Hacker News Reader

Comment by Ajedi32 | Hacker News Reader

AFAIK the DK1 didn't even have positional tracking. Going from that to a Quest would be a huge step up in basically every conceivable metric.

Positional tracking and hand presence make a massive difference.

View on HN

Well, except maybe for actual system performance. No need for a computer is great, but shoehorning in enough CPU and GPU power for this in a $400 budget that also includes all the hardware, combined with the promo videos all being third person views and the items held all being shown as wire frame and single color means there's quite a bit unknown about what it will actually be like (unless there is more info I'm missing).

It kind of reminds me of Virtual Boy...

I'm at Oculus Connect 5 right now and you can try it out. It's a finished product, not a concept or prototype. It's one of the coolest things. EVER!

I've read the screen's refresh rate is 72 Hz, lower than the 90 Hz that Oculus/Valve have been pushing as a minimum for immersive VR. Did this seem like a problem, or does it feel good enough?

I know the Go already runs at 72 or 60 Hz, but without positional tracking it was never going to be particularly immersive no matter what framerate it had.

In Carmack’s talk today he said many people find 60hz uncomfortable, but far fewer need 80hz to feel comfort. And 72hz is 3x 24fps which most videos play in so it helps for smooth video playback. And 96hz would be a substantial performance hit.

For me, it's not a problem. The go runs 72, but most apps don't take advantage of that But it's true that some people are more sensitive to lower refresh rates. As another user said here, Carmack said that with 72 Hz, almost every person should feel fine, so there is almost no upside in bringing it further up.

Fair point, but it's kinda hard to make full use of PC's graphical capabilities when the hardware caps you to a resolution of 640 x 800 per eye and 60 fps.

If the Oculus Go is anything to go by, I suspect most games built for the Quest will match up fairly well graphically against experiences originally built for the DK1.

Oculus Go runs a Snapdragon 821, Oculus Quest has a Snapdragon 835 [1]. Android Central calls it 27% faster at 40% less energy [2]. GPU in the 835 goes from an Adreno 530 to Adreno 540, which looks like a good step up but nothing monumental.

In terms of large scenes with high res textures and detailed models, probably not going to do much better than the Go. It's really all about the resolution increase and 6dof tracking, more than trying to get anywhere near the Rift's visuals.

[1] https://gizmodo.com/the-big-thing-facebook-didnt-say-about-o...

[2] https://www.androidcentral.com/snapdragon-835-vs-snapdragon-...

Yeah, the DK1 I tried was leftover from some research project and ended up at the local maker space for a while. I assume the tracking quality on that is about on par with Google Cardboard.

I don’t have a space for a real “room scale” setup with Rift cameras or Vive beacons, so I’ve been waiting on Santa Cruz’s inside-out tracking to look at VR more seriously.

Pleasantly surprised by price. They’ve confirmed it’s based on a Snapdragon 835, which isn’t a top end SoC, and I think that’s a smart move to get it down to this range. At $500 they’d have a much harder sell, and this is where PS4 and Xbox One started (once they dropped Kinekt, anyway).

By the way, do you have an idea why Microsoft has dropped Kinect?

I couldn't say for sure, but I'd bet that it's because they priced the Xbox One at $500 against the PS4's $400 and between the two options a lot of people would just buy the cheaper one (especuially Santa). Dropping the included Kinekt was a way to get the cost competitive.

I didn't see anything specifically on this, but I think it would be fair to assume the screen will be significantly higher than the DK1. Even the resolution of the Oculus Go is far better.

Resolution is _way_ better than the DK1. The DK1 was 640 x 800 per eye. The Quest is 1,600 x 1,440 per eye. More than quadruple the pixel count.

Absolutely! I don’t have the numbers handy but I ran them earlier and it’s about 1.7x the pixel count of the CV1, which was already better than the DK1.

Reply on news.ycombinator.com