Archived from groups: alt.tv.tech.hdtv (More info?)

The Selling of Cellphone TV
Qualcomm and Texas Instruments have unveiled two competing schemes for
delivering video to cellphones that bypass 3G. Is the market big enough
to get them off the ground?

By ERIC S. BROWN
November 12, 2004 2:13 p.m.


Recent signs of trouble in the large-screen TV business suggest there
might be market-driven limits to how big a television set can be. Yet
might there also be minimum size limits?

Judging from two recent announcements on mobile-video delivery, we're
likely to find out soon. First, Texas Instruments Inc. said that it
plans to introduce a chip called Hollywood that will decode and display
digital TV receptions using new mobile versions of digital television
broadcast standards. TI claims that Hollywood-equipped cellphones, which
should arrive in 2007, will be capable of displaying video at digital TV
resolutions.

Then Qualcomm Inc. disclosed that it was spending $800 million to create
a nationwide network called MediaFLO that should go online by 2006.
MediaFLO will use new wireless frequencies bought last year by Qualcomm
to deliver video and other multimedia content to mobile devices.

Both the Qualcomm and Texas Instruments technologies make it possible to
offload video delivery from new third-generation cellular networks and
place it on dedicated video delivery networks. Considering that a big
part of the 3G hype was the technology's ability to deliver video, this
development is a bit ironic, to say the least.

It's true that video of a sort has recently arrived on 3G. For the last
few months, Sprint Corp. PCS Vision Multimedia Services has been
offering as many as 600 video clips a day to PCS subscribers who own a
special Samsung phone. In October, AT&T Wireless Inc. (now part of
Cingular) formally launched its own multimedia service, which is based
on the same wireless broadcast network used by Sprint: Idetic's MobiTV.
The drawback is that at best, the frame rate is six to 10 frames per
second; users with older phones see one-frame-per-second video, which is
more like a slide show. Generally, 10 to 15 frames-per-second rates are
considered to be the lower limits necessary to create an acceptable
illusion of motion, and the Qualcomm and TI technologies promise to
offer 24 to 30 frames per second -- the latter being the standard used
by TV broadcasts.

A number of technological trends should get streaming cellular video up
into double-digit frame rates within a few years. First, new video chips
from Qualcomm and others have arrived this year that support the H.264
video format, which is designed to get the most out of lower bandwidth
networks. Also, 3G speeds should gradually rise in the coming years to
boost frame rates and overall quality. Yet the next-generation 4G
technology and handsets that could deliver 30 frames-per-second video at
digital TV-quality resolutions may be 10 to 20 years away. This is why
Qualcomm, Texas Instruments, and other companies decided to create
parallel broadcast networks.



Also in MIT Technology Review
(www.technologyreview.com)

 The Physics of Gluttony
 
 Atom Flip Energy Measured
 
 Turn On, Tune In, Pay Up
 


TI's Hollywood chip combines three TV chips into one and promises 24 to
30 frames per second playback of live video. It supports two emerging
digital TV standards for mobile devices: DVB-H (a European-based
standard now under trial that was developed by a consortium called the
Digital Video Broadcasting Project) and a rival Japanese format called
ISDB-T (Integrated Services Digital BroadcastingTerrestrial). Expected
to be available by 2007, both formats use various techniques to reduce
battery consumption compared to standard digital broadcasting
technologies. For example, DVB-H (the H is for "handheld" ), which was
finalized earlier this year, temporarily shuts off tuner chips between
broadcast bursts -- a technique known as "time slicing."

Of the two formats, DVB-H seems to have more momentum in the U.S. In
October, Nokia and cell-tower operator Crown Castle announced they had
begun trials of the technology in Pittsburgh. The DVB-H broadcasting
equipment could operate either in conjunction with TV broadcasters'
standard digital broadcasting equipment or be run by cellular providers
as a separate service from their 3G networks.

The Qualcomm MediaFLO network is based on its FLO (forward link only)
multicasting technology rather than digital broadcasting, and it will
also accommodate the delivery of streaming content over 3G networks. FLO
transmits in the 700-megahertz spectrum (UHF channel 55) and requires 30
to 50 times fewer towers than a cellular network. On the handset side,
Qualcomm will announce a client chip for FLO in the coming year,
promising playback at up to 30 frames per second.

The MediaFLO network, which will be developed and operated by a Qualcomm
subsidiary called MediaFLO USA, is based on a "push" multicasting
paradigm in which battery consumption is reduced by downloading video
only during certain periods when the phone is not in use. With enough
storage on the phone, users could interactively access an accumulation
of content. The technology would also be able to integrate live
broadcasting along with canned "clipcasting" content.

According to Rob Chandhok, vice president of engineering at Qualcomm,
storage costs would not add significantly to handsets. "The thing that
is dropping in price the fastest on handsets is memory," he says. More
important, says Mr. Chandhok, is the issue of battery life, a key reason
why Qualcomm decided not to embrace the more ambitious DVB-H. "We're
going to be significantly better than DVB-H in terms of power savings,"
he says.

Both the Qualcomm and Texas Instruments technologies would permit
synchronization with 3G data services running simultaneously on the
cellphone to enable a miniature version of interactive TV. For example,
cellular providers might pop up an interactive data window over a TV
advertisement, letting users instantly purchase the product over the
Web. Cellular providers may well support at least one of these new
mobile digital video delivery technologies because they solve the
bandwidth (and therefore, quality) limitations of cellular 3G services
by simply bypassing them.

Yet, TI's projection that 70% of new phones will have digital TV
capability within three years seems optimistic. First, having several
competing standards is rarely promising for a fast take-off, and other
contenders may emerge as well. Then there's the phone itself to
consider: even if the cost of the TV chips is comparable to the expense
of adding a camera, an MP3 player or Wi-Fi, all these capabilities add
weight, cost, heat, and complexity while reducing battery life. There
are still limits to the amount of circuitry that can be shared among
these functions, and while memory prices are dropping, video is very
demanding of storage. All of this adds up to a high-priced,
short-battery life phone. Then, of course, users have to pay for the
monthly TV charge (Sprint charges $10 for MobiTV).

While cellular providers are keen on adding services to boost their
bottom lines, there are only so many directions that they and their
customers can afford to move at once. All the competing applications --
photography, music, games, data access -- seem to be a better fit than
TV is for the mobile realm. Cellphone users may occasionally find short
periods of time to watch the tubelet, but the mobile experience -- still
being primarily a professional one -- seems more oriented toward
on-demand clips than TV channel surfing. Few people have time to watch
TV while on the move, and TVs have become so ubiquitous in public spaces
that one of the hottest selling gadgets of late is a rogue device that
turns off nearby TVs. By 2006, the rare moments that cellphone users
will be inclined to watch video will also be the times that their
Wi-Fi-enabled phones will be in range of a high-bandwidth Wi-Fi access
point. And they may also be able to download the videos for later
viewing when they're back in 3G territory.

In short, don't be fooled by the mobile hype -- cellphone users may move
around a lot, but at the end of the day they still veg out at home or in
a hotel room watching a nice big TV (or big laptop monitor). Which
brings us back to the size question. Two-inch hand-held TVs have gotten
dramatically better in recent years, and the digital technology from
Texas Instruments, Qualcomm, and others is likely to be even better.
Yet, even if the resolution and frame rate improve, size matters in the
TV illusion. At two inches, details are still difficult to make out, and
it's a hassle to have to sit and hold your TV in your hand. Even with a
three-inch screen (about the biggest that's feasible on a phone), people
will watch it when the need arises, but it's less likely they'll be
hypnotized. That may be good for our souls, but not so good for the TV
business.