An informative article on Netflix in Bloomberg BusinessWeek

The Netflix article is a profile of the founder, Reed Hastings, and goes into some detail on their technology and strategy. Some of the points that struck me were:

• In spite of the fact that Netflix and Amazon are direct competitors in the IP video market, Netflix is hosted on Amazon Web Services.
• Netflix streams during the day and analyses data at night. They load servers between 2 and 5 AM local time. Shows they predict will be popular are served from flash storage.
• The master copies of all the shows and movies available to Netflix take up 3.14 petabytes of storage space. Netflix compresses the master files creating more than 100 different versions, each tuned for the varying bandwidth, device, and language needs of its customers. The compressed catalog is about 2.75 petabytes.

Check the article for more on the technology, Hastings and Netflix.

North American downstream traffic

-----

Update 6/18/2013

*Cogent says Verizon is not provisioning capacity to handle Netflix traffic*

From a GigaOm article at http://bit.ly/11XvMzn:

Cogent and Verizon peer to each other at about ten locations and they exchange traffic through several ports. These ports typically send and receive data at speeds of around 10 gigabit per second. When the ports start to fill up (usually at 50 percent of their capacity), the internet companies add more ports. In this case, through, Verizon is allowing the ports that connect to Cogent to get crammed. ”They are allowing the peer connections to degrade,” said Dave Schaffer, chief executive officer of Cogent said in an interview. “Today some of the ports are at 100 percent capacity.”

The Aereo local TV case -- new patents and conflicting experts

I've been watching Aereo's effort to stream local TV online since they launched and were immediately sued. They've just won a legal round -- a temporary injunction against them was denied -- but the case may continue.

I'm rooting for Aereo because I want to see all local TV online one day.

I'm a cord cutter, but I'm happy to pay for selected video content. I subscribe to Netflix and Amazon Prime and just paid a one time charge to stream the Tour de France. Fortunately, I can watch local TV over the air with a rabbit ears antenna. All my local stations transmit from the same mountain top, Mount Wilson, so I don't even have to play around with the antenna.

If I lived a long way from Mount Wilson or was surrounded by tall buildings in a city like New York, I would be out of luck with my rabbit ears, but I would be willing to pay Aereo or anyone else a reasonable fee for local channels -- either all of them or ala carte.

The law suit against Aereo claims they are rebroadcasting copyrighted material. Aereo counters that they've developed technology enabling them to assign each user his or her own dime-sized antenna, either permanently or dynamically when they log in. It is as if the user had mounted their antenna at Aereo's location instead of on their own rooftop.

Yesterday, Judge Alison Nathan of the United States District Court in Manhattan denied a request for a temporary injunction stopping Aereo from offering their service.

Aereo won a battle, but not yet the war. Their case still hinges upon the claim that those dime-sized antennas are independent of each other and each is rented to a different user.

When I wrote about Aereo earlier, I expressed some skepticism about their technology. Take a look at your rabbit ears or rooftop antenna -- it's a lot bigger than a dime. That's because optimal antenna size is determined by signal wavelength, and the optimal length of a rabbit ears antenna would be roughly one to eight feet, depending on which channel you were watching. The antenna would also have to be oriented correctly for best reception.

I'm not qualified to say that a dime-sized TV antenna is not possible with enough design and signal processing smarts, but it would be quite an engineering feat.

I was also skeptical of Aereo PR. When Aereo was launched, they stated that CEO Chaitanya Kanojia held 12 patents, but a quick check showed that none were related to antenna design. When I asked Aereo about patents at the time, they would not comment.

What have we learned since then?

I am happy to say that Aereo has now filed for four patents:

• 20120127363 - Antenna system with individually addressable elements in dense array
  
• 20120127374 - System and method for providing network access to antenna feeds
  
• 20120129479 - Method and system for processing antenna feeds using separate processing pipelines
  
• 20120131621 - System and method for providing network access to individually recorded content

I skimmed over the applications and noted considerable overlap. For example, three of the four applications have identical abstracts and the fourth abstract only differs by the addition of four words. The diagram shown above is also included in each application. Still, they have filed patent applications, which hopefully disclose a successful design for a very small TV antenna. The court has also had roughly eleven weeks for expedited discovery and briefing on the preliminary injunction motion. The focus was on the antennas. As Judge Nathan states in her opinion:

The only significant factual dispute concerns the operation of Aereo's antennas. Aereo contends that each of its antennas functions separately to receive the incoming broadcast signals. Plaintiffs assert that Aereo's antennas function collectively as a single antenna, aided by a shared metallic substructure.

Each side presented expert testimony, and the experts disagreed. The plaintiff's expert was Dr. John Volakis. In her ruling the judge summarized Dr. Volakis' finding that

the antennas do not function independently. Instead, according to Dr. Volakis, the antennas are packed on the board so close together that the incoming signal "does not see the loops as separate elements, but rather as one continuous piece of metal," the function of which is further aided by a common metal substructure formed by the circuit boards and the metal rails.

It's no surprise that the defendents experts, Dr. Pozar and Dr. Horowitz, disagreed. The judge wrote that

Dr. Pozar and Dr. Horowitz maintain that the construction of the antenna system requires the antennas to function independently, and Dr. Horowitz has observed numerous (if small) differences in recordings of the same program created by two different antennas. ... Moreover, tests performed at the Aereo site demonstrate that the signal received by Aereo's antennas is 1,000 times stronger than that needed for reliable reception.

Is it an array of small antennas or one big antenna? The judge concluded the discussion of the antennas by stating:

Based on the evidence at this stage of the proceedings, the Court finds that Aereo's antennas function independently. That is to say, each antenna separately receives the incoming broadcast signal, rather than functioning collectively with the other antennas or with the assistance of the shared metal substructure.

That had to make Aereo fans smile, but it is only based on evidence "at this stage of the proceedings."

We will see whether the case continues and, if it does, how it turns out.

As I said at the start, I am pulling for Aereo or anyone else who can get local TV streamed online. Let's assume Aereo's technology claims are indeed true and they prevail in this case. What then?

As much as I hope Aereo wins, they have created a kludge to work around the copyright laws. It would be simpler for local stations to stream their content themselves, eliminating the need for Aereo.

What might Aereo do if the local stations were to do that? For a start they might give up their antennas and do the streaming for the local stations. Their patents cover transcoding and indexing content -- they could provide streaming service to local stations.

Furthermore, if those antennas can really pull in a signal 1,000 time stronger than that needed for reliable reception, I want one. In fact I want more than one. Again, it would be a lot less kludgy to just sell the antennas to end users who cannot get local stations with a rabbit ears antenna, but could with an Aereo antenna. If the antenna works well, it could take over the indoor antenna market.

If the technology works, Aereo can pivot and win even if they don't make it as a streaming company.

Superbowl streamed -- pirates do IPTV better than NBC

When I reported that the Superbowl would be streamed this year, I saw it as a milestone on the road to IPTV.

I was one of the 2,105,441 people who watched the stream on NBC.COM. I tuned in out of curiosity, but after watching for a few minutes and taking a few screen shots, I turned my computer off and watched the game on TV. (NBC reports that the average veiwer remained online for 39 minutes).

I was not impressed. The action was in a small window surrounded by ads and statistics on a black background. The viewer could switch camera angles by clicking on the insert window on the upper right.

IPTV done poorly -- by NBC

It was too busy and too small for my taste. There were also seemingly constant commercials and other distractions. I did not tune in at the very start, but by the third quarter, the Internet stream was a quarter behind the TV broadcast.

On the other hand, spokesmen for NBC and the NFL were pleased.

Kevin Monaghan, SVP, Business Development & Managing Director Digital Media, NBC Sports Group was pleased by the "record traffic that grew throughout the event." He was also happy with "record high engagement numbers" referring to nearly two million camera angle switches.

Hans Schroeder, NFL, SVP, Media Strategy and Development called the live stream "a tremendous success."

They might have been pleased, but I expected more because I have seen better live streaming of a sporting event. The basketball game shown below filled the laptop screen and was identical to and only four secnds behind the TV broadcast shown behind it.

IPTV done better -- by pirates

One small hitch -- the basketball game was pirated.   NBC and the NFL should check out the music industry experience with pirates. A good way to stop pirates is to offer people convenient, high quality content at a reasonable price.

Perhaps TV executives should think of pirates as market research consultants who are showing them what the public wants.  NBC needs to learn from the pirates that the distinction between "TV" and "the Internet" is broken -- it's all bits.

I cut the cord -- no more cable TV

In this video, Verizon tells us the future of home video will be a wireless LAN connecting our TV sets and other devices to a Verizon FiOS server:



I agree that we will be distributing video around our homes on LANs, but don't expect mine to be connected to a FiOS server. For a start, Verizon does not offer FiOS in my neighborhood and from what I hear and read, they have no plans to do so.

Moreover, if they eventually do offer me FiOS, I suspect that it will be expensive and I will have to purchase a bundle of video "service" -- forcing me to pay for a lot of channels that I will never watch.

But, I don't want video service from Verizon, I just want bits.

I want my home LAN to be connected to the Internet (by Verizon or any other ISP), allowing me to watch ala carte IP video.

I've taken my first step in that direction. I “cut the cord" -- dropping our cable TV service and connecting our TV sets to our home LAN using Roku boxes. We (just barely) get local channels over the air using rabbit-ear antennas.

This set up and the available content is far from perfect, but it is my first step toward unbundled IP video.

Have you cut the cord? How do you like it?

100 YouTube tools and resources

If you make or watch YouTube videos, check out this YouTube toolbox . With over 100 tools and resources, there has to one you like.

If you particularly like one of these (or another that is not listed), let us know.

High definition video is coming to the Net

We have seen that new data types become mainstream as technology improves, and high definition video is now becoming available on the Internet.

YouTube is now offering "high definition" video, and NetFlix and others promise to follow soon. These still frames from a YouTube video illustrate the quality improvement.

YouTube claims their high definition video is 720p -- each image is made of 720 horizontal scan lines with every line refreshed in every frame. That would place the quality somewhere between an old style TV set (525 lines, alternatively refreshed every other frame) and high quality new TV sets (1080 lines, refreshed on every frame).

Since YouTube streams video, high definition requires a high speed Internet connection.

Would you be willing to watch 720 line video on a television set? Would you be willing to pay extra for a high speed Internet connection at home if you could watch 720 line movies and television programs? How might this effect movie and television production and distribution?

Consumer electronic devices on your home LAN

We talk about home LANs and their eventual convergence with consumer electronics like TV sets and audio equipment.

This New York Times article surveys some of the Internet-ready consumer electronic devices that were shown at a recent German trade show.

Do you have a home LAN? If so, is your TV, sound system or any other consumer electronic device connected to your LAN?

Computer science research leads to commercial products

Computer scientists at universities and the research labs of companies like Microsoft, Google or IBM work on new ways of doing things that may one day be used in commercial projects. In addition to theoretical studies, they often build hardware or software prototypes that test and demonstrate their ideas. (The slogan "demo or die" is common in such labs).

They also publish papers describing their work in technical journals and conference proceedings. One can get a glimpse of the future by reading those papers and attending conferences. For example, the 2008 conference of the Association for Computing Machinery Special Interest Group on Computer Graphics (SIGGRAPH for short) was just held in Los Angeles. Projects presented at SIGGRAPH foreshadow tomorrow's image and video editing programs. For a first look at features that might one day be in Photoshop or Windows Movie Maker, check these projects:

• Three-D photo viewer: combine many photos of the same object (like the statue of liberty or Notre Dame Cathedral) into a panoramic scene over which the user can pan and zoom.
• "Unwrap mosaics" for video editing: modify an object like a person's face in one still frame, and have the change propagated automatically throughout the video.
• Using photographs to enhance videos of a static scene: combine a video of a scene with a few still photographs to enhance the video by improving resolution, dynamic range, or touching up or removing objects.
• Face swapping: automatically replace faces in photographs.

You will find demonstration videos and technical papers at each of these sites.

The case of Pixar Animation illustrates the research-commerce relationship. In 1986 Pixar, then a brand new company, showed its first film, a short called Luxo Jr., at the SIGGRAPH conference. Watch Luxo Jr. and you will note things like the accuracy of moving light, shadows and reflections and the sinusoidal movement of the lamp cord when Luxo jumps. Achieving that sort of realism was state of the art research in 1986, therefore relevant for a scientific meeting. (The creation of a cute character was not science, but story telling art). The interplay between research and commerce continues -- Pixar employees presented a technical paper and a course at SIGGRAPH 2008.

Entrepreneurs and venture capitalists watch developments at research conferences like SIGGRAPH for ideas -- what are some other computer science conferences?

Free tools for working with Flash video files

Flash (.flv), probably the most common Internet video format, is used by Youtube and many other sites. If you download a video, you you might want to convert it to another format, strip off the audio as an .mp3 file, or edit it. Check out this list of free tools for processing Flash video.

Soon we will have Internet video cameras in our pockets

Video is now a common Internet data type and mobile Internet access devices are becoming available. You can see the convergence of these trends in this video of a test drive of a Tesla electric car:



The audio and video quality is poor, but it was shot using a Nokia N95 cell phone camera and streamed on the Internet as it was being recorded.

(The video is worth watching in its own right -- Tesla's founder discusses their technology and plans for the future while taking a joy ride).

We saw the Apple iPhone this year, and large phone vendors are now demonstrating their own mobile Internet access devices. Soon portable video cameras will be ubiquitous.

Ubiquitous, Internet-connected video cameras will be used for many business and personal applications. Will they also affect politics and government? Would it have mattered if the students at Tienanmen Square or the recent Tibetan demonstrators had been online? How many athletes and spectators will be online in China during the upcoming Olympic Games? How many people will be online in Iraq next year?

Synchronous collaboration -- working in different places at the same time

Several years ago -- before the Web was invented -- I was at the home of a Russian networking pioneer. I noticed that he had a computer in his living room that was connected over dial-up phone line to the Internet connection at his office. He told me he had called the office six months earlier, and, since there was no charge for local calls within Moscow, he never hung up the phone.

The cost of Internet connectivity is fixed -- you pay a flat fee for the month. That fee structure encourages same-time collaboration. Consider the way this programmer describes his work day.

He works with a colleague in a different state, but they remain in constant communication -- as if they were in the same room. Like my colleague in Russia, they open the connection between them when they arrive at work, and leave it open all day. He mentions using several networked applications -- Skype (VOIP), IRC (chat), Wiki (for documentation), and VNC (screen sharing) -- to facilitate collaboration.

(The recorded comment was made by an audience member at a panel discussion on attention).

But, can we have rich, emotional communication over the Internet? In 1980, artists Kit Galloway and Sherrie Rabinowitz connected larger-than-life video displays in Los Angeles and New York using a satellite link. They called the event "Hole in Space," and it was The Mother of all Video Chats. They demonstrated that, with sufficient bandwidth, emotion and presence could surely be communicated. Here are some video excerpts from their experiment followed by a short public-policy rant.

Television video is becoming economically feasible

We discuss the notion that new data types become economically feasible as technology improves. A New York Times article indicates that television quality video may soon be common on the Internet.

As shown here, over 100 million videos were streamed from four network Web sites during December, 2007.



The Times also reported that 2.7 million people watched the season four premier of The Office on the Internet and 9.7 watched it on television. (The Office is a dialog-heavy program, better suited to the computer screen than an action program). The article also mentions a survey conducted last October by Nielsen Media Research which found that one in four Internet users had streamed full-length television episodes online during the previous three months.

Note that these surveys were taken in the United States, where connectivity is mediocre by the standards of developed nations.

Have you watched a television episode on the Internet? If not, why not? If so, how was the experience? These surveys refer to television quality video. Would high-definition TV programs or movies be as common on the Internet?

Who needs the broadcast television networks?

The Los Angeles Times and the New York Times both ran recent stories on entertainment programs being produced for the Internet by successful, mainline TV producers and writers. They are attracted to the new medium by creative control and cost savings on distribution. You can check episodes of two of them at The Fantastic Two and Qarterlife.

For more insight into the background of TV production and what is motivating these successful men to move their talents from television to the Internet, see the companion LA Times article Are Corporate Suits Ruining TV?". (You can guess the answer to that question).

Watch a couple episodes from one or both of these programs. How do they compare to TV? How do you think they could be improved?

Crescent Heights -- the future of television?

Proctor and Gamble, which makes Tide laundry soap and pioneered radio and television soap operas, has an Internet soap opera called Crescent Heights. The three minute episodes are designed to be seen on a mobile device or a computer, and there are no commercials. The Tide logo appears at the end of an episode and, as you see in this screen shot, the characters use it for their wash. The key advertising value is in a set of forums where members can discuss the characters, the story, and household topics like stain removal and washing. The discussion site mentions Tide products and has a Tide logo.

You can read more on Crescent Heights and the business proposition here.

In this case, the Internet lets the advertiser produce and distribute their own content rather than sponsoring a show produced by someone else and distributed by a cable or broadcast network. Would this cut cost? Is this a rough, early glimpse at the future of television?

As mobile and home connectivity improve, how will a program like Crescent Heights improve? Would you be willing to watch a football game sponsored by Budweiser on the Internet? Do you think people would be more likely to purchase Tide soap after watching Crescent Heights and looking at the online forums?