Welp, that was a fail

This is what gets me whenever some new headset comes along with high numbers on their spec sheet. There's so much more to having a good VR experience than just how many pixels you have. The Rift is a fantastic piece of design work that brings together everything (both hardware and software) to provide a top-notch experience. I think we forget how good we have it when we see some of these upstart 4-8-whatever-K headsets and declare "This will be the best VR HMD on the market!!"

Even relying on SteamVR to do some of the heavy lifting, if Samsung can't pull it off, how are some of these other guys going to provide a better HMD than the big boys?

nalex66 said:

This is what gets me whenever some new headset comes along with high numbers on their spec sheet. There's so much more to having a good VR experience than just how many pixels you have.

My issue isn't how numbers don't align with perceived improvement, but the vastly different reports of the latter.

To answer some questions, no, I couldn't see individual pixels but after about 12 hours of tinkering, I think I know what the issue is. It seems to use the cheater resolution technique the current Vive uses. What seems to be going on is while the Rift uses 3 subpixels per pixel (1 red, 1 green, 1 blue for each pixel), the Odyssey does not. It uses 1 subpixel and then shares the other 2 subpixels with its neighbors. This is just a guess on my part, but it explains shimmering I would see around object edges in some games and why text was harder to read. If I'm right, the Odyssey is actually running at about 2/3rds the resolution it claims due to the subpixel setup. That would make it lower resolution than the Rift.

Unfortunately this isn't reflected in GPU usage. In DCS, the Rift eats 5.5 GB of VRAM in a test mission I have, the Odyssey eats 8.5GB at the exact same settings in that mission. The Rift also maintained 90 fps (just barely) while the Odyssey mostly hung out in the 60-70 fps range.

As for why reviewers would say the Odyssey looks better. That's because it does at first glance. The displays are far more vibrant. I was truly blown away the first time I put it on and loaded up the Microsoft Home. It just looks incredible compared to the Rift. Even in DCS where reading text was harder, the initial impression was excellent.

In high contrast scenes, the Odyssey also lacks the massive godrays the Rift has. It does have a little bit of bright blur in extreme cases, but it is far less distracting. Additionally, the SDE is noticeably lower. By SDE I mean the gaps between pixels. I really can't see them in the Odyssey. But as said, the actual resolution seems to be lower despite having more "pixels" and less of a gap between them.



edit:

As a side note, the controllers for it are certainly usable. They don't detect if your finger is touching the trigger or if your thumb is touching the touchpad/stick or not, so pointing is out but they otherwise work well. They aren't as ergonomic as the Rift's, but are are still good. Tracking was spot on too for the most part, though I had to be careful about reaching down to my hips to grab guns. I found it best to look down a little when doing so. Not having to deal with sensor LoS was an improvement too, but I have a dodgy Rift setup at  home due to limited space and a pet rabbit that likes the taste of electricity.

I gave FO4 VR a try while I was at it, the controls suck just as hard with touchpads imho. Actually I think they sucked more despite the game being designed exclusively for them instead of our Rift stick hacks. I just couldn't feel where the center was so I kept switching pages in the inventory when I meant to scroll down instead.

One final comment before I go return it, somebody somewhere mentioned it might have a softer focus. I pulled it back out of the box to try to verify and was able to confirm it. The problem isn't the resolution, it's that the pixels are slightly blurred. This makes reading small text (such as cockpit instruments) impossible. To be fair, without any super sampling at all, it is totally unreadable with the Rift too. The difference is with a  healthy dose of super sampling, the crisper Rift image makes it mostly readable. Super sampling doesn't have the same effect with the Odyssey since it seems deliberately blurred slightly, making sharp images impossible.

BeastyBaiter said:

Been really enjoying my Rift since summer but the low resolution finally got to me. Thus I broke down and went to the local Microsoft store at the mall and picked up a shiny new Samsung Odyssey. Now before Zenbane shits a brick, it turned out to be a bit fail. I have no idea how Samsung managed to pull this off, but despite having vastly superior lenses (no more death by lens flare) and roughly 30% better resolution with the same FoV, it's harder to read instruments in DCS and I swear it's lower resolution overall. I honestly wasn't expecting that at all, the whole point was to get better image quality, not worse!

On the flip side, loading up Robo Recall and From Other Suns with it was far less of a hassle than I was expecting. Plan to return it tomorrow though. I just can't figure out how lower image quality with better lenses and screens is possible. It does have much better contrast and colors though, I will give it that. Really stunning looking if it weren't for the impossibly low resolution. Not a super sampling issue either, tried the same settings (steamVR vs steamVR and in game vs in game, I know OTT uses a different measurement). I even went full retard with the Odyssey and crushed my 1080 TI with super sampling, the Rift still appeared higher resolution with only modest super sampling (1.7 PD for Odyssey vs 1.3 PD for Rift in DCS).

how was the inside out tracking? thats what this uses right? does it support touch or vice wands?

@BeastyBaiter: Yeah, that's what I was afraid of. By simultaneously raising the magic pixel resolution number (which is easy to market) a little bit, pumping up the "vibrance" which always makes everything look better at first, and blurring the image to get rid of the screen door effect the cheap way, they are actually poisoning the market.

So, we'll need more powerful GPUs to actually get less visible detail. And games will be struggling even more.

That's pretty sad news for VR  :'(

danknugz said:

how was the inside out tracking? thats what this uses right? does it support touch or vice wands?

Yes, it uses inside out tracking instead of external cameras. Headset tracking was flawless, for my setup I found it to be better than the Rift's (only have 2 sensors). Hand tracking was mostly good too. With my hands in view, it was perfect but it was less reliable if I reached for holsters at my hips without looking down. It still worked most of the time, but there was a pretty good chance I'd reach down and grab nothing if I didn't also look down. If I did look down, it worked flawlessly since my hands were in view or close to it.

Based on my experience, I think inside out tracking such as on the Odyssey is probably the way forwards. It's cheaper to make, easier for the end user to setup and works well enough.

As for the touch controllers, it uses its own. They are a cross between the Rift touch and the Vive wands. Ergonomics are between as well. The Rift's controllers are substantially better feeling, have finger tracking and more buttons. The Odyssey has both sticks and trackpads though, so no issues playing both Vive and Rift exclusives with it. The subjective feel of the controllers was cheaper on the Odyssey. Felt like cheaper plastic and the buttons weren't as satisfying. That's purely subjective of course.

RedRizla said:

Wildt said:

@BeastyBaiter  - Dang.. your impression of the visual quality goes against everything else I've read about this device. Very odd!

I was told by a friend who has a Samsung headset the same thing and thought his eyes must need testing. I said about this in another thread because the Vive pro is going to get the same resolution. I managed to to see the Samsung headset for myself when I called at my friends house. For me I could see the text a bit more clearer (I wasn't playing DCS) on the Samsung, but overall it doesn't look much different from a CV1 visually. People are going to pay all that money for a Vive Pro and if it's anything like the Samsung headset then I expect they'll be disappointed. I'm going to wait for nothing short of a 2k headset after seeing this for myself.

I think the problem is that the WVR headsets have a MUCH smaller sweet spot going by reports I've read. The Vive Pro will have a larger sweet spot (the same size as the current Vive which is slightly smaller than the Rift's) so it'll be easier to find once you put the headset on.

Try a bit of fiddling around with the placement of the Odyssey and you'll probably find that sweet spot and things will look sharper.

The advantage that the Rift has is that there's considerably less SDE compared to the Vive and the displays are sharper and clearer so you won't notice as much difference doing an A>B comparison with the Rift and Odyssey as you would comparing the Odyssey with the Vive or comparing the Vive to the Vive Pro.

I didn't find the sweet spot with the Odyssey any harder to find than with the Rift. But I do wear glasses under it, so getting the Rift just right always takes a bit of fiddling with.

Here's another recent contradictory comparison by someone who allegedly has most HMDs on the market.
On the negative side, he reports god rays in the same league as the Rift.