Apple Reportedly Working on 3D Sensor System for Rear Camera in 2019 iPhones

[MacRumors]

Apple is developing 3D depth sensing technology for the rear-facing cameras in its 2019 iPhones, according to a new report by Bloomberg on Tuesday. The 3D sensor system will be different to the one found in the iPhone X's front-facing camera, and is said to be the next big step in turning the smartphone into a leading augmented reality device.

Apple is evaluating a different technology from the one it currently uses in the TrueDepth sensor system on the front of the iPhone X, the people said. The existing system relies on a structured-light technique that projects a pattern of 30,000 laser dots onto a user's face and measures the distortion to generate an accurate 3D image for authentication. The planned rear-facing sensor would instead use a time-of-flight approach that calculates the time it takes for a laser to bounce off surrounding objects to create a three-dimensional picture of the environment.

The existing TrueDepth camera would continue to be used in the front-facing camera of future iPhones in order to power Face ID, while the new system would bring the more advanced "time-of-flight" 3D sensing capability to the rear camera, according to the sources cited. Discussions with manufacturers are reportedly already underway, and include Infineon, Sony, STMicroelectronics, and Panasonic. Testing is said to be still in the early stages, and could end up not being used in the phones at all.

With the release of iOS 11, Apple introduced the ARKit software framework that allows iPhone developers to build augmented reality experiences into their apps. The addition of a rear-facing 3D sensor could theoretically increase the ability for virtual objects to interact with environments and enhance the illusion of solidity.

Apple was reportedly beset with production problems when making the sensor in the iPhone X's front-facing camera, because the components used in the sensor array have to be assembled with a very high degree of accuracy. According to Bloomberg, while the time-of-flight technology uses a more advanced image sensor than the existing one in the iPhone X, it does not require the same level of precision during assembly. That fact alone could make a rear-facing 3D sensor easier to produce at high volume.

Late last month, oft-reliable KGI Securities analyst Ming-Chi Kuo claimed that Apple is unlikely to expand its front-facing 3D sensing system to the rear-facing camera module on iPhones released in 2018. Kuo said the iPhone X's 3D sensing capabilities are already at least one year ahead of Android smartphones, therefore he believes Apple's focus with next year's iPhone models will be ensuring an on-time launch with adequate supply.

Article Link: Apple Reportedly Working on 3D Sensor System for Rear Camera in 2019 iPhones

 

[meaning-matters]

This is very hard because you need to measure time-of-flight differences. If they do manage to pull this off, it would be truly amazing.

Example: For depth resolution of 1cm you'd need to be able to measure time differences of 33.3 pico seconds (i.e. 0,0000000000333 or 33.3 trillionths seconds). But 1cm for facial recognition is too coarse; they'd need a fraction of this.

It's been done in high-tech systems taking up more space than a whole iPhone.

 

[bmot]

Im surprised how long this whole AR topic takes to take form.
I remember i had a map app on my 3gs back then that would project map infos into my surroundings on screen.

 

[iPhysicist]

If they do this right this will be bigger then iPhone and everything else they did before. But it will undoubtedly be without any projections. Should be something with 3D camera depth sensing.

Pure awesomeness for the geek and everyone else

 

[GrumpyMom]

Are these laser dots like the lasers used in some autofocus camera systems (only multiplied many times over)?

It would be cool if the new tech meant better focusing, too. No more blurry cat photos. They’d be fuzzy only because the cat is fuzzy.

 

[iModFrenzy]

Cue the comments from people about how they’re glad they didn’t buy the current X and waited

 

[JohnApples]

Yikes, next year’s iPhone is sounding like a beta test. There’s bound to be problems in a gen 1 rear 3D camera system.

Good luck guinea pigs! I’ll be buying the iPhone after next as all the kinks in the rear camera will have been ironed out.

Edit: I thought I was obviously tongue-in-cheek but I guess not. So here’s the tag: /s

These types of comments are so common here that people just assume you’re being serious. Lol.

 

[robinp]

This is very hard because you need to measure time-of-flight differences. If they do manage to pull this off, it would be truly amazing.

Example: For depth resolution of 1cm you'd need to be able to measure time differences of 33.3 pico seconds (i.e. 0,0000000000333 or 33.3 trillionths seconds). But 1cm for facial recognition is too coarse; they'd need a fraction of this.

It's been done in high-tech systems taking up more space than a whole iPhone.

But it doesn't need to work with FaceID if it is the rear camera. It would be about understanding 3D space in a general sense. An upgrade over the dual camera set up for producing portrait mode 'bokeh'

 

[KONVICTED]

Cue the comments from people about how they’re glad they didn’t buy the current X and waited

Everyone has different preferences. i own iPhone X and like it. I upgrade annually and will definitely check out Plus iPhone X coming next year.

 

[futbalguy]

This is very hard because you need to measure time-of-flight differences. If they do manage to pull this off, it would be truly amazing.

Example: For depth resolution of 1cm you'd need to be able to measure time differences of 33.3 pico seconds (i.e. 0,0000000000333 or 33.3 trillionths seconds). But 1cm for facial recognition is too coarse; they'd need a fraction of this.

It's been done in high-tech systems taking up more space than a whole iPhone.

It sounds like this is for augmented reality instead of face scanning so they wouldn’t need that level of accuracy. Also explains why they would use different technologies on the front and back camera. Different use cases - short-range, high accuracy vs long-range, low accuracy (relatively).

 

[BvizioN]

Good luck guinea pigs! I’ll be buying the iPhone after next as all the kinks in the rear camera will have been ironed out.

Looking at the horizon for something better........

 

[dan9700]

In 2030 another new camera leave it out enjoy the X and stop talking about years to come stupid rumours

 

[Jsameds]

In 2030 another new camera leave it out enjoy the X and stop talking about years to come stupid rumours

Do you realise what site you're on?

 

[WatchFromAfar]

But it doesn't need to work with FaceID if it is the rear camera. It would be about understanding 3D space in a general sense. An upgrade over the dual camera set up for producing portrait mode 'bokeh'

How about if that's the only reason for it? Like being able to recognize your friends and family?

 

[Amplelink]

Do you realise what site you're on?

No, he clearly doesn’t! Haha

I always thought this kind of thing is what would really boost AR adoption. I was expecting this for iPhone X and not the front facing camera stuff.

 

[AyeKay15]

As good ol’ Jim Ross would say....

BAH GAWD ALMIGHTY!

 

[GrumpyMom]

Yikes, next year’s iPhone is sounding like a beta test. There’s bound to be problems in a gen 1 rear 3D camera system.

Good luck guinea pigs! I’ll be buying the iPhone after next as all the kinks in the rear camera will have been ironed out.

But what if you’re eaten by the Kraken the year before? As you slide down its gullet or whatever Krakens have, your last words will be “I should have gotten the 2019 IPhone with the awesome state of the art rear sensor...ack...gurghhhhhh.” So tragic.

 

[unibility]

This is very hard because you need to measure time-of-flight differences. If they do manage to pull this off, it would be truly amazing.

Example: For depth resolution of 1cm you'd need to be able to measure time differences of 33.3 pico seconds (i.e. 0,0000000000333 or 33.3 trillionths seconds). But 1cm for facial recognition is too coarse; they'd need a fraction of this.

It's been done in high-tech systems taking up more space than a whole iPhone.

I’m not sure what I just read but I believe.

 

[Enygmatic]

*chuckles* Yeah, let's hurry and get these "2019 iPhone" rumors out there now, so that Samsung has something to rush to market in their 2018 flagships. Then, by the time Apple presents at the Keynote, everyone's "bored".

Rinse and repeat every year. Ming and Bloomberg, like clockwork.
[doublepost=1510666046][/doublepost]

Yikes, next year’s iPhone is sounding like a beta test. There’s bound to be problems in a gen 1 rear 3D camera system.

Good luck guinea pigs! I’ll be buying the iPhone after next as all the kinks in the rear camera will have been ironed out.

Next year is 2018... the article is about 2019.

 

[thisisnotmyname]

Very cool, anything they can do to make AR more powerful I'm in favor of.

 

[WannaGoMac]

Cue the comments from people about how they’re glad they didn’t buy the current X and waited

Or people who don't care about having the latest thing, don't have time or energy to swap phones every year, have learned to never buy first gen apple stuff and its inevitable bugs (iphone 6+ touch disease, etc.), and prefer something that's stable and polished.

But to each their own...

 

[slicecom]

So glad I decided to skip the 2018 model.

/s

Every time a 2018 model rumour is posted, people say they're so glad they skipped the X, so this is the next logical step.

 

[HobeSoundDarryl]

Apple was reportedly beset with production problems when making the sensor in the iPhone X's front-facing camera, because the components used in the sensor array have to be assembled with a very high degree of accuracy. According to Bloomberg, while the time-of-flight technology uses a more advanced image sensor than the existing one in the iPhone X, it does not require the same level of precision during assembly. That fact alone could make a rear-facing 3D sensor easier to produce at high volume.

Somehow, I suspect this or something else will be in "short supply" or have "production challenges" leading into the launch. The show is always the same... and always seem to work (so why not?). If not this, something else will cast a lot of doubt on being able to get the next iPhone at launch and once again create the frenzy of waking up in the wee hours to try to get one of the precious few that will be available before they are all snatched away.

Furthermore, i predict the next, next iPhone will have the exact same issue of short supply rumors flying hot & heavy going into it's launch.

And then the next, next, next iPhone too.

If it ain't broke?

 

[smacrumon]

How difficult could a 3d sensor be for a multi billion dollar company to work out?

 

[Jsameds]

How difficult could a 3d sensor be for a multi billion dollar company to work out?

Not that difficult. Mass producing it in the tens of millions and hitting a defect rate of below 0.0001% is the hard part.