Apple's new iPhones shift smartphone camera battleground to AI
Send a link to a friend
[September 12, 2019] By
Stephen Nellis
(Reuters) - When Apple Inc <AAPL.O>
introduced its triple-camera iPhone this week, marketing chief Phil
Schiller waxed on about the device's ability to create the perfect
photograph by weaving it together with eight separate exposures captured
before the main shot, a feat of "computational photography mad science."
"When you press the shutter button it takes one long exposure, and then
in just one second the neural engine analyzes the fused combination of
long and short images, picking the best among them, selecting all the
pixels, and pixel by pixel, going through 24 million pixels to optimize
for detail and low noise," Schiller said, describing a feature called
"Deep Fusion" that will ship later this fall.
It was the kind of technical digression that, in years past, might have
been reserved for design chief Jony Ive's narration of a precision
aluminum milling process to produce the iPhone's clean lines. But in
this case, Schiller, the company's most enthusiastic photographer, was
heaping his highest praise on custom silicon and artificial intelligence
software.
The technology industry's battleground for smartphone cameras has moved
inside the phone, where sophisticated artificial intelligence software
and special chips play a major role in how a phone's photos look.
"Cameras and displays sell phones," said Julie Ask, vice president and
principal analyst at Forrester.
Apple added a third lens to the iPhone 11 Pro model, matching the
three-camera setup of rivals like Samsung Electronics Co Ltd <005930.KS>
and Huawei Technologies Co Ltd [HWT.UL], already a feature on their
flagship models.
But Apple also played catch-up inside the phone, with some features such
as "night mode," a setting designed to make low-light photos look
better. Apple will add that mode to its new phones when they ship on
Sept. 20, but Huawei and Alphabet Inc's <GOOGL.O> Google Pixel have had
similar features since last year.
In making photos look better, Apple is trying to gain an advantage by
way of the custom chip that powers its phone. During the iPhone 11 Pro
launch, executives spent more time talking its processor - dubbed the
A13 Bionic - than the specs of the newly added lens.
[to top of second column] |
CEO Tim Cook presents the new iPhone 11 Pro at an Apple event at
their headquarters in Cupertino, California, U.S. September 10,
2019. REUTERS/Stephen Lam - HP1EF9A1EM211
A special portion of that chip called the "neural engine," which is reserved for
artificial intelligence tasks, aims to help the iPhone take better, sharper
pictures in challenging lighting situations.
Samsung and Huawei also design custom chips for their phones, and even Google
has custom "Visual Core" silicon that helps with its Pixel's photography tasks.
Ryan Reith, the program vice president for research firm IDC's mobile device
tracking program, said that has created an expensive game in which only phone
makers with enough resources to create custom chips and software can afford to
invest in custom camera systems that set their devices apart.
Even very cheap handsets now feature two and three cameras on the back of the
phone, he said, but it is the chips and software that play a huge role in
whether the resulting images look stunning or so-so.
"Owning the stack today in smartphones and chipsets is more important than it's
ever been, because the outside of the phone is commodities," Reith said.
The custom chips and software powering the new camera system take years to
develop. But in Apple's case, the research and development work could prove
useful later in products such as augmented reality glasses, which many industry
experts believe Apple has under development.
"It's all being built up for the bigger story down the line - augmented reality,
starting in phones and eventually other products," Reith said.
(Reporting by Stephen Nellis in San Francisco; Editing by Lisa Shumaker)
[© 2019 Thomson Reuters. All rights
reserved.] Copyright 2019 Reuters. All rights reserved. This material may not be published,
broadcast, rewritten or redistributed.
Thompson Reuters is solely responsible for this content.
|