Broadband Video is overrated too !

No i dont hate technology. I just prefer a little realism with my technology planning. Broadband Video is smoking hot right now. Why ? Because thats where the money is. Advertisers have gone bonkers over it. For good reason. Its a great use of the net for advertisers. A video directed to the right demo can generate far more value for an advertiser than many other advertising opportunities. Advertisers have caught on and are putting their budgets to rich media. Thats a good thing.

Its also why content providers, particularly major media players are putting content into broadband offerings. If you can get a great CPM and sell “name brand” content that commands a great CPM, you are stupid not to do it.

But offering broadband content in order to gain a bigger share of the advertising pie is a completely different realization of broadband than the expectation that broadband video delivered over the internet is going to be a viable alternative or as some future gonzos are suggesting, a future replacement for tradtional delivery of TV. A replacement for TV ? Over the open internet ? Now that is crazy ! (for definition’s sake here, Im exluding private end to end controlled networks like cable or telcos who are using IP delivery of content. Switched delivery of content on an end to end owned network is not “internet delivery”)

The first problem for broadband is bandwidth. DVd quality at 1mbs is fine today, but have you noticed you cant do live consistently for any size audience at even this minimum bandwidth level ?

If someday the internet can support live delivery of 1mbs unicast streams, the cost will be prohibitive (every live stream requires a direct stream from source to end viewer), It adds up very fast. 1 stream per person. X number of streams per server. All the routing and internal backbone equipment to get it on the net. All the monitoring equipment to make sure it gets to the viewer in some semblance of decent quality. All the people to make sure that all works. Thats big bandwidth and overhead and hosting costs. Which is why 350k simultaneous streams at 300k quality for march madness and a concert were considered huge events in 2006. Every single incremental user for a 300k stream of a 2 hour event can cost more than $1 PER USER. (Dont think so ? call a broadband video provider and ask them how much they will charge to stream a live 300k stream to 350k simultaneous viewers with TV level quality of service).

Compare that with the cost of delivering TV today.

Then of course there is the consideration that if broadband will replace TV, what happens when we go High Def ? Lets see we can get by with the lowest quality and only 6mbs of bandwidth (and all you “we have a better codec people, your stuff still looks like crap doing HDTV at 6mbs) . If its a challenge and costs a fortune to delivery 300k streams at 350k , DVD quality , how long do you think it will be before we can do the same over the internet with 6mbs or the required 8mbs for low end and 12 mbs for high end content ?

It aint gonna happen anytime soon. Not this year. Not next. Not 5 years. Not 10 years.

Want to deliver the SuperBowl or American Idol in HD in realtime to 10s of millions of simultaneous viewers ? Not in this lifetime without some breakthrough technology that hasnt been invented yet. (DO NOT SEND ME EMAILS SAYING YOU HAVE THIS. YOU DONT)

So if you want to look at broadband as the future of tv delivery, you can completely eliminate live programming. Of course thats not a big deal. Right ??????

Its feasible of course to do on demand, in a netflix type model. Let it download over night (equal to postal overnight). But it does tie up your PC, so shipping on a hard or optical drive is much more efficient, and thats not the internet.

But wait there’s more. You still have to pay for that bandwidth somewhere. Yes peer to peer helps save bandwidth at the originating end. But it doesnt help at the destination end. 100 peers on a network segment will still use the same amount of bandwidth on that segment as 1 destination with no peers. 10gbs of programming still has to find its way to the destination. So clogged pipes in that last mile are going to clog further as more content is delivered is delivered at higher bit rates. Which in turn mean that fewer broadband bits can be delivered at busy times to last mile users. Net Neutrality will pretty much guarantee that this is a problem forever and ever.

Finally, there is the home experience. Everyone who thinks that analog and sd channels look Worse on an HDTV than they did on an analog tv raise their hand. Now everyone who thinks its worth the hassle to hook a PC up as a media center, connect it using HDMI or DVI to their new HDTV and then connect to the net as their primary source of content, raise the other hand….

You can hook any HDTV today up to pretty much any recent PC and use it as a conduit to get internet content to your HD set. You know how many people are doing it ? Not many at all. I dont know of any families gathering around their brand new HDTV to watch internet content they just grabbed from Youtube. I know millions that are doing it for HDNet, HDNet Movies, TNT HD, HBO HD , ESPN HD.

The reward for connecting a PC to an HDTV isnt worth the hassle and that wont change for years. Dare i say, not even in 10 years, if ever.

Two years ago i wrote that it will be easier to deliver content on a hard drive than it will be over the net. That is the case today, and it will be the case for a long time to come.

Remember, your broadband throughput hasnt increased 5 mbs in 5 years. What makes you think that its going to give you 5mbs more of THROUGHPUT to enable you to do a SINGLE hd stream at a time to your home in the next 5 years ? Or 15mbs more to do 2 HD streams ? (you do want to use your PVR for HD shows dont you ?)

And finally, for all you content creators that think broadband video, whether to a PC or PDA will be the greatest opportunity to make money since…. Podcasting, well read my posts on podcasting. You aint gonna make crap from your broadband video efforts that are drawing zillions of downloads (dont you realize a download isnt the same as actually watching ? do you watch every minute of everything you download ???) unless someone hires you to create programming for TRADITIONAL VIDEO DISTRIBUTION METHODS LIKE TV or DVD !

49 thoughts on “Broadband Video is overrated too !

  1. oh dear, you poor US consumers really have it bad , the UKs Virgin Media cable covering a mear 52% of the UK got several bumps in IP rates from 2Mbit, then 4Mbit,10Mbit to todays 20Mbit, and soon 50Mbit Docsis3.

    for the readers that dont realise it,sure the internet has had TV broadcasting capability from nearly day one, that being IP MULTICAST AND IP BROADCAST capabilitys…

    the simple fact is it the ISPs that are holding you back as they seem to think its clever to turn off the multicast and broadcast capabilitys in all the comercial grade routers and related kit to and from the end users so called last mile conection.

    if you really want to see some movement in all this , then write to your very own ISP boardroom CEOs and demand they turn initiate an across the board re-enableing of IP MULTICAST on all available IPv4 and IPv6 installed kit.

    pay or become active in the azureus torrent developers circles and get some java multicast patches into the current open codebase, to start you off in your search look to the java multicast DHT codebase from the bamboo DHT website

    \”Marcel has also written a report about his experiences building a multicast protocol on top of Bamboo. It may also be useful for tutorial purposes\”

    hell if you cant even manage to get your ISPs to turn back on multicasting all the way to your user end ,then you can always do a poor mans multicast through a basic multicast tunnel, and get what theres also a very old Mbone (the old multicast fre network)free java app Mtunnel
    \”The mTunnel is an application that tunnels multicast packets over an unicast UDP channel. Several multicast streams can be sent over the same tunnel while the tunnel will still only use one port. This is useful if tunneling through a firewall.
    The applications primary goal is to allow for easy tunneling of multicast over for instance a modem and/or an ISDN connection.

    The mTunnel has a built in Web-server allowing for easy access to information about current tunnels. This server listens by default on port 9000 on the machine where started.

    The mTunnel also listens on session announcements for easier tunneling of known sessions.

    If you download and install this package please send me an email! 🙂 (

    The latest public version is 0.3 released 980102. \”

    so the codebase is there, go do something interesting and get your ISP to stop limiting this multicast DHT torrent/p2p
    becoming a reality for the good of your US selves, and the world at large.

    BTW incase you didnt realise, AVC/H264 is becomeing the new EU/world standard for DVB broadcasting/multicasting transport streaming not your old US mpeg2 so theres a massive bandwidth saving there to be had today…..

    Comment by popper -

  2. Actually the internet it hitting a wall

    Comment by Pallet Rack -

  3. Someday the internet will have so much capacity that it really can carry everything. The total potential demand isn’t infinite, any more than the demand for electricity is infinite.

    Comment by Sun -

  4. Mark:

    I agree that the Internet will not replace the broadcasting of live events such as the Superbowl but the deliver of HD content overtime. I.E. The Netflix business model will and does work.

    When Netflix came out people were saying they would not make it because people had to wait on the delivery of their DVDs and Blockbuster was just down the road. Well, Netflix model works and it is kicking Blockbuster’s *ss.

    Now is the time to take the model to the next level and allow all content to be delivered to the home via broadband over time of course and it must be cost effective. Netflix cost per dvd on shipping alone is $0.72 not to mention overhead.

    Netcast HD has developed the ability to deliver HD content to consumers over broadband for about $0.60 per content piece and this includes overhead. Netcast HD can also deliver faster than Netflix and in HD. Our HD content is at 5Mbit for 1080p and I would encourage you and anyone else to check it out at.

    I see the internet becoming the large DVR that allows you the consumer to get what they want when they want it.

    If you want a baseball game, Superbowl or Mavs game, tune into your local broadcast. If you want old TV shows, documentaries or movies, tune into the internet.

    Thanks for your time.

    Todd Bryant

    Comment by Todd Bryant -

  5. Hi All!

    First of all I must tell that english is not my native language… but I hope everyone can understand anyway!? 😉
    I have read some of the comments above and I do not agree to that it will take 10 years or not in our lifetime before a technology solve the issues for a reasonable price etc. I know that big telcom companies like Cisco, Juniper, Alcatel etc. has really big problems with this stuff as soon as the network grows a little and it’s not so strange; The technology they are using is not initially created for that kind of use. First of all this kind of network needs to guarantee 100% QoS (zero packets loss), so we can forget all kind of packet technologies. This leaves us circuit based technologies like Standard SDH/Sonet which Lacks native interface, Low utilisation of fibre capacity and its expensive. To overcome these problems there is a technology that combines the best from packet based with best from circuit switched based technologies. The technology is circuit based and was from the beginning invented to deliver video in networks. It’s the only technology that I know of that promise that they can deliver 100% QoS End to End, just like a telephone network! 😀

    Here is a short description of the technology:

    Today there is only one company that offers this technology and it
    has an impressive list of reference customers, including
    European Broadcasting Union, Westdeutscher Rundfunkt in Germany,
    Broadwing in the US and Versatel.

    The company is called Net Insight and this comment comes from their webpage:
    “Net Insight develops video, voice and data networking equipment that delivers guaranteed 100% QoS with maximum network utilization and multicast to provide a network that can efficiently and economically deliver advanced video services such as HDTV, Video on Demand and Digital Television combined with Internet and voice traffic. Our Nimbra platform enables our customers to launch new revenue generating TV and video related services for Broadcast TV, CATV, Telco Triple Play and DVB-T with significantly reduced CAPEX and OPEX. Net Insight is quoted on the Stockholm Stock Exchange’s O list and has offices in Sweden and the USA. For more information, visit

    I have followed this company for some years now and they actually solve stuff others networking companies can’t and this time it’s all TRUE!

    /// Per

    Comment by Per -

  6. Follow up to above: The reality is that it will still be very difficult, read expensive, for an independent “internet” content provider to be carried, since it would require either streaming, or paying the content distributor to carry it. With or without net neutrality the content provider is going to have to be affiliated via some business transaction with the content distributor (cable company or sat provider) as the type of infrastructure is going to require a closed system unless everyone has fiber. I suspect that the TimeWarner’s and Comcast’s will make it just proprietary enough that they don’t have to worry about competing with the internet, since their quality is so much better. Also, one of the reasons that bandwidth speed increases haven’t kept up is, what would all that extra bandwidth do. VOIP is impacting the RBOC’s, what happens to Comcast if you get 100mb internet and IPTV allows you to turn off everything but internet access. There’s no real benefit and significant costs; both in building out the infrastructure and potentially supplying competitors with an access point to their captured customer, to supply more than 10mb or so.

    I would agree with Mark when you talk about an independent filmmaker getting distribution via the internet, too painful to all but the technology savvy, it’s complicated even for the geeky among us. If you want to make money, you need to have a system that even your mom can use, and my mom can barely use a DVR box.

    Comment by Rick -

  7. One of the things that this discussion misses is that millions of people have PC’s connected to their TV’s, we just call them DVRs. Fundamentally they are purpose built PC’s and as we’ve seen with Tivo, they have a high capability to add functions. While I agree that true VOD, live streaming, is probably not possible without something like FIOS, which is far from implementation in most places, there are other options. The Moviebeam model is pretty compelling, if we didn’t have the limitations of over the air broadcast only providing limited availablity, and the catalog is fairly poor at this time.

    If you could add the ability to preload via the wire, or satellite, a selection of features that are timely, then many more folks would use this functionality. The technology is already in place, how you implement has many options, p2p on your cable box possibly. You could do day and date with theater release, charge $14.99 or $19.99 for some timeframe of access, and you’d have a second distribution channel for initial release and still have the ability to sell DVD’s later. If you go to the theater you’re going to spend at least that much, and going to the theater is a hassle for some folks, hence DVD rentals have had a huge impact on theater seat sales. If the cable companies use something like bittorrent to load your DVR and share the load inside their cable infrastructure, and your DVR could decode WMV-HD ,H.264, Mpeg4, DIVX-HD, you could do HD, without a huge physical server farm to dedicate, its just software at that point. It may take days to preload, DIRECTV and Dish would actually have an advantage in this case, but most content providers know when their release is coming so you can schedule it, or have the users select what they want a la Netflix or CinemaNow et al. I’ve seen your post about the networks and the iTunes model and this would be similar, just a different distribution method.

    Customers should be willing to shell out a few more dollars to get day and date with the theatrical release vs. the delay and DVD release. But once you get the DVR in house, you can order both the premium (i.e. new release) movie as well as the budget (i.e. DVD release timeframe) movies at the same time, you can even have premium extras like a DVD has with the DIVX format. And both the studios and the cable companies can make money.

    The only problems I see with this is that most major studios aren’t necessarily mastering HD content at the same time as film, or even DVD, at this point. So you get back to the problem of distributing an SD and HD version of your content, if you get HD you can distribute it and convert to SD, just not vice versa. Maybe once more filmakers start leveraging the HD cameras or distributing to DLP theaters then content stops being the issue and distribution is your only hurdle. The other problem is that “Hollywood” is in a conundrum today with the whole HD DVD vs BluRay format war, I’m not sure they want something else confusing the customer, but they should get proactive on this instead of reactive. Once someone cracks the DRM, and I’m sure there’s going to be a crack, there hasn’t been a encryption system that is both consumer friendly and uncrackable to this point, their content is going to get pushed out via the internet anyway, and they’ll not be able to hire enough lawyers to stop it. Better to get in front of it and take the revenue from all the folks that don’t want to go to the theater, and also don’t want to mess about with setting up a PC for pirated content.

    There has been a similar discussion every time a new media distribution model comes out. When VHS rental came out there was still some benefit to going to the theater; video quality and sound, but still the studios didn’t embrace it for years. Then DVD’s came out, the rental model was entrenched by then, but since the quality of the media was much better, studios have had a tough time, many movies don’t show profitability until rentals are calculated. Now with 1080p HD content and sound, your quality of the media is fabulous, and you don’t have to worry about the guy on the cell phone or getting up to get more popcorn and missing 15 minutes. Yes, your wife is going to complain about that monstrous TV, but she still likes it better than fighting the army of 13 year olds to get into the theater on Friday night. And that’s the point really, going out to a movie used to be an event, now it’s a chore. The fact that we work more, and are tethered to work via 72 different devices, it’s just more relaxing to stay in. The media companies have been scared to death of the internet for 10 years, but look at music, eventually they started to embrace things like iTunes or Rhapsody, because their customers were already using it and the companies weren’t seeing any revenue, in fact declining revenue. See Napster. Fundamentally there really isn’t a difference between music or video, just cost of developing content and the contents size.

    This is what I want personally, I moved my whole house to HD 5 years ago, when the available content was just not there, plus just built a dedicated home theater room specifically for all this stuff. I’d rather wait 3 months and watch the movie on one of the DVD formats than go to the theater, and I live within 3 miles of 34 screens, all of them new. It’s Vegas, everything’s new. I’d gladly pay $20 for a day and date release without leaving the house, but maybe it’s just me. I just don’t think that there is a technical issue; psychological, maybe; inertia, probably; but with all the smart people around the technical side is probably solvable. Of course, if it were my money, I’m not sure that I wouldn’t want to be second to market instead of the first.

    Comment by Rick -

  8. Mark,

    I have to disagree. You are correct in saying quality broadband TV is not possible RIGHT NOW, but I file your ten year prediction in the same category as some other famous quotes.

    “640k ought to be enough for anybody.” — Bill Gates, 1981

    “I think there is a world market for maybe five computers.” — Thomas Watson, 1943

    “The reward for connecting a PC to an HDTV isnt worth the hassle and that wont change for years. Dare i say, not even in 10 years, if ever.” — Mark Cuban, 2006

    Let me break your argument down into two sections. The first argument is that the backbone can’t handle the load. To this I point to an overabundance of dark fiber, the gradual rollout of 10 Gbps connections (Sprint is in the process of upgrading their entire backbone from OC-48’s to OC-192’s), and the improvement of multicast technologies. I think within ten years the Internet will easily be able to handle 10 to 30 Mbps into the home. Dare I say within five years?

    The second argument is that the end user won’t have the bandwidth. I point to two rollouts:

    The first is the 30 Mbps service Verizon is offering. The second is the 2.5 Gbps service France Telecom is offering. Yes that is 2.5 Gigabits per second. These services are rare right now, but give them two or three more years and we should have a large enough potential audience to justify the creation of high-quality Internet video stations. (Should we start calling these Intervision stations?)

    In response to your comment that its taken five years to go from ISDN to 5 Mbps, I would like to point out a few things. Bandwidth increases exponentially… (10 Mbps, 100 Mbps, 1000 Mbps, 10000 Mbps). That today’s infrastructure (primarily Cable modems) has the capability to deliver higher speeds but there just isn’t a demand for it. The reason last mile speeds have stagnated is a lack of applications for high speed bandwidth. I see that changing. Online gaming is probably the biggest driver, but online video is quickly becoming another driver. Look at the adoption rate for ESPN 360 and Xbox Live if your still in doubt.

    My prediction is that Intervision will be common within five years.

    I normally don’t post comments on your blog; but then again normally you don’t make it so easy to prove a billionaire wrong.

    Marquis Daniels

    Comment by Ladar Levison -

  9. As several other commenters have noted, hybrid satellite/cable/over-the-air and broadband IP hybrids are probably the best solution for the short term. And, why not? The extremely high def live video that needs to be displayed on a 56″ widescreen does not beg to go portable – who cares about high-def on a telephone-sized screen or on a 13″ laptop screen, for that matter? So, IP broadband for low and medium def content, digital over-the-air/satellite/cable solutions for live and high-def.

    Mark’s comment about media-center pcs… I think that these systems are where component Hi-Fi was in the early 70s. The adopter profile is similar and the mainstream adoption rate will probably repaeat that pattern closely. I know that several of my friends (non-geeks, by the way) have bought media center pcs and are using them, either with HD or without. I have been successfully using a Mac Mini since February 2005 as a media center (yes, well before the Intel version) and have been extremely satisfied with it. One little box does TV/DVD/Music playback and acts as a DVR (with EyeTV). Easy to set up.

    SO – hybrid delivery, small, stylish, easy-to-set-up media center pcs sum up to solve the problems that Mark lays out. Eventually we’ll get to IP delivery of all content, but – as someone suggested – we’ll be using it to teleport to the live event, not watch it on high-def :).

    Comment by Steve -

  10. Cuban is wrong with this: “If its a challenge and costs a fortune to delivery 300k streams at 350k , DVD quality , how long do you think it will be before we can do the same over the internet with 6mbs or the required 8mbs for low end and 12 mbs for high end content? It aint gonna happen anytime soon. Not this year. Not next. Not 5 years. Not 10 years.” Gigabit internet and 10Gigabit internet is closer than he thinks…several Asian countries are known to adopt technology much faster than the U.S. and Hong Kong already delivers 1Gps (symmetric – that’s correct upstream and downstream are the same speed) to its residential users (about 800,000 users are currently wired and more will be added). My personal experience, while on a trip to Hong Kong, was good – watching MPEG-2 videos at speeds ranging from 4.5Mbps to 10Mbps were of DVD visual quality.

    Comment by redline -

  11. South Africa: 30 Mbps Broadband over Power Lines Deployed- IPTV next !!!

    Powerline Deployed in South Africa And Uganda – IP-TV Trial This Autumn
    Balancing Act (London)
    July 17, 2006
    Posted to the web July 17, 2006

    By Russell Southwood

    Powerline technology has always seemed to promise much but never seem to quite come to the boil. But Goal Technology Solutions (known as GTS) has rolled out operational 30 meg connections in South Africa and is currently deploying in Uganda for UTL. And come October this year it will be trialling IP-TV. Russell Southwood spoke to its CEO Adrian Maguire about why it had succeeded where others have failed.

    GTS is a spin-off of the Power Line Communications division of Grintek Telecom. The GTS team worked for Grintek in this division for two years before setting up GTS in September 2004. Adrian Maguire, CEO, GTS is very honest about the early years:”The first two years we had relatively little success. We went through a number of suppliers who worked not quite well enough for commercial deployment and there were issues of cost.” But it is now the only integrator and value added reseller appointed by Mitsubishi Electric for the SADC region for its PLC technology. Maguire told us:”It’s DS2 technology but we’ve done quite a bit of local development. These are small tweaks to get reliability. And with that, it’s now gone over the curve of let’s see if it works.”

    It went live with its first application in November 2005 with 130 houses in Pretoria. According to Maguire:”We were trialling second generation powerline technology and it was our first large-scale trial. The customer (Tshwane Municipality) wanted 4-6 meg per home but we were able to deliver a 30 meg connection per house.” The company is focused on providing “last-mile” solutions.

    It has worked hard on the applications that can be delivered using the technology so that it does what it’s supposed to do. It has run voice and Internet, installed high-quality security cameras and deployed water meter reading devices. Maguire says:”The reliability is such that we don’t have to keep going back.” It will conduct its first IP-TV broadcast in October this year.


    Comment by trexia -

  12. Its because broadband was the “in” right now, take a little more time again and you’ll know there will be something new arises. Broadband was just for now. Take for an example on the birth of computer, it had its glorious time but eventually more development came and new issues had been focused into. I have heard great ideas from friends on webdate*dot*com that comes up with new ideas on a brand new better technology.

    Comment by Edz -

  13. So if you want to know where the market is going, look at porn. People aren’t bothering with puny MP3s anymore. Now, they’re downloading ENTIRE ripped-off DVDs over high download bandwidth into humongous disk drives (400GB for $200). Then playing it on their PC or loading onto their 60Gig iPod. Advertising is in the dirt–nobody reads it, and streaming is in the mud–nobody wants to bother with it. But downloading is still in the game. The traffic is so heavy that the Megauploads of the world have to throttle it back to keep from swamping their boat. And yes, you can get a 300GB disk with 1000 hours of your (or somebody elses) favorite porno on it. Next step: download to an integrated TV/disk or Movie-to-disk to TV over USB. Piece of cake.

    Comment by ronniebeegood -

  14. How much does pay Level3 to download a 4 minute Video over the Net to the Consumer???? THANKS for your INPUT and also the $1 per streaming real life FINANCIAL input on how much it would COST to stream traffic real time like March Madness! WOW, I had NO idea they were getting a Dollar per stream!!! THANKS !
    PS when does the Basketball Team give away FREE VOIP with a Season Ticket???????

    Comment by skibare -

  15. To summarize your point – broadband doesnt have the bandwidth to support TV-quality video. I agree with that, but — “so what?”

    The demographics point to people spending more time looking at their computer screens than TVs. Poor-quality of My-choice trumps High-quality of Your-choice.

    Comment by yesdi -

  16. That’s comparing apples to oranges… or rather, comparing TV to anything. There are people who will defend it to the death that video on their mobile phone looks GREAT.

    All things considered, I’d rather view a youtube video on my 19″ monitor.

    HD content is best viewable on my HD TV. So to each, its own.

    Comment by Vidizer -

  17. Convergence between devices picks up speed as technology gaps between TV and PC diminishes, and The Internet has become a global infrastructure for communication, and has proven itself as a viable content delivery infrastructure (just look at Itunes!).

    The problem is live streaming of high bitrate content, because then you need infrastructure online which can support one-to-many webcasts (rather than traditional internet point-to-point broadcasting – unicast). The problem with multicasting is that it is hardware-based and all users must be on the same network – but by using p2p streaming networks instead you get in effect a software based system for multicasting!

    The cost of webcasting increases proportionally with the audience size. This is why P2P streaming has become an important issue on broadcasters’ agenda. Even big national broadcasters such as the BBC are currently testing different solutions for using peering technologies in their streaming. When using this technology, the bandwidth requirement of the broadcast is intelligently distributed over the entire network of participants, instead of being centralized at the broadcasts origin.

    This makes broadcasting over the Internet completely scalable and eliminates the success penalty broadcasters normally experience. RawFlow Inc. is one of the leading suppliers of P2P software solutions today (
    Peer enabled streaming helps broadcasters get around bottlenecks and minimizes bandwidth costs, but at the same time it is relying heavily on the upload capacity in the network, and therefore there is a limit on the quality of the stream. Thus, it is not yet possible to stream full-quality movies live. However, we can expect that the high-speed internet penetration in the population will increase dramatically within a short timeframe.

    Comment by Ingjerd -

  18. All i’m feeling here is a bunch of nega-frickin-tivity! I hope you sense the sarcasm in that comment.

    I disagree with your time frame. If you wanna see high efficient bandwidth utilization, download BitTorrent and try to download an episode of a popular TV show. You can probably get it in HD, and have it in less time than it takes to watch it (meaning it could be streamed to you). A 30 minute episode of Family Guy takes around 20 minutes to download (assuming you’re on a decent tracker). People get too bogged down in the infrastructure – Peer2Peer is where it’s at, BitTorrent is incredibly efficent (like 98% pipe utilitzation – which is borderline unreal). I can download a whole DVD of content in a couple of hours.

    Btw, my bandwidth has increased by 5Mb in 5 years. In ’01 i was on dialup.

    At the end of the day Mark, technology always finds a way. Mb/s is the new Mhz. Saddle up.

    Comment by Josh -

  19. Broadband over Electric Power Lines is one of the answers for Bandwidth problems… BPL speed is usually Symmetrical download/upload using 200Mbps technology !!!!!

    Comment by TREXIA -

  20. Hmmmmmm . . . . didn’t YOU make a billion of an overrated Internet video/audio company?

    Comment by Jason -

  21. The future will be on “friend networks” where I can watch stuff on my DVD player and pump it back into the internet so that my friends can watch the fun. I’ve already been buying up content in DVD sets. I’ve got more programming than TNT.

    What will also be making a major impact is those toaster size 2TB hard drives – you can load up hundreds of hours of old TV shows. Who needs Nick At Nite? You’re going to be able to create your own stations – set up auto programming cycles – minus those 10 minutes of ads per 30 minutes.

    The future belongs to narrowcasters.

    Comment by Joe Corey -

  22. There has to be a bit more credit given to connectivity than what you’re trying to say. As someone who works closely in the hosting industry, I’ve seen a lot of new opportunities to bring very good bandwidth to homes, especially those in urban areas. And because of this, along with the fact that Microsoft and other big names are developing computers that are effectively home entertainment systems to be used in the family TV room, I do think there’ll be a growing demand for delivery of content to the home. YouTube is an example of user-driven content but we’ll eventually see sites that mimic the YouTube ease-of-use with professional content. In fact, Microsoft has been creating sitcoms and various other shows to push this new type of media center. Your rule of hardware falling in price, will also eventually be seen in tier-1 bandwidth getting to our homes.

    Comment by Rohit -

  23. It’s the same reason why ISP’s will not be able to deliver an acceptable video experience over anything less than a FIOS like fiber service. I think IP video will work for stuff like Rev3’s content, systm diggnation and the likes (which are not HD) – but true HD is just such a bandwidth hog it’s not gonna happen.

    Comment by Andrew -

  24. Mark:

    Having been the operations executive for one of the largest data networks in the world, I agree with you that broadband internet is far from being the cheapest distribution system for mass media. That distinction goes to satellite technology. As I said in another post a while back, with the penetration of DVRs, the current satellite TV systems can be viewed as having N unique streams, each available 24/7. So if you have a 100 channel system, there is capacity for 2400 unique one hour time slots each day. I think we’re still pretty hard pressed to find that much programming (e.g. look how often you repeat shows on HDNET!)

    While advertising will be a component of revenue for a long time, its form will change: more imbedded placement, more commercials as entertainment, etc. But I believe that what we think of as network TV today will go to a subscription model, where individuals subscribe to particular shows (or more likely, bundles of shows), and their DVR/computer records it whenever it happens to be broadcast. Past episodes might well be available for on-demand download, but those can potentially be delivered on a low-bandwidth internet channel (which goes to your point about multiple bandwidth/price points for internet service). Or the consumer can pay a premium and get it delivered on demand at full bandwidth.

    Someday the internet will have so much capacity that it really can carry everything. The total potential demand isn’t infinite, any more than the demand for electricity is infinite. You don’t hear folks in their home saying that they wish their electricity wasn’t so restricted. Pretty much everyone (in the US)can get all they want, and the system is scaled to deliver it (with occasional adjustments, eg when a/c was introduced, or when computers got to be everywhere).

    Same thing with telecommunications. Our eyes have use for only some many pixels/sec and our ears can only hear so much. So it is possible to build enough capacity to meet every reasonable demand, but it’s still too expensive relative to reasonable alternatives (e.g. satellites).

    Someday we’ll figure out the whole Star Trek teleporter stuff, and will want another couple orders of magnitude of internet capacity to handle all the bits associated with that. Before then, we’ll have public teleporters, and prices will be based on distance and volume (us big guys will cost more I suppose). I wonder if they’ll have MPEG328454 as the standard compression scheme for teleporting? You’ll come out kinda like you went in…

    Comment by Paul Lambert -

  25. Also broadband video needs huge broadband. Servers need upgrades or even replacements, networks too. It means money for service. Of course it’s profitable

    Comment by Dexter -

  26. Good call on the bandwidth barrier.

    However, I think the internet scenarios like MySpace (and all the possible things that connect to the internet like iPod and Xbox) are the dominating factor… more so than the specific application of video. Time is a zero sum game for the most part. (unless we’re talking about online gaming in Korea)

    Now I don’t know much about the cable business and I’m relatively new to this blog, but is CPM a big factor in the HD cable business model? Agreed that video over broadband is uncompetitive today, but it’s not a far stretch to imagine highly targeted CPM video inventory being integrated in to non-video (and even non-browser) applications. And today’s pipe is fat enough for some 30 second ads.

    Comment by ventureless -

  27. Have you ever looked into Flash Media Server? There’s a way that you could stream unlimited amounts of data to a ton of people. Of course, this would be extremely expensive. Here are the costs:

    For an unlimited amount of bandwidth, you can stream content to up to 150 people for $4500/license. So lets say FOX was going to use FMS to stream American Idol to 20,000,000 it would end up costing them only $600,000,000

    That’s $30/viewer. Surely advertisers could cover these costs!

    Hahaha. Someone needs to invent a whole new way of data transferring and streaming.

    What if you could mesh certain geographical areas together and have everyone share the data that’s being streamed to them to the nearest, say, 5 connected computers? So every computer acts as a server, sending patches of the stream to other computers. The computers that receive it build up all these packets of data into one viewable stream. That’s a mess for the servers that are transferring the data to begin with, and doesn’t even seem possible.

    Wow, that’s a terrible solution to.

    This whole video over the internet concept is so far behind it’s not even cool…


    Comment by Sam Purtill -

  28. Mark You I agree with you on live streams its not ready for Prime time even though the people at NFT will try and convinve you differntly ,But companies like Red Swoosh claim they can deliver recorded content more effiently as long as they have scale

    Im sure you have read Travis’s Red Swwosh blog on what he thinks about Live content delivery .

    “How much of the TV you watch is truly “live”? Don’t look at me funny, it’s a serious question. What fraction of the TV you watch is only a few seconds old, straight from some camera to your TV without editing?

    Let’s break it down:
    – Is anything on Must-See-TV live? No
    – How about Saturday morning cartoons? No
    – Leno?, Letterman? Nope
    – The Sopranos, Everybody Loves Raymond, Soaps, Opera Winfrey? Nada

    How about the national news? If you’re on the east coast, maybe it’s live, not on the west coast though. If you’re watching a 24 hour news channel, the *vast* majority of play time is pre-recorded segments. How about 24? That clock doesn’t make it so 🙂

    Now isn’t the holy trinity of media distribution – broadcast, cable, satellite – perfectly designed and suited for live distribution of Television? Of course it is.

    So why isn’t most of TV live? Here are a couple reasons:

    – Editing makes TV better.
    – For 95%+ of the TV people watch, live doesn’t matter, NOBODY CARES

    Even the notion of linear programming seems to be dwindling in every direction I look. Comcast and their MSO brethren have launched and are aggressively expanding their on-demand offerings. 99% of what is distributed on the Net is non-live content (the 1% is made up of the few radio stations that still do live streams online – the most popular radio services are simply playing pre-recorded songs in a pre-set playlist).

    So, why the heck is everybody knocking down my door about live P2P streaming? Usually they’re frenetic, intense and on a mission, like Indiana Jones about to get the holy graille or something. Cheap, live, high-performance video distribution online. Previously, only in the sci-fi books. Does it make me feel kind of special to be the guy, behind the guy, behind the guy, that can make their dreams come true? Of course it does, and I bask in their praise every chance I get.

    But then reality sets in:
    How much of Headline News is Live? I’m going to guess less than 10%
    How much of ESPN is Live? Less than 50% for sure.

    Now don’t get me wrong, if you’re going to want live, it’s going to be in one of 3 categories: sports, news, events (like concerts). But even in these categories, sports is the only one where a majority of its viewed content is Live. Long-term customers probably range only at 15-20 max. My point isn’t that Live doesn’t exist or that there’s no business there, but it’s not even in the same order of magnitude as the big business of pre-recorded content delivery.

    So how come all the hub-ub, the frothy mouths, and the Venture Capitalists circling around like vultures?

    The REAL reason for this recent move toward live online P2P is licensing.

    Content owners are scared of losing ad revenues in an on-demand world (Tivo on the Net is worse than the Second Coming for these guys) especially one that’s online. Cable companies are scared that the pipes they built to double up for Internet access will be used to get around their $80/month/subscriber cash cow. Content companies are scared that their cable deals will get screwed if they try to “go around” the cable companies by doing on-demand on the Net.

    Bottom line is that for the next couple years, on-demand, ad-supported TV licensing for the Net is a non-starter.

    And so with all those 100 million people on the Net but no way to get them on-demand programming online, some brilliant guy trying to make his numbers came up with a genius holding pattern:

    “Why don’t we take pre-recorded TV content, and broadcast it Live over the Internet?”

    Don’t even get me started on the technical lameness of the proposition, but this is the world we live in for a couple years, until online, on-demand licensing makes the world better for consumers everywhere.

    In the meantime, I’ll bask in the temporal sunny rays of the Live P2P streaming spotlight.

    – chief swoosher “

    Comment by Matt -

  29. Multicast will solve some of these issues for live broadcasts:

    Although it breaks down with millions of viewers.

    For the rest of the programming, delivery of the program before it “airs” solves the limitation of the final mile. Using a protocal like bittorrent also solves the providers bandwidth issue.

    In a perfect world, you’d get a box with a hard drive in it. You’d tell it what shows you want to watch, and it goes out and downloads them in the off hours. All in HD quality. This can be done today, with today’s technology. Heck, an iptv provider can just give people a Mac mini as the box, and use iTunes as the delivery mechanism. Make it easy to subscribe and you’re good to go!

    Comment by Ryan Collins -

  30. I agree that we won’t have a good way to get video content on the current web. It is not because of technical difficulties, but political difficulties. There are pirates that deliver a variety of content. The size is 350MB per 42 minute show. HD content is the same size. It looks better, but I don’t have an HD screen.

    Comment by Allen Cole -

  31. Sorry your underlying premise is wrong. HDTV can’t deliver any more 30-second spots than conventional or broadband TV. The consumer is already consuming 17 minutes for every 60. Does HDTV’s magazine quality dpi change that ratio? If HDTV can’t then the advertiser has no reason to pay more; therfore conventional and broadband TV is good enough.

    Comment by HighAnkleSprain -

  32. Mark you are right on. If only criticism is allowing the Net Neutrality inmates to dictate the terms of the debate.
    You need to get a network engineer to explain to the politicians what the internet is and is not.

    The problem is the Internet is based on 40 year old technology design to carry eight bit ASCII data over multiple network paths between University and DOD computer installations in a way that can survive a nuclear attack.
    The Internet was never designed to handle video traffic. The latency alone for breaking up a video frame and sending them over multiple networks and reassembling them in order at the destination knowing the data pack will not arrive in the same order at the same time and having to request retransmission of missing packets. The best I can get is a 4 inch screen on my 17″ LCD with occasional video hiccups and. I can not see the same technology used to transmit video worthy of a 26″ HD LCD I am ogling at Best Buy.

    Comment by Richard -

  33. Your comments about the increase in bandwidth to the home over the next 10 years is spot on. I face a similar problem on a smaller scale when I’m trying to backup client machines to a central location within a corporate infrastructure. And every time I hear about the problem, I’m reminded of a quote. “Don’t underestimate the bandwidth achieved by driving a bunch of memory sticks along a highway.” The quote used to be ‘a bunch of tapes’, but I’ve updated it for the timed.

    Given that the cost of memory is falling so far, so fast, I’m sure that delivering high quality content using some version of this model will be practical long before critical mass is reached for residential super-high bandwidth. And will continue to be practical for years once it does.

    Comment by Bruce Johnson -

  34. Mark, I think you are being a little short sighted with the potential market of internet tv. You ask about the capability of sending the SuperBowl or American Idol in HD live over the internet, but the thing is that doesn’t need to happen. That problem has already been solved off of the internet by the use of UHF transmissions in all major cities (It’s free too). The point is that the content delivery over the internet is a different market than traditional tv and the technologies can, will, and should coexist together. Eventually it will be significantly cheaper for Netflix to send a movie over the internet compared to shipping via trucks and airplanes. This could actually be done right now for dirt cheap with a bittorrent like technology like a previous poster suggested. Its just going to take the right implementation of technologies that already exist and it will be done much earlier than 10 years from now. Live tv, such as sports, news, and shuttle launches will always be most appropriate for standard broadcast, but a movie or a sitcom or CSI can be delivered and then accessed by the viewer when it is convenient for them.

    Comment by Jay -

  35. Mark:
    You’re spot on about everything, except for maybe bandwidth evolution. I have Verizon Fios for my TV, Intenret, and phone service. I have 15mb/2mb for Internet, overkill at this point. There are 20mb dedicated to the TV(wicked amount of HD channels BTW), and the voice stuff Im not sure about, but who cares, it works fine. I love my connection and I hear I’ll soon be able to wirelessly transmit content from my PC to my PVR. That might make the whole process more palatable. I realize that Verizon customers represent a lucky minority at this point, but I’m wondering if the other telcos/cable companies won’t be forced to follow suit to remain competitve.

    Comment by Mark -

  36. Mark, do you know how broadband gets into my house? Via the same cable that delivers my television content. Let’s think about that…. Internet and TV via the same wire, hmmm. On my television I can get video on demand over that cable. Start, stop, rewind, fast-forward, all over that coax cable. It also delivers digital cable for high-definition. That’s a lot of bits and bytes. Okay, I’ll stop being snarky.

    My point is that the hardware and the technology for great Internet video pretty much exists. While TCP/IP packets may not be an elegant way to transmit data-heavy mediums like video, we know that movies and sound and such can stream quite nicely over coax and fiber-optic. The two signals can even co-exist in the same wire. It’s just a matter of using commands passed through TCP/IP to control the other, and since I can control video on demand I am assuming that technology exists too.

    Instead of hybrid cars we might soon be buying computers with hybrid-cards. Cards that allow our computers to receive different types of transmission technologies and to have those data streams work together seamlessly.

    I don’t think we’re going to get to the Jetsons via TCP/IP alone, but the answer is out there Scully. The answer is out there.

    Comment by Tom Schmitz -

  37. You may be right in saying that the net isn’t capable in its current form of handling demands that will be place on it in the future, but what do you think should be done about it? Do you think the current legislation is a good solution?

    Comment by Stuart -

  38. Business as usual!!!

    Your only reason for batching broadband video/tv is because you can not make much money out of those customers, not yet. People seeking online video/tv are not interested in watching Idol, the super bowl or any of your HD programming. They just one simple, free and good entertainment. I personally see most of my news online and stating to find more programming form networks now online. I can see them when I please!!

    You are right, but broadband TV is not to transmit live to millions Idol, and that is not what we want it for either. Do not lose sight of the market because is just a completely different audience and will just get better!!!

    Comment by Jose -

  39. I agree with what you are saying about
    live events, which basically have to be steamed.
    If data (like new movie releases) can be
    delivered overnight, or in advance (with a key
    to prevent early access), it could work using a
    bittorrent model. The math changes dramatically
    in a peer-to-peer model. I could divide the
    file into 10 chunks, send each chunk to 10
    different clients, and those clients happily
    send chunks back and forth until the all have
    a complete copy of the movie. And they tell
    10 friends, and they tell 10 friends, and so
    on, and so on.

    My download link is about 10 times as fast as
    my upload link. This problem caused a real
    problem for peer-to-peer sharing, but bittorrent
    or a similiar scheme allows me to download
    from 10 similarly upload impaired peers.

    Maybe a NetFlix like company can offer a
    25% discount if you allow your computer
    to serve as a distribution peer.

    And I imagine some smart person can design
    a micro-bittorrent that would work on live
    broadcasts, with maybe a couple minutes
    of delay to ensure all the chunks arrive
    in time.

    see ya,

    Comment by Robbie -

  40. In japan consumers can get DSL speeds around 50mbs. Here in the states I’m still waiting for 6mbs.

    Comment by Stewart -

  41. Mark,

    We bantered some time ago about the merits of IPTV as a viable means of delivery. I am in your court on this, as there is still no SOLID system that is going to allow for live delivery.

    Even in an MCAST world, as a designer speaking, there are still too many obstacles. Let’s say that I could distribute via MCAST and getting to the last mile was no problem. Delivery from the last mile is difficult to say the least.

    Unless you get Fiber to the *x factor* we will still have an inferior service. The best HD at 6meg is a blotchy mess most of the time and any decent HD capable display show’s this. Compression has merits and has a place, but it does not substitute for live viewing.

    As for connecting my PC to my HD monitors for viewing, the novelty wore out for me, oh about 2 seconds into it. It is nice to have multi-head systems and use fat ass Sharp Aquos screens as a virtual desktop, but the fun stops there. It’s like going to the dentist to get your teeth clean, you always walk away feeling good, but at the end of the day, the results just are not what you expected.

    It will be nice to see SBC and Alcatel attempt to deliver cable and satellite substitutes, but at the end of the day, my cost savings (if any) for licensing and box fees and such will not be large enough in a quality/dollar ratio to sway me from satellite or cable.

    Now to all the nay sayers, I am not saying it will not eventually happen, but the time is not today, 1, 3, or even 5 years from now. When the IPTV solutions hit mass market, you will be limited to the number of simultaneous streams. Again we are talking about live (real-time) content, not buffered. I would expect to see an increased number of delayed content delivery services once IPTV is deployed in mass, broadband makes that model very attractive. The price will be right, and perhaps the quality will be just as good as typical HD over the air, cable, or sat. But then we are not talking about broadcasting streams of live content anymore.

    As an architect dealing with both commercial and technical issues, IPTV is still cost prohibitive. Someone can correct this, but I think Verizon is spending somewhere in the neighborhood of $750 to $1,000 per subscriber to deliver their IPTV services. That is an incredible cost, and ROI will not come for years. Remember that only a small portion of your bill actually makes it to the profit margin, making returns more and more difficult.

    I like the Netflix model of content delivery using BB as the medium. Perhaps with DRM and other protections available set-top-box viewing will become more popular.


    Comment by Wayne Downing -

  42. I agree with everything Mark said and the problems.

    Something interesting though – I live in a semi-rural spread out suburb in Ohio, and here we have a local ISP rolling out fiber to the house. They are feeding phone(voip) / internet / cable through the pipe.

    Just like broadband it will take time for the technology to catch up with a mass of the population, but I do think if companies keep investing the speed to the home could raise significantly.

    Of course that only solves a small piece of the problem. Traditional channels are here to stay for quite a while. The real sea change will occur when shows start leaving the “9 o’clock schedule” and going to On-Demand.

    Comment by Chris Crawford -

  43. I would have to say you are dead wrong about the value of hooking a PC up to your TV. It will come a lot sooner than you think (some would argue that it already has with DVRs being just a limited PC). Not only do you get the benefit of net content (limited now but growing) but you get to take advantage of the awesome display of your HD TV for other purposes (like watching home movies, picture slideshows, DVDs, listening to music, etc).

    You’re right, it will take a while for mass adoption, but to dismiss it as you do is rather shortsighted as it doesn’t take into account the other useful functions that happen once you connect the two. Once you try it you will never go back.

    Comment by Jesse Chenard -

  44. Hum, well I think we may be too focused on the IP/Broadband use for all of the delivery….how about using a mix of the two? how about incorporating into IExplorer plug-in features that will allow USB connection to a TV/Cable line and IE using XML and “Service Oriented Technology” would mix the two and allow live streaming to originate at the box and not from the ISP. This way all the decision making and/or Programming of the live stream is delivered using IP/Broadband, but when it comes to the content or actual TV Program, that is delivered using the existing techniques (Cable/Dish). The PC takes both streams from IP and Cable/Dish, and mixes them and allow a TIVO like environment. This will reduce IP unicast issues, and will make use of cable/dish to reduce network contentions. Basically, a boxset technique built into the IEexplorer.

    Comment by Mitchell -

  45. Broadband is definitely smoking hot. We can agree on that much. But let’s not lose perspective of one simple fact: the Bells have been promising fiber to the home for YEARS now, while pocketing our money and doing nothing to improve last mile service. They’re scared that they’re going to install the fiber, and someone else is going to make money delivering content through “their” pipe! Don’t hold any illusions – the Bells are out to make money not only by distributing ones and zeroes, but also by getting you to “prefer” their own content to that of YouTube, etc.

    The only reason connecting a PC to an HDTV is worthless right now is because the quality of video is low. Once quality improves, content will improve, you can count on that. What are we supposed to do with telcos who keep dragging their feet on last mile fiber because they know they can get a hefty tax break or subsidy every year?

    The best part that we’re missing is: if this fiber WERE installed to the home, that would be up AND downstream. Force everyone who is watching your program to host it simultaneously.

    Damn, have to go back to work, so no time to proofread this comment. I look forward to what others have to say on this issue.

    Comment by anton -

  46. Broadband took the embarrassment out of the internet. It made the internet a useful. That extra speed to download pages made it okay to spend and afternoon on the internet “researching.” Otherwise, in the days before broadband and afternoon surfing meant that you were on the net just waiting for pages to load. It was such a waste of time to get anything actually accomplished before broadband.

    Comment by Antonio Howell -

  47. So, you’re saying the internet is a series of pipes?

    I don’t mean to jump on one line of your post, but Net Neutrality is a lot more than just “clogged pipes.” The core issue at hand is whether or not the big telcos will get to determine what parts of the internet I can access freely. As technology evolves “clogging” will become a non-issue but if the main service providers can determine who sees what content based on your economic level then that’s a problem.

    Comment by Josh Byers -

  48. My only additional comment is that multi-cast technology has the potential to overcome some of the problems you raised. Regular TV is of-course multi-cast and so this is an obvious direction for improving the internet based video experience.

    My recent attempt to watch the shuttle launch via ‘NASA TV’ is a case in point of why unicast internet technology will never succeed. 45 mins before the launch, watching NASA TV was a smooth and pleasant experience. With 15 minutes before the launch the problems began. More people logged in. The picutre and sound froze. After a minute of so, enough people got fed up with the poor quality of service and closed their windows media players. This freed up the media server, and the rest of us had the sound and picture restored. But only for a couple of minutes before more people logged in and the cycle repeated. Needless to say the media server was jammed for the few minutes before and after the actual shuttle launch. A useless application of internet technology.

    Perhaps content providers should impose a small cost on viewers? This would restrict demand and result in a better chance of providing a quality service. Of course this approach is incongruous with the culture of internet surfers. Is CNN Pipeline successful?

    Comment by Tim Adler -

  49. Mark your right, The internet is not Open for everyone or Free for development at any point. You must invest, someone with stupid money probably more than you will have to implement a better method of delivery. The internet needs to have its own T.V. channels and so On with programming that can be downloaded…

    – Bandwidth is too expspensive, yet I hear of all this “Black fiber” and “Dark fiber” and “Un used optic” – isn’t that referring too extra pipeline we aren’t using? … I also heard google is buying it up.
    All that to say, Saying Delivering TV on internet connection bandwidth. Is as good as saying FREE WIFI is coming…

    Yeah google tried it with earthlink and others and failed. In Lousiana and in California. Both are miserable failures. The cost of Tech persons to keep it up and the equipment. can’t be supported by ads.

    – Thanks for sharing good solid information mark, – Richard BOwles

    Comment by Richard Bowles -

Comments are closed.