A world of producers

There’s a good chance that the best picture you can put on your HD screen doesn’t come from your cable or satellite TV company, but from your new HD camcorder. As time and markets march on, that chance will only get larger. That’s because the there is a trade-off between the number of channels carried and the quality of each channel. To squeeze in more channels, the carrier squeezes out picture quality through compression. The result is “artifacts.” See here:


The titles get “jaggies,” the football field gets pimples, and everything gets blurred and/or re-painted by he compression algorithm as an approximation of the original image.

Carriers compete more by the number of channels they carry than by the quality of each channel. (There are exceptions to this, but on the whole that’s the marketing think.) Meanwhile your camcorder quality only goes up.

And as camcorder quality goes up, more of us will be producing rather than consuming our video. More importantly, we will be co-producing that video with other people. We will be producers as well as consumers. This is already the case, but the results that appear on YouTube are purposely compressed to a low quality compared to HDTV. In time the demand for better will prevail. When that happens we’ll need upstream as well as downstream capacity.

So here’s a piece in Broadband Reports that shows how carriers can be out of touch with the future, even as they increase the capacities of their offerings. An excerpt:

In upgraded markets, Comcast is not only upgrading existing speed tiers ($42.95 “Performance” 6Mbps/1Mbps and $52.95 “Performance Plus” 8Mbps/2Mbps tiers became 12Mbps/2Mbps and 16Mbps/2Mbps), but is adding two new tiers to the mix ($62.95 “Ultra” 22Mbps/5Mbps and the aforementioned $139.95 “Extreme 50” 50Mbps/10Mbps).

One recurring theme we’ve seen in our forums is that the new speeds have many users downgrading. In both forum threads and polls, many customers on Comcast’s 16Mbps/2Mbps tier say they’re downgrading to their 12Mbps/2Mbps tier — apparently because they don’t think an additional 4Mbps downstream is worth $10. Customers used to be willing to pay the additional $10 for double the upstream speed, but there’s no longer an upstream difference between the tiers.

That last line is the kicker. Comcast apparently still thinks that downstream is all that really matters. It isn’t. For anybody producing a lot of photography or video, upstream not only matters more, but supports activities where the user can see the difference.

In fact there isn’t a lot of perceived difference between 12Mbps and 16Mbps on the downstream side. Either is fast enough for a YouTube video. But on the upstream side, you can see the difference. In my case, that difference appears in the progress bars for pictures I upload to Flickr.

A few months ago I upgraded my Verizon FiOS service from 20/5Mbps to 20/20Mbps. The difference was obvious as soon as it went in. The difference will be a lot more obvious to a lot more people once those people start sharing, mashing up and co-producing higher-definition videos.

Just watch.

7 responses to “A world of producers”

  1. There are some other matters to consider, as well. I live in an area with some of Comcast’s oldest cabling. That puts a real cap on up- and download speeds. I’ve recorded a 630Kbps download speed on a system that’s supposed to deliver 6Mbps. Until the cabling system is replaced, I don’t think it really matters which “Plus” “Performance” or “Extreme plan I pay for.

  2. I think the frustration with upload speeds will be with us for quite some time. I’d put 300 Gigabytes of stuff up on Flickr if I could, but there’s no way I’m going to wait that long.

    If I had more time, I’d have already gotten my advanced sneakernet design up and running. The basic idea is to carry your stuff around, plug in to a family or friends PC, and have it automatically distribute things while you visit. (Straight up copy of content, disk to disk).

    If the internet were free and the ends were secure, we could just share folders with the people we trust…. perhaps some day we’ll get there. For now we have to use Flickr, YouTube and others as a guarantee of the safety of our stuff against virii, worms, etc.

  3. You live in a paradise, I’m really envious of your connect speed, I’m still linked to the Internet with 2Mbit/512Kbit ADSL line. At the speed of your connection you can also share without uploading to an external server but directly from your home.

  4. Comcast offers three different upstream rate caps: 2, 5, and 10 Mbps, and the vast majority of users choose the lowest rate. This would seem to invalidate your arguments that “Comcast apparently still thinks that downstream is all that really matters” and that users really care about their upstream rate.

  5. Stefano, you’re right. We have two homes. Our house in Santa Barbara is our dream home, literally. It took us years to build and we wouldn’t trade it for anything else around, which says a lot in a town like that one. Our apartment near Cambridge, Mass — the one our kid calls “alt.home” or “SHIFT_HOME” — is the upper half of a creaky 1920s apartment house. It has a few old-house charms, but it’s far less desirable than the Santa Barbara place in all but two respects: 1) it’s near great public transportation to Harvard, Boston, and lots of other places we need or like to go; and 2) there are three competing carriers with cables on the poles out by the street. I took one look at those cables and declared this place paradise. I dearly love the fact that there is real competition between three carriers — two with fiber and the third with pretty good co-ax offerings — for my business.

    I think all three could do a better job of competing. RCN (the other fiber company) only offers 2Mb upstream and in every respect appears not interested in providing more than a subset of what Verizon offers with FiOS on the same poles. (Or did the last time I checked.) They should zig to Verizon’s zag, and (at the very least) target the large tech/academic population here with services that appeal to them. But for now that’s a quibble. What matters is that we don’t have a duopoly here.

    So, what makes our neighborhood a paradise is not that we can get 20Mb symmetrical service from somebody, but that there is a competitive market here.

  6. Richard, here are the prices and rates from the quoted piece:

    “Performance” — $42.95 for 8Mbps/2Mbps
    “Performance Plus” — $52.95 for 16Mbps/2Mbps
    “Ultra” — $62.95 for 22Mbps/5Mbps
    “Extrme” — $139.95 for 50Mbps/10Mbps

    Customers can double their downstream at “Performance Plus” for $10. They less than double it at “Ultra” for another $10. Now they’re at $63. They more than double that when they jump to “Extreme” at $140. But that’s a helluva lot more money than they’ve ever spent for Internet — especially when the primary offering is still TV.

    Thus the marketing here encourages users to stick with one of the two bottom tiers, regardless of their upstream demand. And Comcast is right to assume that downstream is what matters most, to the most people. That will probably always be the case.

    So I also agree that the upstream demand isn’t there yet. My main point is that it will be. At some time in the future, demand for upstream capacity will drive some carriers to compete on that feature. (In fact, Verizon is already doing that, though not very aggressively. Yet.)

    It is inevitable that data demands will exceed TV and telephony demands. This is already starting to happen with cellular systems as well. The iPhone is a data device that also happens to be a phone. It’s also a harbinger. When data demands reach that point, upstream capacity will become much more of a selling feature for the carriers.

    I look forward to productive competition there, and minimal regulatory interference with it.

  7. I went back and checked the arguments we made when we first discussed the asymmetry of broadband networks in 2003, and I see that some of the facts have changed since then. Your piece was “Saving the Net”, http://www.linuxjournal.com/article/6989
    and mine was “Symmetry, Control, and Progress”, http://bennett.com/blog/2003/07/symmetry-control-and-progress/ .

    I speculated that DSL would be more symmetrical than cable, because of the way DOCSIS 1.x worked, but it hasn’t worked out that way. The Comcast line of services is the most symmetrical high-speed home offering short of the 20/20 plan Verizon offers, with asymmetry ratios ranging from 4.5:1 to 8:1. AT&T’s vDSL ranges from 6:1 all the way to 12:1, and they won’t sell you more than 1.5 Mbps upstream for all the money in the world. They offer a low-ball 1.5 Mbps down, 1.0 Mbps up option, but I doubt many people buy it for the symmetry.

    Most of the demand for broadband bandwidth is to accommodate video downloads as people begin to supplement cable tv with VoD. The Netflix deal with TiVo is a good example: I watched a movie from Netflix via TiVo over the Comcast network last night that was alleged to be HD. I’ve been using Netflix streaming over the HTPC I have connected to the same TV set, but the TiVo implementation has better audio and is easier to control.

    It seems to me that the principal driver for upstream bandwidth is P2P, which the carriers are beginning to embrace as it becomes domesticated by P4P, caching, and the IETF efforts. At some point, the carriers will actually be the ones pushing more symmetrical broadband in order to relieve their own P2P-related transit costs; it’s much better for them to keep the transfer in the neighborhood.

    So what we’ve seen in the last five years is broadband becoming both faster and more symmetrical, with the carriers (other than AT&T) becoming boosters of fatter upstream pipes.

    Neither of us predicted the 20/20 service five years ago, for example, but there it is.

Leave a Reply

Your email address will not be published. Required fields are marked *