The Inside Story of Oculus Rift | Hacker News Reader

The Inside Story of Oculus Rift

wired.com

80 pointsrelampago28 comments

It doesn't really mention that Palmer Luckey was working with Mark Bolas' VR group at USC [1] at the time of the Kickstarter. The Rift isn't the product of a lone 18-year-old VR autodidact.

I'm not 100% certain regarding the timeline, but it's starting to seem odd how every article about Oculus seems to marginalize everyone but Palmer and Carmack.

[1] http://interactive.usc.edu/2012/06/12/mxr-lab-members-and-al...

Most of these articles marginalize key players in relation to the public face

It seems that geeks are less excited about Oculus in general since the acquisition thing.

Did Valve really contribute the most important part to Oculus that is the no-nausea experience? If so, on what terms did they share the technology with Oculus? I'm wondering if Valve is the winner in the shadows or the biggest loser.

IIRC, Valve wants to make games for VR, they don't want to make VR.

This is why they let Jeri Ellsworth go with all the IP to CastAR.

That's the current state, but surely if they patented the no-nausea bit of tech, they wouldve earn millions just licensing it. It makes me wonder whether the Valve claim in the story is accurate or perhaps there is another story behind the official one.

So, that's likely absurd, right?

The "no-nausea" bit is more likely a combination of factors: low-latency, head tracking with thus-and-such a precision, field-of-view tweaks, etc.

It'd be like patenting the ideal air-to-gasoline burning ratio: it's just a fact of life.

> It'd be like patenting the ideal air-to-gasoline burning ratio: it's just a fact of life.

That never stopped people from trying to patent such facts of life, often succeeding and making tons of money with it or forcing competitors out of the field. Patents are like swords, use or abuse, to some only winning matters and they count their wins by their pocket-books.

My impression is that Valve contributed two qualitative improvements: replacing accelerometer/gyro with optical tracking and replacing LCD with OLED and blanking the OLED between each frame to reduce persistence.

The first is a very obvious step in improving absolute pose estimation--one implemented in various videos around the time the devkit came out.

The other seems like a natural evolution of display technology.

Valve didnt invent this tech.

Short flashy exposures of the display is pretty much standard in gaming monitors and everyone knows it is _the_ thing to do to minimize motion blur (LightBoost).

Using camera to extract absolute position instead of drifty relative mems trackers is also no brainer. This is what N originally did with Wii remote. N also got sued at least four times over this exact technology.

Sidenote: wired.com articles have a truly annoying UX error. The slider/gallery goes full screen when the user presses 'F', which hides the article.

Not very helpful when you want to find something written in the article by Ctrl+F!

For me, almost every single time: Ctrl+Shift+K, push to Kindle, close the browser.

Rubin helpfully fills in a lot of blank spaces in what we already knew about this. The one nagging doubt that still lingers sits somewhere between what is doable, versus what is done. The Kinect, for all it's technological triumphs and impressive market performance, still leaves us waiting for whatever it was that made it seem like it was going to change everything. While few who have ever had the Oculus experience have any doubt that it could change everything, we mustn't forget that we are also all still waiting for the first concrete confirmation that this expectation will be met.

Kinect rode on Microsoft >$1 billion marketing campaign of lies. Every PR clip showed people interacting with fast paced action packed games in very responsive way.

Reality of hardware was totally different. M$ wasnt ready, Project Natal was couple of years away. They scrambled in panic and bought Primesense. Kinect was basically repackaged Primesense dev kit.

Kinect was supposed to be answer to Wii. Problem is there is a huge difference in speed between tracking 4 points of light versus complex DSP on a point cloud. Kinect had a 100-500ms delay between something happening and skeletal tracking on the screen. Kinect One only manages to cut that in ~half.

You simply CANT do responsive input with latencies in the hundreds of milliseconds. This is why Kinect has failed to deliver what $1 billion of M$ PR promised four years ago.

Kinect is great for puzzle games, art projects, slow complex input.

do you have sources for the latency numbers for kinect on xbox one? I've searched and can't find anything concrete and the XDK isn't publicly available.

and this might downvote me into oblivion, but is using "M$" still a thing? I don't see how that helps a discussion.

Unofficial Kinect one docs say 20ms depth sensor latency, this is same as 50fps vision camera so no lag there. They also say 60ms for tracking versus 90ms of original Kinect. Problem is original kinect did >100ms at minimum, and topped at 500ms

http://www.nrl.navy.mil/itd/imda/sites/www.nrl.navy.mil.itd....

   We experienced maximum latencies of nearly 500 ms

M$ will cease to be a thing when Microsoft stops dropping Billion dollars on every problem (kinect pr, original xbox pr, surface writeof)

M$ is juvenile. Big companies will tend to spend up on building and then promoting their products. What's the shame in that?

The shame is with enough money pigs can fly. If the product you were making was something of quality that people wanted, you wouldn't need billions to ram it down the public's throat. (You'd probably only need hundreds of thousands.)

How are you going to quantify what constitutes normal advertising and what is "ramming it down throats"?

You decide, how would you classify paying people from third world countries to spam YT with "product reviews" in broken english?

They arent spending on building brands, they are trying to directly buy users. Bribes in form of discounts, "cross promotion" deals that in effect mean they give you stuff for free just to brag in the press how you use it. Huge FUD campaigns. M$ does it all.

Recently mentioned on HN Gopro is an excellent contrast. Bottom up organic marketing, investing in events, users and activities. Touching peoples lifes. Concentrating on positives, not gimmicks or attacking competition.

Microsoft is simply hostile and out of touch. Money is their only language.

GoPro does cross promotion with TV/media companies and they often run discount coupons to attract sales. This is what companies do to promote their products. Microsoft also invests in events, users and activities.

I didn't know it had shipped to retail already

It hasn't. That's why the article says: "Now Oculus is hard at work on its long-awaited headset for consumers, which the company predicts will be released later this year, or more likely early next year, or perhaps even not so early next year."

Yes I get that but the article says how it became a reality. It's been a non-consumer reality for quite a while in arcades. It's still not a reality unless you have a dev kit.

Wired, your pages are slooow.

Reply on news.ycombinator.com