HTC Vive Pro Headset Review: A High Bar for Premium VR

HTC Vive Pro Performance Analysis

As noted previously, we ran all our tests using both our GTX 980 Ti, as well as the lesser GTX 970. We would not recommend buying a Vive Pro unless you have a powerful GPU. As you’ll see, our GTX 970 doesn’t quite have what it takes here. It will do the trick in less-demanding titles, but it won’t provide a great experience for all games.

Rick & Morty: Virtual Rick-ality

Rick & Morty: Virtual Rick-ality from Owlchemy Labs is the least-demanding title in our tests suite, and our GTX 970 had no trouble delivering buttery-smooth frame rates to the Vive Pro headset.

Despite the high rendering resolution, the GTX 970 kept the frame latency below 7ms. Curiously, we observed lower latency with the Vive Pro than with the Samsung Odyssey with the same hardware.

When we switched to the GTX 980 Ti, we were surprised to see that the tables had turned. The higher-power GPU kept the latency below 7ms in the Odyssey, while the Vive Pro flirted with 8ms and beyond. We’re not sure how the GTX 970 kept the latency so low.

We were surprised to see how consistent the performance results from the 9-megapixel comparison test were. The Odyssey headset gave the GPU a harder time than the Vive Pro, but the difference is frankly irrelevant, which means the Vive Pro doesn’t actually offer an advantage in visual fidelity in Rick & Morty.

Gorn

Gorn is another lightweight VR game, although it's not as well optimized as Rick & Morty. As you can see from the graphs, the frame timing in this game is not as consistent. Even with the GTX 980 Ti, the graph is all over the place. We even observed a handful of dropped frames in Gorn with both graphics cards, especially with the higher-resolution HMDs.

With both the Vive Pro and the Odyssey set with 9-megapixel rendering targets, the performance was consistent between the two headsets. We expected a performance deficit when using the Odyssey headset, but Valve updated SteamVR between our initial tests and the 1.8x render-target tests. So we could be seeing the fruits of recent optimizations.

Arizona Sunshine

Arizona Sunshine is a more demanding title than Rick & Morty and Gorn. In fact, Vertigo Games added features to the title that it claims work best on high-end Intel i7 CPUs. Our test rig features an older yet still upper-class i7 processor, so we turned up all the goodies for our tests.

Our GTX 970 kept the latency below 10ms on the Vive and Rift with all the graphics settings dialed up. However, the higher pixel count of the new headsets pushed the limits of what the GTX 970 can handle. With the Odyssey headset, the frame timing was inconsistent. But with the Vive Pro, the experience was much worse.

Note that the graph shows that we experienced high dropped frames several times in our 60-second test. Those were moments when the chaperone came into view. Our GTX 970 was unable to render the game and provide basic chaperone information simultaneously.

We didn’t observe any issues when we used the GTX 980 Ti to drive the headsets. 

Serious Sam VR: The Last Hope

We can’t say that we were surprised with the results from our Unity games, but we wouldn’t call the results predictable, either. We would describe the performance in Serious Sam VR, though, as predictable. We already know that Serious Sam is well optimized, and our GTX 970 delivers low latency and stable frame rates with the Oculus Rift and HTC Vive alike. The Odyssey headset demonstrated some performance issues, but Windows MR headset support on SteamVR is in beta, so we didn’t expect perfect results there.

The Vive Pro results surprised us, though. Despite the 9-megapixel-per-second workload, the GTX 970 kept the latency below 9ms for the whole test. The GTX 980 Ti kept the latency of the Rift, Vive, and Odyssey well below 6ms, but the Vive Pro pushed the latency up beyond 7ms.

With both headsets on level playing fields, the frame timing is more stable with Vive Pro, but with slightly more latency.

Island 359

Island 359 is the heavy hitter of our test suite. The game is optimized to deliver smooth frame rates for an HTC Vive from a GTX 970, but Cloudgate pushed the limits to get there. With our GTX 970, we can run Island 359 at high settings on our Vive, and we get excellent performance, with few dropped frames and no major hitches. However, a close look at the graphs reveals that our GPU just barely managed to keep the frames coming in under the wire.

Faced with the added strain of rendering another 4 megapixels per frame, our GTX 970 fell flat on its face when we ran the Vive Pro off it. The GPU dropped virtually every second frame and delivered frame times in the mid-20ms range.

Predictably, the GTX 980 Ti handled the workload with much better results. However, the previous-generation halo-GPU still struggled to keep the frame rate up to par. The first half of our benchmark was buttery smooth, but the last 30 seconds started to get a little rough, and the GPU started dropping frames. It appears our GTX 980 Ti’s Kryptonite is Island 359 with High settings on a Vive Pro.

This thread is closed for comments
18 comments
    Your comment
  • ingtar33
    well glad the screen door effect is discussed, somehow that's left out of many of these VR display discussions, as anyone who's actually used one for 5 minutes will tell you it's the biggest problem you'll face in VR for the most part.

    Good news on the design improvements and resolution improvement. Shame that god rays are still an issue. Seems like we're still a generation or two on this tech before it becomes something viable for the mainstream.
  • jfkeenan
    There are several VR solutions available at a variety of price points. I wish people would stop complaining that HTC is offering a premium product. When I learned the news of the pre-order I was like, "Shut up and take my money!" I just got my tax return so this was a no brainer for me. Like Kevin said, if you can't afford a Mercedes don't be mad at them for making luxury cars.
  • kcarbotte
    1027081 said:
    well glad the screen door effect is discussed, somehow that's left out of many of these VR display discussions, as anyone who's actually used one for 5 minutes will tell you it's the biggest problem you'll face in VR for the most part. Good news on the design improvements and resolution improvement. Shame that god rays are still an issue. Seems like we're still a generation or two on this tech before it becomes something viable for the mainstream.


    It's really not a big problem unless you're trying to do things that require reading text.
    As someone who's put hundreds of hours in with the Vive, I can honestly say that screen door effect is not a real problem. It's just something that people like to complain about to give them justification not to adopt VR.

    I won't deny that it exists. But I don't buy for a second that it's a real hindrance.

    That said, if that's the biggest problem that you perceive, then VR is doing pretty good, I'd say.
  • FreyjasChosen
    I'm with you, JFKEENAN. I pre-ordered the day it was available. I even bought the original HTC Vive because I needed the base stations and controllers (before the bundle was announced).

    I see a lot of complaints about the price tag of the Vive Pro, but no one's forcing anyone to buy it. I'm happy to pay for the best available product on the market.
  • Giroro
    If HTC wants to charge a premium, then the device has to BE premium. There is a definite difference between "legitimately expensive" and "overpriced".

    At the end of the day, the Vive pro is still a tethered plastic headset that uses an inferior Samsung pentile display (forcing your system to waste 30% of its resources rendering subpixels that don't actually exists). Is the design a little better thought-out? sure.. that's how product iteration works. You change the design over time to improve quality and to decrease cost. What has HTC actually done here to justify the fact that they've more than doubled their asking price?
    At the end of the day, nobody is going to pay "Lamborghini" prices for a Nissan Leaf E; not even one with particularly nice paint comfortable seats.
    Of course, that's not a perfect analogy, because a used Nissan leaf is probably less expensive than a Vive Pro if you add in a computer powerful enough to deliver 90fps to such an impractical display.
  • uglyduckling81
    Kcarbotte can I ask which titles you have put the most time into?
    I've had a vive since launch but have used it less than 30 hours. I haven't found a game to put a lot of hours into.
  • rhysiam
    The issue with your luxury car analogy is that in the case of VR, the product needs a vibrant software market to really succeed long term. When you buy a luxury car you're able to use a well-established road system that's developed over many years. That's not the case with VR. We have a classic chicken-and-egg situation where software developers are understandably nervous about investing heavily in VR experiences because the price of entry for VR results in a tiny potential market for their VR game/experience, while on the flip side, sales of headsets are being impacted by the limited games and experiences on offer. Lower headset sales means poorer economies of scale and higher per-headset prices to recoup the huge R&D investments, which just exacerbates the issue.

    Those of us (myself included) who really want VR to thrive long term do IMHO have some cause for concern with products like the Vive Pro that reinforce the fringe nature of VR and don't do a lot to entice the all-important software developers into the ecosystem.

    Another threat to VR, and perhaps a bigger threat at that, is race-to-the-bottom headsets which compromise the experience. I know several people who have tried cheap, flawed VR experiences and written off VR as a whole. At least Vive is clearly marketing - and making strides to provide - a genuinely good VR experience.

    I'm genuinely hoping for the day when I can pop on a wireless headset with 4K (or more) per eye and well-implemented foveated rendering and explore foreign worlds with interesting characters and complex stories. That's a platform that I would be more than willing to pay a lot of money to get access to. But getting there will require significant investment both on the hardware and software side. The Vive Pro headset is absolutely a step in the right direction from a technical hardware perspective, but is an (effectively) $1100 product right for the VR industry as a whole in the long term? I'm not so sure.
  • uglyduckling81
    I would say yes.
    Like you said a bad VR experience is horrendous and immediately turns you away.
    When I got the Vive I only had a GTX 970 and it made me very sick every time I tried to use it. I was on the verge of selling it and putting VR into the non compatible with me section.
    I took a different approach and upgraded to a GTX 1080 and haven't gotten sick since.
    Turns out my hopeless travel sickness body is extremely susceptible to any frame time issues and the extra GPU horsepower did the trick.
  • Ionlydothis
    The problem isn't that about offering a premium product. It's just that the Vive Pro IS NOT premium. It's exploitative. Minor improvements for a huge mark up but still the same critical flaws? Yeah, no thanks!
  • svx94
    You can position yourself like Rolex or Mercedes of luxury market, they are built for "EXCLUSIVENESS"... that magic word won't do you any good in gaming market.
  • gdmaclew
    208770 said:
    I'm with you, JFKEENAN. I pre-ordered the day it was available. I even bought the original HTC Vive because I needed the base stations and controllers (before the bundle was announced). I see a lot of complaints about the price tag of the Vive Pro, but no one's forcing anyone to buy it. I'm happy to pay for the best available product on the market.


    Saying "no one is forcing anyone to buy it" as an excuse for VR's high cost is like saying "no one is forcing anyone to buy a lamborghini". The difference here is that there are countless cars available for a much much lower price.

    Not so for VR. A few choices maybe but not enough to make this tech go mainstream.
    VR won't go mainstream in my estimation for at least another 5 years when there will be a much better selection available.
  • Specter0420
    Same FOV + same God-rays + sameish resolution + more comfort = setting a high bar?
    This is Gen 1.5 at best... I guess the Pimax will be the first Gen 2 if they can deliver.
  • dx_houle
    Specter0420
    1. Go see Tribal Instincts' review on YouTube. You won't be saying "sameish resolution" after that.
    2. Pimax uses LCD screens at a lower refresh rate and isn't even on the market. Why compare? I remember going from the Galaxy Nexus (OLED) to the Nexus 5 (LCD). The vastly reduced screen quality made me miss my old phone despite the newer one having a bigger screen and higher resolution. I'm positive most reviewers will point that out once there is actually a consumer version available.
  • computerguy72
    DX_HOULE - I tried the Pimax 8k at CES and totally disagree with the tribal instincts impressions. The guy reviewed a brief exposure to a prototype like it was actually on the market.
    I tried for quite a time and the refresh rate didn't bother me (really couldn't tell compared to my Vive). The screen resolution and fov of the pimax was epic. It does indeed have a lot of rough edges (literally as the edge of the screen had a small reflection which pimax claims is fixed).

    I genuinely hope pimax does well as that device will push VR into better experiences than it has today. I mean think about it Oculus with the Go is going literally in the opposite direction with a snapdragon 821 that is like 1% the performance of a PC.
  • dx_houle
    Computerguy72

    I'm glad to hear the Pimax 8k is actually pushing VR technology forward. I agree that Oculus is hurting VR yet again, first with exclusives, and now with inferior quality headsets.

    As for the Tribal Instincts video, I was referring to his official Vive Pro review where he shows the best comparisons of the screen quality between it and the original Vive. Those comparisons alone should make everyone saying that the Vive Pro is not enough of an upgrade to put their foot in their mouth.
  • kcarbotte
    1886042 said:
    If HTC wants to charge a premium, then the device has to BE premium. There is a definite difference between "legitimately expensive" and "overpriced". At the end of the day, the Vive pro is still a tethered plastic headset that uses an inferior Samsung pentile display (forcing your system to waste 30% of its resources rendering subpixels that don't actually exists). Is the design a little better thought-out? sure.. that's how product iteration works. You change the design over time to improve quality and to decrease cost. What has HTC actually done here to justify the fact that they've more than doubled their asking price? At the end of the day, nobody is going to pay "Lamborghini" prices for a Nissan Leaf E; not even one with particularly nice paint comfortable seats. Of course, that's not a perfect analogy, because a used Nissan leaf is probably less expensive than a Vive Pro if you add in a computer powerful enough to deliver 90fps to such an impractical display.


    More than doubled the price? How did you come up with that figure?
    The Vive launched at $800.
    The Vive Pro is now $800 + $300 for the extra controllers. That's nowhere near twice the price.
    As for justification; I went into pretty significant detail about that in the review. If you don't agree, that's fine.
    Every other headset maker failed to make anything reasonably close to as comfortable as the Vive Pro. I'm of the opinion that comfort is worth paying more for.
  • kcarbotte
    464372 said:
    Kcarbotte can I ask which titles you have put the most time into? I've had a vive since launch but have used it less than 30 hours. I haven't found a game to put a lot of hours into.


    I haven't spent more than 40 hours in many single titles, but I've spent hundreds of hours combined in the various headsets that I have at my disposal.
    The title that I've spent most of my time in is Space Pirate Trainer. I often use it to get a bit of exercise. Many people find the game boring, but in my experience, those people don't get very far. I like to challenge myself to try to reach to top 10 list with various weapons.

    I haven't spent a lot of time racing lately, but I also put a lot of time into Project Cars. I have a couple racing wheels, so I sometimes setup race days with a friend.
  • kcarbotte
    1287211 said:
    The issue with your luxury car analogy is that in the case of VR, the product needs a vibrant software market to really succeed long term. When you buy a luxury car you're able to use a well-established road system that's developed over many years. That's not the case with VR. We have a classic chicken-and-egg situation where software developers are understandably nervous about investing heavily in VR experiences because the price of entry for VR results in a tiny potential market for their VR game/experience, while on the flip side, sales of headsets are being impacted by the limited games and experiences on offer. Lower headset sales means poorer economies of scale and higher per-headset prices to recoup the huge R&D investments, which just exacerbates the issue.


    You're looking at my analogy from the perspective of today's car market, but VR can't be compared to a market that's well established over decades, with all the third party pieces in place.
    You have to look at it from the beginning of the automobile market for a proper comparison.

    We are somewhat dealing with a chicken and egg problem, but the Vive Pro doesn't do anything to exacerbate that problem. We're now in a market with many choices. There's no reason why a luxury device shouldn't exist for those willing to pay for it.
    Higher price tag usually means higher margins, which means HTC can invest more into R&D for future headsets.

    Better headsets in the future will draw more people in, which in turn will draw more developers into the fold. However, to suggest that developers aren't interested in VR is silly. Many devs have quit their jobs to pursue VR development. Big studios have big bills to pay, so they can't afford to take the risk, but small studios are jumping into VR all the time, and they are growing.

    It takes time to build software. Moreover, it takes time to figure out what works in VR and what doesn't. Cheaper hardware ins't going to fix that. The Vive Pro doesn't need to be cheaper and attract a ton of new people, because the software is still immature. We're still very much in early adopter territory.


    1287211 said:
    Those of us (myself included) who really want VR to thrive long term do IMHO have some cause for concern with products like the Vive Pro that reinforce the fringe nature of VR and don't do a lot to entice the all-important software developers into the ecosystem.


    When there's more than a dozen headsets on the market, a single high-priced device is not doing anything to scare of developers. If anything, they are excited for a comfortable headset that they can wear for extended periods while working. This kind of headset makes tools like the Unity and Unreal VR editors useful and practical.
    No one wants to wear an uncomfortable headset for hours on end. Better hardware should help in that regard.


    1287211 said:
    Another threat to VR, and perhaps a bigger threat at that, is race-to-the-bottom headsets which compromise the experience. I know several people who have tried cheap, flawed VR experiences and written off VR as a whole. At least Vive is clearly marketing - and making strides to provide - a genuinely good VR experience.


    So, which would you rather? Cheap headsets with corners cut, or expensive headsets with innovative new design features?


    1287211 said:
    I'm genuinely hoping for the day when I can pop on a wireless headset with 4K (or more) per eye and well-implemented foveated rendering and explore foreign worlds with interesting characters and complex stories. That's a platform that I would be more than willing to pay a lot of money to get access to. But getting there will require significant investment both on the hardware and software side. The Vive Pro headset is absolutely a step in the right direction from a technical hardware perspective, but is an (effectively) $1100 product right for the VR industry as a whole in the long term? I'm not so sure.


    The Vive Pro isn't going to be a long term solution. HTC said the original Vive would be in production through 2018. I suspect we'll see a Vive 2.0 announced later this year or around CES for release next year. That will be the true replacement for the Vive.
    At that time, I would expect the Vive Pro to become the cheaper model, and the Vive phased out.

    The headset that you want is pretty much the headset everyone wants. Ergo, it will happen one day.
    The technology just isn't there yet.