in DCS so you won't loose yours if decide to roll back. And stuff like enable/diable QV and switch between Quest and VD is easy, so should be no problem.

Sorry, but in your last two posts it doesn't seem like you're reading me!! ;-)... I've already stated the CPU and GPU I have, I've already said twice that my SS/PD is 2, I don't even want to try 1.2. With this supersampling (2.0), I need reprojection, not even with a 4090 would I get 90FPS in most scenarios, which is what I'm looking for. 72 FPS, simply, my eyes notice it and I perceive the screen slightly flickering. It's clear that with your configuration, I don't doubt that you have proven that VD works better for you than cable link, but I'm telling you that with SS/PD 2.0 and reprojection, my tests in VD were a sea of stutters compared to cable link. Understand me, we're talking about different configurations and expectations.

Qcumber · February 2

16 minutes ago, gonvise said:

I've already said twice that my SS/PD is 2

What is your final resolution? Quoting multipliers can be confusing. For example, I start at 2112 in Meta and apply 1.6 for a total of 3379.

Fiztex · February 2

44 минуты назад, gonvise сказал:

Sorry, but in your last two posts it doesn't seem like you're reading me!! ;-)... I've already stated the CPU and GPU I have, I've already said twice that my SS/PD is 2, I don't even want to try 1.2. With this supersampling (2.0), I need reprojection, not even with a 4090 would I get 90FPS in most scenarios, which is what I'm looking for. 72 FPS, simply, my eyes notice it and I perceive the screen slightly flickering. It's clear that with your configuration, I don't doubt that you have proven that VD works better for you than cable link, but I'm telling you that with SS/PD 2.0 and reprojection, my tests in VD were a sea of stutters compared to cable link. Understand me, we're talking about different configurations and expectations.

My bad you need to target 3500 resolution before foveating no matter with Link or VD, then on top you'll put DLAA which will supersample further and you would probably end up with a similar PD as you have now but for a fraction of the frame time. You can also go 60 with SSW at 120 Hz which maybe better for your eyes than 45 ASW. I fly 80 full normally and only for really heavy stuff switch to 72.

Tomorrow I'll get a 4090 and update my settings. I have no doubt I'm keeping DLAA though, it's excellent with QV.

Edited February 2 by Fiztex

nikoel · February 3

Yeah boys, I concur. You’re better off going into the Quadview Log (the one inside the AppData folder) and quoting end resolutions. This “multiply by 2, divide by 3, and then square” BS gets old very quickly, not to mention that big elephant in the room of encoding image you're actually sending to the headset (or think you're sending to the headset, more on this in a second)

@gonvise, I’m just going to state this bluntly—if you think you’re running a 2x supersample by showing the Oculus settings page but not showing the end resolution in the quadviews log, it’s easier to say that you’re not and this is not how resolution and encoding works

I strongly believe it’s a matter of taste, and both have their place depending on what the end user wants and what they are sensitive to. Additionally, no two systems are the same, and you can indeed have better experiences with one or the other.

The min/max setup you’re running and whatever absurd supersampling you’re pushing is honestly hilarious. I’m a firm believer that friends don’t let friends use reprojection, but at that stage, I don’t care how much supersampling you’re cranking up—your fake frame injection is absolutely wrecking your image. It’s like splurging on the latest Sony A7 with top-tier lenses and then smearing Vaseline all over the sensor—an iPhone would give you a better result. ASW, in my view, makes fake frame injection so jarring and distracting that I’d gladly take a full 1.0 native image with everything set to medium over that mess. It almost completely nullifies any meaningful discussions about DLSS 4, ghosting, or image quality because of how much it butchers the visual experience. If I had to put it in perspective, sure, having a pimple and some dark circles might hurt your self-esteem, but maybe deal with the black eye and missing teeth first. That said, I’ll admit this is a strong personal preference, and you might see things differently

As others have pointed out, while the numbers in Oculus settings do go higher, for some, the end image is no better and is often worse. This has led people in other forums to hypothesize that larger numbers do not always properly apply—but they do give you a nice fuzzy feeling because you see a higher number in the settings. For many the difference between HVEC @150 and Cable Link at almost 1K is so minuscule that it makes me believe those people have a point

To me running higher DCS settings (everything to the right) brings more meaningful increases in fidelity that are undeniable compared to supersampling beyond ~1.3 due to lens effects

Either way, we know for certain that some settings inside the debug tool are ignored by the Oculus Runtime. As an example, set it to ignore the proximity sensor and see what happens

Good luck with your quest

Edited February 3 by nikoel

gonvise · February 3

7 hours ago, nikoel said:

The min/max setup you’re running and whatever absurd supersampling you’re pushing is honestly hilarious. I’m a firm believer that friends don’t let friends use reprojection, but at that stage, I don’t care how much supersampling you’re cranking up—your fake frame injection is absolutely wrecking your image. It’s like splurging on the latest Sony A7 with top-tier lenses and then smearing Vaseline all over the sensor—an iPhone would give you a better result. ASW, in my view, makes fake frame injection so jarring and distracting that I’d gladly take a full 1.0 native image with everything set to medium over that mess.

It's very difficult to get anything positive out of these types of posts when you are able to say things like that without even having tested it. Do you think I'm putting those values because I'm a teenager (with all due respect to my teenage son ;-)) who has a 4090 and I want to go around saying that I put 2 for supersampling? I'm an engineer with more than 30 years of experience, in hardware and software, and I assure you that I spend more hours testing this type of things than flying the simulator itself. If I say that my supersampling value is 2.0, it's because I've tested 1.9 and I notice a worse difference. If I say that, currently, I hardly notice ghosting with ASW, it's because it is like that, since it keeps improving with each version (although others think it isn't like that). I also "thought" that putting 2 was an outrage, but when I put it, I almost cried, the image quality that it achieves with the lenses of the Questo Pro is incredible. I'm attaching my queadview.log if you want us to start the typical resolution discussion to see "who has the biggest one", but for me it's the opposite, it's a matter of what we see on the screen. I understand that if you think I'm crazy for putting that SS, you've tried it, right? It's very simple; I put 2.0 in the ODT because the slider in the meta application itself only lets me put 1.5. And I've tried 1.6, 1.7, 1.8 and 1.9... and the best quality that my eyes perceive is 2. That's it. Of course, then I put ASW, because with that SS, even with a 4090 it's impossible to maintain 90FPS. And I use ASW, because, nowadays, it's almost magical, and combined with the new DLSS 4 (I only use DLAA), I've even realized that the little ghosting I had was caused by DLSS 3 DLAA. These posts always end the same. Lessons about resolutions and why "mine is better than everyone else's". But I've tried everything you say, I've tried VD, I've tried lowering SS, I've tried with and without ASW, and if you tell me to try something that can improve what I have, I will do it, I won't go around saying things like my SS 2.0 is "absurd and hilarious" or that ASW reprojection is wrecking my image, because my visual perception say it's not, and if you say that it's simply because you likely haven't tried to test it or simply the quest pro lenses are superior to the Quest 3.

Quad-Views-Foveated.log

Edited February 3 by gonvise

nikoel · February 3

1 minute ago, gonvise said:

It's very difficult to get anything positive out of these types of posts when you are able to say things like that without even having tested it. Do you think I'm putting those values because I'm a teenager (with all due respect to my teenage son ;-)) who has a 4090 and I want to go around saying that I put 2 for supersampling? I'm an engineer with more than 30 years of experience, in hardware and software, and I assure you that I spend more hours testing this type of things than flying the simulator itself. If I say that my supersampling value is 2.0, it's because I've tested 1.9 and I notice a worse difference. If I say that, currently, I hardly notice ghosting with ASW, it's because it is like that, since it keeps improving with each version (although others think it isn't like that). I also "thought" that putting 2 was an outrage, but when I put it, I almost cried, the image quality that it achieves with the lenses of the Questo Pro is incredible. I'm attaching my queadview.log if you want us to start the typical resolution discussion to see "who has the biggest one", but for me it's the opposite, it's a matter of what we see on the screen. I understand that if you think I'm crazy for putting that SS, you've tried it, right? It's very simple; I put 2.0 in the ODT because the slider in the meta application itself only lets me put 1.5. And I've tried 1.6, 1.7, 1.8 and 1.9... and the best quality that my eyes perceive is 2. That's it. Of course, then I put ASW, because with that SS, even with a 4090 it's impossible to maintain 90FPS. And I use ASW, because, nowadays, it's almost magical, and combined with the new DLSS 4 (I only use DLAA), I've even realized that the little ghosting I had was caused by DLSS 3 DLAA. These posts always end the same. Lessons about resolutions and why "mine is better than everyone else's". But I've tried everything you say, I've tried VD, I've tried lowering SS, I've tried with and without ASW, and if you tell me to try something that can improve what I have, I will do it, I won't go around saying things like my SS 2.0 is "hilarious" or that ASW reprojection is wrecking my image, because my visual perception say it's not, and if you say that it's simply because you likely haven't tried to test it or simply the quest pro lenses are superior to the Quest 3.

Good for you. I have tried it and this is how I formed my opinion. It looks like you have yours and I am glad you are getting a superior experience with compromises that work best for you

gonvise · February 3

3 minutes ago, nikoel said:

Good for you. I have tried it and this is how I formed my opinion. It looks like you have yours and I am glad you are getting a superior experience with compromises that work best for you

Well, you could have started there ;-)... and it's not that I have my opinion, I don't want to be right or wrong, if you can advise me something that improves performance or even quality, I'll be happy to try it and even more happy that it works. That's why I opened this post.

Fiztex · February 3

2 часа назад, gonvise сказал:

Well, you could have started there ;-)... and it's not that I have my opinion, I don't want to be right or wrong, if you can advise me something that improves performance or even quality, I'll be happy to try it and even more happy that it works. That's why I opened this post.

Log says your initial resolution before QV was ~4000x4000. If you take Godlike VDXR and apply 1.2 - 1.3 PD on top of it (e.g. in DCS) then it will be around 4000x4000 as well - I literally suggested you that but you dismissed me just because you didn't like a particular multiplier...

And then you can lower periferal till 0.25 and boost focus further to 1.1 with QV and new DLSS 4 J/K. And on top new DLAA will supersample even further. You seem to be doing all that with the link, but now with QV you are still rendering 12.5 MP which is way too many and IMHO can be optimized further, especially for displays of Quest Pro. E.g. my settings in QV for fixed foveated target full resolution in the stereo overlap area of Quest 3 and somehow give reduction of 67% while yours are at 60% with a much smaller focus area. I'm thinking you probably miss out on not using vertical offset and have yor vertical FOV in not optimal position, though maybe it's just perfect by default for Pro, who knows. Definitely not for Quest 3 though, focus view is shifted up a lot there, so I'm just suprised about you not using this.

But hell, if you like your settings then this discussion for sure doesn't make much sense, we are not here to convince you to do anything. Repro is worse on VD, so stay off it for sure if you like repro. BTW my 4090 has arrived, will test this all in the evening and see if I can bring everything to ~9-10 MP rendered which by my estimation should be great for a 4090 without repro.

Edited February 3 by Fiztex

gonvise · February 3

I like my current setup, but I have no problem changing them, in fact, I'm always trying and trying, I spend my life like this ;-), and I'm open to anything that improves quality and performance...

I don't want to convince myself or anyone else, just prove that it's what I perceive better DCS with.

To start with, as you say, comparing the Quad View configuration on a Quest Pro, which really has eye tracking, with a Quest 3, which doesn't, can lead us into difficult comparisons. I don't need vertical offset, with that QVFR configuration and eye tracking, I simply don't perceive the transition from the center foveated area to the peripheral area.
Also as you say, if the reprojection is not as good as ASW, for me it's a negative point, but of course I will try everything you say as soon as I can.

Edited February 3 by gonvise

Qcumber · February 3

3 hours ago, gonvise said:

If I say that my supersampling value is 2.0,

For reference it would be useful to know your final resolution rather than the multiplier so we can compare it more meaningfully. Apologies if you have already said this but there is a lot of dialogue. What is your initial resolution before applying any SS?

nikoel · February 3

2 minutes ago, Qcumber said:

For reference it would be useful to know your final resolution rather than the multiplier so we can compare it more meaningfully. Apologies if you have already said this but there is a lot of dialogue. What is your initial resolution before applying any SS?

He is running 12.5 Megapixels [final resolution] brochachi

Edited February 3 by nikoel

gonvise · February 3

56 minutes ago, Qcumber said:

For reference it would be useful to know your final resolution rather than the multiplier so we can compare it more meaningfully. Apologies if you have already said this but there is a lot of dialogue. What is your initial resolution before applying any SS?

Quest Pro (with eye tracking), mbuchia QVFR, DLAA (with last DLSS 4 dll and preset K)

Fiztex · February 3

7 часов назад, gonvise сказал:

To start with, as you say, comparing the Quad View configuration on a Quest Pro, which really has eye tracking, with a Quest 3, which doesn't, can lead us into difficult comparisons. I don't need vertical offset, with that QVFR configuration and eye tracking, I simply don't perceive the transition from the center foveated area to the peripheral area.

That's exactly my point. I have better pixel reduction with Quest 3 than you with Pro, which doesn't make any sense. It should be the opposite - I'd expect you to end up with >70% savings with Pro while I can expect to be around 60-65% with Quest 3 and not perceive the difference as well. You should be able to reduce peripheral multiplier much further down with new DLAA and not perceive the difference, at least in daylight conditions. For starters try 0.3 and see if you can notice ultra clouds shimmer. Also not sure 0.2 smoothening is necessary with new DLAA, I'd try 0.1.

And I see in your log that your vertical focus offset is not working even though you set it in the app - that was another point of mine. I've ditched the app and been adjusting the file directly for a long time already, the app is missing parameters for focus offset.

Edited February 3 by Fiztex

Qcumber · February 3

2 hours ago, gonvise said:

Quest Pro (with eye tracking), mbuchia QVFR, DLAA (with last DLSS 4 dll and preset K)

OK. So with 1888 baseline X2 is 3776 pixels. My final output is about 3400 so very similar and not particularly excessive.

To get the same resolution in VD you need to supersample by 1.15. Try adding this in via QVFR to the centre, or even as high as 1.3. Combined with DLSS 4 it should give you nice visuals.

Edit: have you tried disabling link sharpening or using "quality" instead of "normal". The sharpening is way too much. That might make meta look "better" than VD.

Edited February 3 by Qcumber

Fiztex · February 3

Here is what I ended up with on a 4090 to fly A10C2 in Caucasus:

2025-02-03 20:33:57 +0100: Found option 'smoothen_focus_view_edges=0.1'
2025-02-03 20:33:57 +0100: Found option 'sharpen_focus_view=0'
2025-02-03 20:33:57 +0100: Found option 'horizontal_fixed_section=0.5'
2025-02-03 20:33:57 +0100: Found option 'vertical_fixed_section=0.6'
2025-02-03 20:33:57 +0100: Found option 'vertical_fixed_offset=-0.25'
2025-02-03 20:33:57 +0100: Found option 'peripheral_multiplier=0.20'
2025-02-03 20:33:57 +0100: Found option 'focus_multiplier=1.0'
2025-02-03 20:33:57 +0100: Found option 'turbo_mode=1'
2025-02-03 20:33:57 +0100: Eye tracking is not supported
2025-02-03 20:33:57 +0100: Recommended peripheral resolution: 614x644 (0.200x density)
2025-02-03 20:33:57 +0100: Recommended focus resolution: 1536x1930 (1.000x density)
2025-02-03 20:33:57 +0100: Stereo pixel count was: 19759104 (3072x3216)
2025-02-03 20:33:57 +0100: Quad views pixel count is: 6719792
2025-02-03 20:33:57 +0100: Savings: -66.0%
2025-02-03 20:34:03 +0100: Session is using quad views
2025-02-03 20:34:03 +0100: Edge smoothing: 0.10
2025-02-03 20:34:03 +0100: Sharpening: Disabled
2025-02-03 20:34:03 +0100: Turbo: Enabled

Runs Quest 3 at 80 Hz VDXR Godlike, DCS PD 1.1.

So as you can see only 6.7 MP + DLAA without DLSS with profile J pretty much maxes out Quest 3 optical resolution capability. I have zero clue why would one need 12 MP on Quest Pro with DFR... Sure if I need to run something heavy I can still drop PD one step, reduce visibility to high and switch to 72 Hz. Or even switch to DLSS with some low flying helicopters. But I definitely don't need ASW/SSW...

johnbowzer · February 4

I have been experimenting with the Q3 for months on and off with both oculus link usb, and vdxr wireless. And numerous multipliers, etc, like mentioned here, on both platforms. One thing that I get with vdxr no matter what, is airplane ghosting. Any fast mover going by, even on its six when it banks away hard, its like a second bandit in tow. Terrible. And this with other modern jets, Korean era and WW2 era bandits. I get very little, mostly none of that with oculus link. No different using the new DLSS4 and new driver. I retried various tests after setting up DLSS4 but same thing, and I figured that would not make a difference.

Qcumber · February 4

6 hours ago, johnbowzer said:

I have been experimenting with the Q3 for months on and off with both oculus link usb, and vdxr wireless. And numerous multipliers, etc, like mentioned here, on both platforms. One thing that I get with vdxr no matter what, is airplane ghosting. Any fast mover going by, even on its six when it banks away hard, its like a second bandit in tow. Terrible. And this with other modern jets, Korean era and WW2 era bandits. I get very little, mostly none of that with oculus link. No different using the new DLSS4 and new driver. I retried various tests after setting up DLSS4 but same thing, and I figured that would not make a difference.

What is your refresh rate and FPS? Are you using reprojection?

gonvise · February 4

Thanks @Fiztex, but again the same results with VD. Quest Pro, at 90 Hz, VDXR Godlike (Via dedicated wifi-6 router), DCS PD 1.3. Same configuration as I already mentioned in QVFR, DLAA (DLSS 4).

- Visual quality is worse than with cable link, I have tried several speeds and I always notice some compression. I have also tried to increase the PD.

- Similar performance talking in "FPS", but taking into account that with the quest pro I notice the flickering of 70Hz, I have to put 90Hz, so although I have between 60-80 FPS, without reprojection I can't stand the lack of fluidity in view rotation at those FPS, and with VD reprojection... I better not talk about it, it is inferior to ASW from my perception, with ASW the movement of my head is practically natural (like in life itself), without stutters and with very little ghosting, only perceptible in some very fast scenes or sudden head turns.

Overall, although I could lower my expectations a bit in terms of quality, I can't get anywhere near the fluidity that ASW gives me.

Again, I've tried, as I told you, I bought a wifi 6E router just for this, but in the end each of us has our own perceptions, and yet, no matter how much I turn up everything, I still can't get the image quality and fluidity that I have with a cable.

The point is that each one uses what works best for them. Thanks!

johnbowzer · February 5

18 hours ago, Qcumber said:

What is your refresh rate and FPS? Are you using reprojection?

Tried both 72 and 80 in both link and vdxr, and can get full or almost full fps without repro, and tried with and without repro. I get the ghosting in vdxr always. And it is looking out front, or to the side at other planes. And it seems to be a DCS thing. No ghosting in MSFS2020 with VDXR. I set up a scenario to get a fast pass with another plane and no ghosting.

I realize there are many variables we deal with, but I have tried many things to isolate. But no joy to this point.

Qcumber · February 5

4 hours ago, johnbowzer said:

Tried both 72 and 80 in both link and vdxr, and can get full or almost full fps without repro, and tried with and without repro. I get the ghosting in vdxr always. And it is looking out front, or to the side at other planes. And it seems to be a DCS thing. No ghosting in MSFS2020 with VDXR. I set up a scenario to get a fast pass with another plane and no ghosting.

I realize there are many variables we deal with, but I have tried many things to isolate. But no joy to this point.

Do you get any ghosting if AA is off?

Sign In

Virtual Desktop vs Cable Link (Quest), something new?

Recommended Posts

gonvise

Phantom711

gonvise

Qcumber

Fiztex

gonvise

Qcumber

Fiztex

nikoel

gonvise

nikoel

gonvise

Fiztex

gonvise

Qcumber

nikoel

gonvise

Fiztex

Qcumber

Fiztex

johnbowzer

Qcumber

gonvise

johnbowzer

Qcumber

Recently Browsing 0 members