Wednesday, August 29, 2012
Well, that restriction is starting to fade away. Ting announced that during the next quarter, they will begin allowing customers to use their own Sprint phones. That means the selection of new phones will increase dramatically and old Sprint phones will be welcome as well. They point out that, "at first", there will be restrictions -- the iPhone, BlackBerry push to talk and a small list of specific handsets won’t be included.
If you check Ting's prices, you will find significant savings over the major cellular companies. Now that you can bring your old Sprint phone or choose from a wide selection of new phones, Ting (and other MVNOs) seems like a terrific deal for those in areas with good Sprint coverage. The deal is even sweeter if you live in an area which is or will soon be covered by Sprint LTE.
I don’t know how Ting does it. Perhaps Sprint’s wholesale price is very low or perhaps Ting is less greedy and more efficient than the major carriers. Somehow, it offers a better retail deal than Sprint or any of the others. Maybe it is just that it charges more rationally than its competitors. Since 2009, U.S. data traffic has exceeded voice traffic and the gap is growing rapidly, but the cell phone companies still gain most of their revenue from voice.
The cellular and cable companies seem to have arrived at a "gentleman's agreement" to reduce what little is left of Internet service competition. The cellular companies are getting out of the land line business and the cable companies are getting out of the mobile business.
Will the MVNO's like Ting succeed in disrupting the mobile market?
Monday, August 27, 2012
MOOCs and other online classes give us an opportunity to learn what works in presentations. For example, I've proposed studies of the effect of playback speed or including a talking head in a presentation on comprehension, retention and engagement.
I'm not sure how these studies would turn out (though I have my suspicions), but I am sure that talking heads are important in coversation.
My certainty is based on a single anecdotal experience (of no scientific value). Some friends and I meet for a weekly podcast bull-session on technology (Yet another tech show). We do the podcast using a Google Hangout.
As you see here, video thumbnails of all the participants are shown under a larger head-shot of the person who is speaking. (Switching is automatic). But, check the thumbnails and you see that the second from the left is blank. That is because one of us had a slow Internet connection that night and could not stream video. We could hear what he said, but not see his face.
If you had asked me before we did that podcast whether not seeing him while he spoke would matter, I would have said "no" -- figuring that the audio conveyed all the information in a technical discussion.
If you ask me now, my answer is "yes, for sure." It had nothing to do with information or content. We understood what our invisible colleague was saying and he participated fully, but it felt weird.
Talking to disembodied people on the phone feels natural because there is no point of reference. But, in a mixed conversation, where you can see some, but not all of the participants, the difference is palpable. I don't know if it is because of subtle clues as to who gets the floor next, some primitive fear or something else, but it feels more natural to talk with someone you can see.
Saturday, August 25, 2012
I would like to see studies testing the hypotheses that seeing a talking head or a hand writing on the screen improves student comprehension and retention.
I've developed a collection of modules for my digital literacy course. Each module contains an annotated PowerPoint presentation and other material. As I find time, I record videos of the PowerPoint presentations, using the annotation as a script.
To conserve bandwidth and storage space, I only use still images of the PowerPoint slides in my videos. There is no video of me speaking. I reasoned that my students knew what I looked like, and viewers who were not in my class could see what I looked like in a photo.
But, I notice that Google (left), Coursera (right) and others include video of the speaker's face during their presentations.
Is there evidence that talking heads improve comprehension and retention?
We see a similar contrast in the screencast videos produced by Udacity (left) and the Khan Academy (right}. Udacity videos are screencasts with superimposed video showing the teacher's hand and pen while drawing and writing on the screen (with noticeable parallax). Khan produces screencasts in which only a cursor is shown.
Does seeing the hand and pen improve comprehension and retention?
(If these research questions sound interesting, see a proposal for a related study of the effect on comprehension and retention of varying playback speed and bullet point images).
Friday, August 24, 2012
Here is how the industry sees the deal:
Dan Mead, president and chief executive of Verizon Wireless: “This purchase represents a milestone in the industry, and we appreciate the F.C.C.’s diligent work to review and approve the transaction.”
David L. Cohen, executive vice president of Comcast: “a smart and efficient way for Comcast to deliver a broader array of wireless services, and is an efficient deployment of this spectrum.”
And here is the opinion of Gigi B. Sohn, president and co-founder of Public Knowledge, a nonprofit organization based in Washington that promotes an open Internet:
“By allowing Verizon and the cable companies to sell each other’s services, the D.O.J. and the F.C.C. are acknowledging what has been clear for some time — that broadband competition policy in the United States has failed."
I have to agree with Gigi Sohn. The US congress tried to foster competition with the 1996 Telecommunication Act. The idea was that the incumbent telephone companies would allow competitors to use their facilities at wholesale rates to compete as retail level service providers. The telephone companies drug their feet and fought at the state and local level to keep that from happening, and they succeeded.
This is what William Kennard, who, as chairman of the F. C. C. from 1997-2001, was charged with implementing the Telecommunications Act, had to say near the end of his term:
“All too often companies work to change the regulations, instead of working to change the market.”
“[regulation is] too often used as a shield, to protect the status quo from new competition - often in the form of smaller, hungrier competitors -- and too infrequently as a sword -- to cut a pathway for new competitors to compete by creating new networks and services.”
“companies invest in lawyers, lobbyists and politicians, instead of plant, people and customer service.”
Some have argued that wireless connectivity would eventually bring competition to the Internet service industry, but it seems we have instead a gentleman's agreement in which the cable companies will provide land-line connectivity and the cell phone companies will provide wireless connectivity. Furthermore, sometimes they can sell each other's service.
My experience illustrates the gentleman's agreement. I was a Verizon DSL customer for years. I was about to leave them when I started seeing ads and billboards saying I would soon be able to get their fiber service, FIOS. After a couple years wait, they announced they would not be bringing FIOS to my neighborhood after all. Then they cut my DSL speed from 5 to 1.5 mbps! I switched to cable -- my one alternative.
My current slim hope for getting connectivity that close to what is available in Riga Latvia is Google's pilot in Kansas City.
The FCC okayed Comcast's deal with Verizon, but maybe they are getting tougher now. They've put the Time Warner Cable merger on hold:
In a public letter to Comcast and Time Warner Cable executives, the FCC announced that it is hitting the pause button on its review of the proposed $45 billion merger. Citing inadequate responses by both cable companies to earlier FCC requests for additional information, the agency is stopping the clock on its 180 day review period until late October.
The end game has been revealed -- Verizon will exit the wireline business within ten years:
A utility company familiar with Verizon's business plans says Verizon has told it the company plans to exit the fixed-line broadband business entirely within the next ten years. Stop The Cap directs our attention to a filing at the FCC made by Florida Power & Light, urging the FCC to reject Verizon's decision to offload its Florida, Texas and California DSL, FiOS and POTS customers to Frontier Communications.I guess they are counting on 5G cellular, but I hope public WiFi hotspots give them some stiff competition.
During a recent earnings call, Verizon CFO Fran Shammo said that FiOS will be available to about 70% of their east coast, fixed-line customers once the company finishes the sale of their Texas, Florida and California fixed-line networks to Frontier Communications.
If they stick to their plan of leaving the wireline business, they will have 100% penetration the day before selling off the last customer.
While no further sales are imminent, Shammo stated that all options remain on the table. "As I have always said you never say never," Shammo noted. "We're always open to new options and we always keep our eyes open for that but at this point we're very committed to the wireline business."
Verizon is living up to the aforementioned gentleman's agreement with Comcast. In my personal case, that means Time Warner Cable will remain my only viable Internet service provider until Google or Elon Musk or someone else comes to my rescue.
Wednesday, August 22, 2012
Google has been awarded US patent #8,239,662 for "a system for providing an operating system over a network to a local device."
The patent describes the system architecture and goes on to make 3 independent and 15 dependent claims for protection, but how unique are those claims? There have been several "thin client" products and Oracle marketed their Network Computer. Furthermore, is downloading a program that is an operating system conceptually different than downloading a program that is a Web browser or word processor -- where exactly do you draw the line between an operating system and an application program?
I spoke with intellectual property attorney Thomas Ewing of Avancept LLC about those 18 claims and he pointed out that "the number of claims a patent has is only very loosely correlated with its inventiveness." He added that while multiple claims are common in the US, they are less so in other nations and that to infringe upon a dependent claim, one must also infringe upon its corresponding independent claim.
He noted that the patent examiners initially questioned the novelty of several claims in Google's application and it took two rounds of rejection and appeal to finally get the patent, but characterized this as a "fairly typical back and forth with the patent office." He also noted that the patent examiner cited references to prior art in 1 issued patent and 6 published applications (listed below).
Ewing said that to infringe a claim literally, someone would have to do everything listed in the claim. I would think that engineers working on a competing product could find work-arounds that would differentiate their work from Google's. “There is a possibility for arguing infringement under what’s known as ‘the doctrine of equivalents,’” he said, adding that such arguments are often difficult to win.
Ewing also mentioned that it is unusual for someone as senior as David C. Drummond, Google's Senior Vice President, Corporate Development and Chief Legal Officer to sign the power of attorney for a patent application. He acknowledged that that may be standard procedure at Google, but, if is not, it may give us an indication of their commitment to the Chromebook.
With their Motorola acquisition, the Chromebook, Google TV, Nexus tablet and the quickly withdrawn Nexus Q, it looks like Google will end up a hardware manufacturer, heeding Alan Kay's advice that "People who are really serious about software should make their own hardware." They may also have noticed Apple's profit margins.
If you would like to really geek out on this patent, you can see all of the accompanying documentation by requesting information on patent number 8,239,662 at the Patent Office portal. I did so, and was impressed (and depressed) by the effort needed to secure the patent -- there are records of 46 transactions, 69 documents on file, 63 adjustments (some minor) made during the application process and contact information for 82 attorneys and agents (though only two attorneys actually filed the documents in the case).
Whether it is eventually upheld or not, this patent gives Google and its lawyers a weapon to use in the sad court fights that are so characteristic of today's computing scene.
Prior art referenced by the patent examiner:
- 7,343,560 - Method and system for generating dynamic images - owned by Novell
- 20030126242 - Network boot system and method using remotely-stored, client-specific boot images created from shared, base snapshot image - owned by Compaq/HP
- 20040153694 - Reliability of diskless network-bootable computers using non-volatile memory cache - owned by Microsoft
- 20050091349 - Automatically configuring a computer - owned by SAP
- 20060010314 - Methods and systems for running multiple operating systems in a single mobile device - owned by Intellectual Ventures
- 20060047946 - Distributed operating system management - owned by MaxSP Corp.
- 20090083404 - Software Deployment In Large-Scale Networked Systems - owned by Microsoft
I watched both the BBC and NBC Olympic coverage on the Internet and the BBC did a better job than NBC. Four BBC pluses come to mind.
1. The BBC Web viewer was more interactive, less passive and TV-like, than NBC's and they are continuing to improve it. As you see here, the user could display and hide data on demand and jump to the start of a "chapter" corresponding to an event.
2. I watched the BBC stream via a VPN, which, as shown below, slowed my connection considerably.
In spite of the increased latency and decreased speed, the BBC's video experience was as watchable as the video when using YouTube's content delivery network in the US.
3. Many of the NBC events I watched were video feeds with no commentary and every event I watched on the BBC had commentary. That could have been random chance and I watched more NBC than BBC, but it was my experience. (Commentary is more important for some events than others and some people dislike it. Users should be able to toggle commentary and the ambient sound on and off).
4. The BBC was commercial free -- paid for by the British TV license fee.
The BBC also wins the gold medal for Olymic blogging. The developers of BBC sports have a blog with suprisingly informative posts, and around 20 of the recent posts pertain to the Olympics.
Here are snippets from a couple that caught my eye.
Phil Fearnley, General Manager, News and Knowlege at BBC Future Media outlined their goals for the Olympic coverage and placed it in historical context.
The goal was to deliver over 2,500 hours of live sport online via 24 high definition streams - to cover every sport, from every venue on every day – across four screens: PC, mobile, tablet and connected TV. There were not 24 events going at all times, but the screenshot below was taken at a moment when there were -- the user could choose among 24 video streams.
Finley saw this as the start of a new era, writing
Broadcast television’s first big moment was the coronation in 1953, which brought the nation together around the TV screen for the first time. Our aspiration is that 2012 will do for digital and connected televisions what the coronation did for TV.
Cait O'Riordan, Head of Product, BBC Sport presented some statistics and analysis of the site traffic.
Below, you see the number (millions) of people watching the six streams with highest peaks on the first of August. People could easily switch between streams, and chapter links on the time-line let them find specific events they were interested in.
This graph shows relative usage during the day for each of the four screen types between July 28 and Aug 9. It looks like people watched on their computers (at work?) during the day and took their tablets to bed during the evening.
Matthew Clark, the Senior Technical Architect for BBC Online's Olympic website and apps described the site architecture. Here is his architecture diagram -- frustratingly small and impossible to read, but it conveys complexity:
In the post, Clark states that "Over the coming months we'll be working on bringing many of these features to the rest of Sport, and perhaps other parts of BBC Online too."
Alex Perry, Project Manager for the Live Interactive Video Player, outlined its key capabilities, including:
- Easy switching between up to 24 live streams at any time
- The ability to pause and rewind live video or jump straight to key moments you may have missed by using chapter markers (eg the Men's 100m final)
- Alerts for the key events that have happened or are coming up so you don't miss the moments that matter to you
- Extra facts and info on the sports and competitors you are watching alongside the video
this gives you a taste of the blogs -- if you are a live-stream event coverage geek, be sure to check them out.
Why did BBC do better than NBC? Two things come to mind. BBC's non-commercial culture makes a difference -- the viewers are their customers while NBC's customers are their advertisers. There is another clue in two of the tags on the BBC blog posts open source and open standards. Those tags indicate that BBC "gets" the Internet, and maybe NBC doesn't.
Regardless, the BBC team gets the gold medal for streaming live events.
(I've written several other posts on NBC's coverage of the Olympics and the Tour de France. You can see links to my observations as well as predictions for future event coverage here).
The bar chart below sums it up -- Amazon's Glacier is cheap -- 1¢ per gigabyte per month. (Follow the link for pricing detail).
Amazon has published a Glacier API, and, for now, you have to be a programmer to use it. But they have committed to releasing FTP access very soon, and it will not take long for developers to release backup programs and services with graphical user interfaces to Glacier.
Friday, August 17, 2012
For this, the short version, I will list a few observations and conclude with some predictions for the future of live-event coverage on the Internet. First the observations, in no particular order.
- There were examples of three sources of revenue for live streaming of events: the BBC's coverage of the Olympics (free with a license fee), NBC's coverage of the Tour de France (pay per view) and NBC's Olympic coverage (advertising). My favorite was pay-per-view, but there is a lot of room for improving NBC's advertising presentation and many people would rather pay by watching ads.
- The BBC's Internet coverage was better than NBC's. It was more interactive, had better video quality and commentary and no advertsing. BBC also has a terrific sports blog. They win the gold medal because the viewers are their customers while NBC's customers are their advertisers. BBC "gets" the Internet.
- NBC could learn something about Web design from BBC. The BBC Web site was more interactive, giving the user control, while NBC's more closely followed a passive television model. They also need to clean up glitches like broken links.
- NBC's Olympic coverage needed better direction -- adding expert commentary to all the streams, better timing of commercials and intelligent camera selection.
- Sadly, NBC deleted the archives of the Tour de France soon after end of the race. I believe they have an information stewardship obligation and should maintain the archives of important events for analysis by journalists, scholars, fans, remixers, etc. The cost of doing so would be low.
- NBC's Olympic archive is still online, but they have not said for how long. Since it contains ads, they may leave it up. The BBC archive will remain online until January.
- NBC allowed viewers to watch two streams at once (one in a larger window), but then they had to watch two streams of ads.
- The quality of NBC's Web stream was very poor at first, virtually unwatchable. The video quality using their phone and tablet apps was much better. The video quality of NBC's earlier streaming of the Tour de France was nearly as good as that on television.
- NBC's Web video quality seemed to improve during the Olympic games. I did a subjective evaluation on day nine and, while it was far from perfect, it was in the same ballpark as BBC's streaming coverage. (Note that I watched the BBC coverage using a proxy server, which limited my download speed to about 2.3 Mbps while I was getting around 15 Mbps when accessing NBC).
- NBC's video seemed to improve even further by the end of The Games. Looking at various performance tests during the games, I tentatively concluded that the bottleneck was in the network, although there are too many variables to be sure.
- NBC made changes to their user interface during the Olympics. They were acting a bit like an agile Internet start-up -- release early, revise often -- tweaking the network to improve performance and revising the user interface.
- The Olympic Organizing Committee did its best to control leaks of Olympic material. For example, WiFi hotspots were not allowed in the stands and they did their best to stop social media leaks.
- NBC did a much better job on the Tour de France than the Olympics, but it was an easier event to cover, they had experience covering it in past years and they did not have to deal with the presentation of ads.
Now for a few fearless predictions.
NBC paid the International Olympic Committee $4.38 billion for the media rights to 2014 - 2020 Summer and Winter Olympics. We will see major changes in the coverage and the industry by 2020.
There will two separate markets. The first will be for producing content -- video, commentary, interactive features, Web design, and so forth. NBC was the content producer in this case. The second market will be for efficient networking to distribute that content. During these Olympics YouTube (Google) provided the networking expertise.
Both the content producers and the networkers will have learned a lot from past experience and will do a much better job than they did this year. We have even seen improvement during these Olympics. This is just the start of the IP TV era -- it is like the crystal set era in the early days of radio.
Given the expected improvement, the majority of people will be watching live events and other video on the Internet by 2020 -- we will have cut the cord and be watching over the top.
While there will be both network and content industries, the content industries will not have a lock on production. As the comedian Louis C. K. showed us, the production company will be dropped in some cases. Louis C. K. decided that he could do a better job of producing his comedy DVDs than Sony, and the Internatinal Olympic Committee might decide that it can do a better job than NBC or another production company.
It will be interesting to see what Google does in this area. If their fiber pilot study in Kansas City looks promising and replicable, they may become a networking powerhouse. At the same time, they are developing content and learning by working with folks like NBC. They may end up being able to offer the whole package.
Of course, there is a disclaimer to these predictions -- never underestimate the power of well-funded, entrenched lobbyists and huge corporate legal staffs. On the other hand, it is hard to find good buggy whips these days.
Google has posted the content of their Internet search MOOC. It consists of six 50-minute classes, and I went through the first one, to see how Google is approaching MOOCs.
The class consisted of an introductory video followed by five lessons with activities for each lesson.
Four of the lessons were videos showing screenshots and screencasts with a talking head in a small window, as shown below. The fifth lesson was a professionally produced video on how search engines work.
Each activity started with a boxed text summary of the points made in the associated lesson, as shown here.
The summary was followed by three or four activities. There were three types of activity:
- Multiple choice/true false questions, followed by the answers
- An admonition to think about something or answer an open ended question before proceeding (for example "What is something you have wanted to find that color filtering might have helped you locate faster? Reflect on this before proceeding to the next lesson.")
- An open ended question, followed by the response of an "expert" to the same question
#mooc #digitalliteracy #edreform #highered #bonkopen #pedagogy #edtec
Some examples I found in an article in the Chronicle of Higher Education are:
- Using Meetup.com to form local face-to-face (FTF) groups
- Weekly Google Hangouts with the professor and interested students (John Boyer also uses weekly online office hours to supplement his massive on-campus class)
- FTF “office hours” in different cities
- Student organized groups on Google Plus
- CompScisters, a Facebook group to encourage women to take and complete MOOC’s in mathematics and science
Thursday, August 16, 2012
1. Peter Norvig, The 100,000-student classroom. This is a six-minute TED talk describing the MOOC Norvig and a colleague taught at Stanford along with some of the principles that guided the course design.
2. A twenty minute audio interview of George Siemens, a professor at Athabasca University. Siemens was one of the first to teach a MOOC. (He may have coined the term). He describes his decentralized approach to MOOC design and, like Norvig, focuses on the need and opportunity for pedagogical research.
3. A twenty minute Ted Talk by Coursera co-founder Daphne Koller presents their MOOC techniques and highlights a few pedagoical research results.
4. Two blog posts that give a sense of the video production process at Udacity, a prominent, venture funded startup:
- Katy Reichelt, video editing manager at Udacity: Video editing and the learning process
- John Regehr, Associate Professor of Computer Science at the University of Utah describes the Udacity production process from the point of view of a teacher.
Disclaimer -- these are all presentations by enthusiasts. I've written several posts on MOOCs in the last year or so, and see a lot of potential, but there remain many unanswered questions.
With the Olympics and Tour de France, we saw examples of three live streaming revenue models: license fee, advertising and pay per view. Here is a quick look at each.
BBC Olympics, license fee
The BBC streamed 2,500 hours of free, live coverage, but you had to have a British TV license. The TV licenses cost £145.50 per year for a color TV and £49.00 for a black and white TV. The fee allowed you to watch the video streams on any device and gave you permission to store, but not distribute, the content. If you came to the site from a non-British IP address, you were blocked, but that was easily circumvented using proxy servers. In the future, they may be more aggressive about discovering and blocking proxy servers.
Archived footage will be available until January.
NBC Olympics, advertising supported
To view their live coverage of the Olympics, NBC required you to have a cable TV account with an authorized provider and a subscription level with access to several NBC cable channels. (That restriction may have been imposed by NBC's owner, Comcast). On top of that, they showed a lot of ads.
Like the BBC, authorized users were free to view the live coverage on the Web or their tablets and phones. One could circumvent the account restrictions by using the user name and password of a friend with a qualifying account. Again, in the future they might be more aggressive in blockng that sort of access.
NBC's archives are on line at this time, but they have not announced their plans for future access. Since they run ads, they may leave them up indefinitely.
NBC Tour de France, pay per view
For the Tour de France, NBC charged a $29 fee, which, as with the other events, entitled the user to watch on the Web, a phone or tablet. There were no ads, but, unfortunately, access to the archive footage was dropped shortly after the end of the event.
My least favorite this year was the ad support model, which was done poorly. When navigating the Web site, it seemed that every new page view had a pre-roll commercial, and arbitrary commercial breaks in the middle of event coverage were irritating and distracting -- the screen would go black for a few seconds in the middle of an event, then they would run as many as six commercials before resuming the action. To add insult to injury, they boosted the audio volume during commercials.
I preferred pay per view, but many viewers would opt for ads, and there is a lot of room for improving the experience. For a start, NBC needs directors to watch the action and insert ads intelligently. Ad targeting -- showing individuals ads based on demographic, social networking and search signals or asking them for hints as to the kinds of ads they were interested in -- would also improve the experience. There could be hybrid models like allowing the user to choose between pay per view with no ads and paying a smaller fee and seeing just a few, well-placed ads.
Wednesday, August 15, 2012
I am attracted to the big picture -- studies of the global diffusion of the Internet. For example, Telegeography tracks the spread of the physical infrastructure in the form of undersea cables and Akami tracks use and traffic in their quarterly reports on the State of the Internet.
Akami's report for the first quarter of 2012 is just out. It is a detailed report, but a couple of big picture items caught my attention.
Akami saw a 6.0% increase in the number of unique IP addresses connecting to their global content distribution network over the previous quarter, and the global average connection speed was 2.6 Mbps.
They also noted the continued growth of mobile data, which passed mobile voice during the fourth quarter of 2009.
Check the full report for a lot more of this sort of thing and to drill down into the details.
For old timers, let me throw in links to a couple of earlier papers on Internet diffusion by my colleagues and me:
An Internet Diffusion Framework, CACM
The State of the Internet: Growth and Gaps, ISOC 2000
I wrote an earlier post on the production and direction of NBC's Olympic coverage. That note focused primarily on the production of their video content, but we also noted a few Web site production problems like broken links, slow response time and screen blackouts during transitions to commercials.
Many of these glitches depended upon the network as well as the Web design, but they should be able to improve both of those based on their experience this year.
In addition to fixing those problems, they can take some interface ideas from the BBC's Olympic Web site, which was more interactive than theirs. NBC was more like television -- the viewer watched the video passively, and NBC producers popped up extra information like standings and scores on the screen when they decided to do so. (They also popped up commercials -- something the BBC or Tour de France producers did not have to deal with).
The BBC gave the user more control. If the user clicked on Extras after hovering over the bottom of the screen, they could display information about the athletes and the game. They could also control playback using a chapter-oriented time line.
In the example shown below, the video timeline is divided into "chapters." In this case, the chapters correspond to pre and post-game material and quarters or parts of quarters of the game action. The user is watching the second video chapter (yellow) and has displayed the game statistics of one player, Carmelo Anthony. The chapter-oriented timeline and Extra choices gave the BBC user more control than the NBC user had.
The folks on the NBC Olympic crew can also borrow from their Tour de France team. For example, I really liked the 15-second instant replay button on Tour videos. Again, this example shows a way to give control directly to users, who can replay as much or little as they wish.
The user interface shown above was BBC's default, but they offered two optional user interfaces. The accessible player had text-oriented controls that were designed for keyboard and screen reader access. They also had a pop-out player for people who wanted to keep the action streaming in a small window while the worked on something else.
This is not to beat up on the NBC Olympic team. The Tour de France team had a simpler event to cover -- a single race comprised of stages that all followed the same format. Furthermore, they had previous experience covering the Tour and neither they nor the BBC crew had to deal with commercials.
That being said, I am sure NBC will look back at what they and their competitors did this year, and improve their site design.
Monday, August 13, 2012
The Olympics were brought to us by two organizations -- YouTube and NBC. YouTube ran the network, and their task was to deliver smooth, high-resolution video streams. NBC handled content production and direction, using video feeds supplied by Olympic Broadcasting Services, OBS. Most of my posts to this point have focused on the former -- the network. How about direction and production?
While some of the content was good, there were significant gaps.
For example, a cycling road race is a difficult event to cover -- it is spread over a large area, the action is continuous and the cyclists break into groups. NBC streamed the video of this and many other events without commentary and with relatively few on-screen graphics. While that may be satisfactory for an event like volleyball or weight lifting, one tends to get lost in a cycling road race.
The camera might be focused on a single group of riders, but, who is in the group, are some of the race favorites or top sprinters in the group and where us the group relative to the rest of the field? Commentators and graphics like NBC used in their coverage of the Tour de France provide that context. (I spent less time watching the BBC's live coverage, but all the streams I saw had live commentators).
Extended tree-obstructed shots like the one shown below are evidence of loose direction. I do not know whether NBC had the ability to switch cameras or they only got one video feed from OBS, but someone should be watching and switch cameras when the view is obstructed or there is something else of more interest to show. Ideally, the person switching cameras would also be in communication with the commentators.
Cuts to commercials at arbitrary points in the action were also evidence of loose direction. Furthermore, seeing a commercial on a TV set is not the same as seeing it on a laptop screen. We subconsciously expect commercials on TV and the transitions are fast, but on the Web they are jarring. You think something is wrong at first when the screen goes black during the transition to a commercial.
NBC's Tour de France video was superior to that of their Olympic video, but that was to be expected. The Tour de France video feed was supplied by France television, working with Euro media France (EMF). France TV and EMF have years of experience, while Beijing was the first Olympics for OBS. They also had only one event to cover, so needed only one set of experts and commentators. Finally, since the coverage was pay-per-view, they did not have to deal with timing of and transition to commercials.
There were also production problems with respect to the Web site -- broken links like the one shown below, slow response, poor user interface design, screen blackouts during transitions to commercials and so forth. (The slow Web site response time is a shared problem between the Web site design team and the networking team from YouTube).
I don't mean to beat up on NBC -- covering 302 events in 32 sports is a huge task, and I am sure they learned a lot about how to produce live events during the Olympics. They even made some mid-stream adjustments like changing the user interface on the live stream menu and perhaps adding more commentators near the end of the games. I cannot be sure of the latter, but it seemed like after the first week or so I found fewer live events without commentary, and the commentators often had English accents.
As video coverage of live events moves online, the production experience NBC and OBS gained at the Olympics and the Tour de France, will serve them well.
You will find a description of EMF's operation during the Tour de France here.
I didn't use Dabbleboard for production work, but it was a terrific tool for teaching students the difference between vector and bit-mapped images and illustrating synchronous collaboration.
This illustrates one of the risks of cloud computing -- you are stuck if you depend upon a service that goes out of business, raises its prices, etc.
I wonder if the folks at Dabbleboard would consider open sourcing it?
|Tiled UI with over 4,000 choices|
There are full events, event highlights, interviews, expert analysis, footage from earlier Olympics, medal ceremonies and more. You can filter the videos by these categories, sport, or top athlete and sort them by by date, popularity or recommendation.
But, what about that user interface? I would have called it a Metro-style user interface, but I guess Microsoft has dropped the name "Metro." Maybe a better name would be square icons user interface (SIUI)? Or tiled user interface (TUI)? Or cell-phone user interface (CPUI)? Or an iPhone user interface (iUI)? (Pronounce each acronym and see what you think).
Seriously, whatever you call it, this sort of user interface might be more appropriate for a cell phone or tablet with a few installed apps than a library of over 4,000 videos. (On the other hand, it lends itself nicely to using oversize tiles for ads).
As I noted in a post the other day, NBC deleted their Tour de France archive. I asked their PR department what the plans were for the Olympic archive, but got no comment. Maybe they will leave this one up as ad bait.
I want to add a note on the quality of the archival video streams. I expected that it would be better than the original broadcast stream because demand would be lower, YouTube would have had time to push content out to caches around the Internet and there would have been time for additional compression.
To test my theory, I watched the replay of the final basketball game between the US and Spain at 1080p resolution. The streaming was a bit jerky, but the game was watchable.
I could not subjectively say whether it was better or worse than the original live stream at that resolution, so I looked at the performance profile while streaming from the archive (below). As you see, it is considerably different than the live streaming profile.
Whether it is better or not, I am confident that YouTube and other major distribution networks will learn from this Olympic experience and improve performance.
I also tried the BBC archive. The BBC provided up to 24 simultaneous live streams adding up to 2,500 hours of coverage during the Olympics. That material has all been archived, and it will remain online until January.
The BBC streams their video at a variable rate, starting slow and adding speed as buffering and bandwidth allowed. I watched the final basketball game between the USA and Spain. The motion was fairly smooth. There was some stuttering, but less than with NBC. However, the resolution was lower, not exceedinig 512 x 288. Bear in mind that I was watching via a proxy server. The average of three download speedtests was 2.34 Mb/s, about 20% of my local download speed.
At this speed, the following is representative of the best image quality I saw.
Like YouTube, I would expect the BBC network to improve as a result of this experience. I wish they would reconsider the decision to delete the archive in January.
|All that is left at the Tour de France Web site.|
I just went back to the site, and discovered that the archive is gone. That is a shame, because it would have been a valuable resource for historians, journalists, fans and remixiers.
What a waste. Doesn't NBC get the Internet? Haven't they noticed Amazon's success with low-selling, long tail books and other retail items?
They need to talk with Brewster Kahle, who founded the Internet Archive to build an Internet library to offer "permanent access for researchers, historians, scholars, people with disabilities, and the general public to historical collections that exist in digital format."
In endorsing the Internet Archive, Stewart Brand wrote that "civilization is developing severe amnesia" due to the rapid turnover of digitized information, that the archiving "norm is total loss."
NBC -- how many early TV video tapes and kinescope recordings have you lost? Don't keep repeat your past mistakes.
Added notes: NBC's Olympic archive is still online, but they have not said how long it will remain there. The BBC Olympic archive (up to 24 live HD streams and 2,500 hours of coverage) will be available until January.
I just learned that the January date was not decided by the BBC, but imposed in their contract with the International Olympic Committee (http://www.bbc.co.uk/blogs/bbcinternet/2012/08/digital_olympics_reach_stream_stats.html?postId=113427062#comment_113427062).
I understand that one can argue that the IOC paid a lot of money and should have the right to maximize their profit on that investment, but I also feel that there is a class of event that should be kept available as part of our cultural and political history for which there is a stewardship responsibility. Do not know where to draw the line, but in this case, I would urge the IOC to plan on keeping the archive available to the public and adjusting the terms of their contract to take that into account.
Sunday, August 12, 2012
I like the ability to watch two events at a time, as shown here. The audio and video streams are independent of each other, and the user can swap the large and small screens with a single click.
There is a menu listing the currently available live streams below the right hand screen, and the user can switch among them with a single click.
There is buffering time whenever you make a change and the two streams compete for bandwidth, but it is a welcome feature with a well-designed user interface.
My gripe is with the advertising. When an ad pops up in one screen, the other stops and displays the message shown below.
I guess stopping the video preserves bandwidth for the HD commercial and forces the you to pay attention, but there is a major glitch -- it pisses the viewer off. I may never buy a Coke again, and, if I do, it surely won't be in one of their special Olympic Games cans.
Thursday, August 09, 2012
At the start of the Olympics, NBC's live video stream was unsatisfactory, but we noted improvement by day 9.
More good news -- the video seems even better now. On day 12, I was comfortable watching events streaming at 1080p at full screen on my Dell laptop with 4 GB of RAM an Intel Core 2 CPU with a 3.06 Ghz clock speed and an Nvidia Quadro FX 770M display chip.
This is not to say the video was perfect. It stuttered a bit, but was watchable. As you see below, there were thousands of dropped frames:
When the video was running smoothly, the frame rate remaiined in the low 20s per second, but during stutters, it dropped to single digits. I played around with resolution a bit, and was surprised to see that the dropped frame rate did not change much when I switched from 240p to 1080p. I checked the Task Manager, and noted some, but not much, increase in processor load variablity when displaying 1,080p (shown between the arrows).
Based on this, it looks like my old laptop is not limiting video quality, and my next laptop will be faster and will not be running Flash. The link from my house to my ISP runs at around 15 Mbps, so that is not a problem. It seems like the frame-rate variability is in the network.
But, given the compression and network improvement NBC and their streaming partner YouTube have made during the last 12 days, I am confident that by the time of the next Olympics, Internet video will be TV quality.
Yesterday Google demonstrated experimental search extensions that utilize semantic information contained in Gmail messages about things like Amazon purchases and airline flights, extending their earlier work on semantic search.
Semantic information enables Google to provide better search results and answer specific questions. For example, as shown here, Google "knows" the values of some of George Washington's attributes, including the fact that Mount Vernon was his home.
It also knows some of the values of some of Mount Vernon's attributes:
Google will test the use of semantic information in Gmail messages in a trial limited to 1 million volunteers. (You can sign up here).
If you sign up, Google will use information about your airline flights and Amazon purchases that it finds in Gmail messages. For example, it will use the values of the attributes of airline flights like airline, departure and destination airport, and flight date to find things like your "flight to Cancun last year."
Google's strategy is to improve search a little bit at a time -- baby steps.
Monday, August 06, 2012
There are many Olympic venues, so there may be several live streams, upcoming streams and streams of recently completed events at a given time. This wealth of choice presents a user interface problem -- how do you show the user what is available?
At first, NBC and the BBC offered different solutions. BBC displayed a scrolling menu with thumbnail impages, as shown below.
I liked that a lot better than NBC's menus, which were hard to find and mostly character oriented. I was ready to give kudos to BBC until the tenth day of the Olympics, when NBC changed their user interface to the one shown here:
It looks like a taller version of the BBC interface, but, what you cannot see here is that those small thumbnails are in fact video live-windows. NBC has brought a Microsoft Metro interface to the Web.
You can expect to see a lot more "Metro on the Web" after Microsoft ships windows 8 with its new Metro interface.
NBC has seemed a bit like an agile Internet startup during the olympics. First they improved their video streaming quality and now they have revamped the opening user interface for their online coverage -- another quick, significant improvement. Maybe someone at NBC has read Mark Zuckerberg's letter to investors on The Hacker Way.
I tried watching BBC's live video on my Dell laptop with 4 GB of RAM an Intel Core 2 CPU with a 3.06 Ghz clock speed and an Nvidia Quadro FX 770M display chip using a proxy server.
I ran two tests. For the first, the laptop was running several other applications and used over 3 GB of memory as shown here.
Note that I accessed the BBC stream through an English proxy server, which may have affected performance. (I used it only to write this review).
Sunday, August 05, 2012
Being a cord cutter and sports fan, I tested NBC's online coverage of the Olympics during the first few days and found the video quality to be unacceptable on my Dell laptop with 4 GB of RAM an Intel Core 2 CPU with a 3.06 Ghz clock speed and an Nvidia Quadro FX 770M display chip, but OK on an iPad or iPhone.
To get smooth video during the first day or two, I had to drop the horizontal resolution down to 240p, which, as you see below, is totally blurry on my 1,920 by 1,200 laptop screen. (Click the images for actual size and clarity).
It is now day nine of the Olympics, and I am happy to say that I am back on my couch watching NBC's streaming coverage on my laptop. They have improved the video significantly. It now runs smoothly at 480p resolution, which looks good, as you see here:
If I up the horizontal resolution to 720p, the video is noticeably sharper, but becomes a little jerky, but watchable:
I got greedy and tested 1080p, but that was too slow and jumpy to watch.
I do not know what NBC changed -- more bandwidth, more servers, different compression algorithms or all of the above -- but the performance went from bad to good.
It is noteworthy that Google is NBC's streaming partner. Maybe NBC has learned something of the "Internet way" from them -- deploy early and improve fast. The folks at NBC might also have talked with their colleagues who streamed the Tour de France in July -- that video was great from start to finish.
Lest you conclude that I'm no longer a curmudgeon and have become a soft-hearted fanboy, let me be clear -- the commercials still suck.
Steve Crocker, Internet pioneer and current chairman of the Internet Corporation for Assigned Names and Numbers, has summed up the roles of government and industry in an excellent article. He explains why the Internet could not have been created by private industry without government’s help and also points out that once the initial infrastructure was in place and its feasibility demonstrated, it was vital for industry to step in and develop products, software and services.
That initial infrastructure began with research and theory leading to the ARPANet, which demonstrated the feasibility of a large scale network. Critical mass and training were achieved when the government funded the spread of networking to universities through CSNET for computer science departments, the National Science Foundation NSFNet for connecting all US colleges and universities and the NSF International Connections program, which brought in university and research networks in 28 nations, most of which were in the developing world.
In a 1996 Communications of the ACM article Seeding Networks: the Federal Role, I documented the cost of that government effort. The ARPAnet cost $25 million, CSNET $5 million, the NSFNET backbone $57.9 million, connecting universities to the backbone $30 million, and the NSF International connections program $6.6 million. The US taxpayer got a pretty good return on an investment of $124.5 million.
Of course it is not only the US Government. Donald Davies was at the National Physical Lab in England when he coined the term "packet switching" and built an early test network and Tim Berners Lee was working at CERN, which was funded by 20 European governments and the European Union, when he invented the Web.
The Internet is one example where Federal funding of research has paid off handsomely. For other examples, see this post on Presidential science initiatives.