Jukebox in the sky, or in our hands?
June 8, 2009
Further to our discussion of streaming vs downloading a couple of weeks ago, here’s a short article from Digital Renaissance, “The Future of Private Copying,” that came out about a year ago, in which the author gives projections for the average size of a hard drive. I’ve taken the data from that article* and plotted it; note that it’s a straight line on a semi-logarithmic scale (that is, the size of hard drives is increasing exponentially, Kryder’s Law). Also note that the average song is about 5 MB (0.005 GB) as an MP3 and about 25 MB in a lossless format. As the author writes, “How will this development affect private copying? When music fans can say: “I have all the music from 1950-2010, do you want to copy?” What kind of business models will be viable in such a reality?”
I’d love to see something similar forecasting the use of the wireless spectrum, which I know is a bit up in the air right now because of the delays in switching from analog to digital television broadcasts. If anyone can point me to a good reference, I’d really appreciate it.
*I haven’t sourced the data, but a quick troll through Dell and Apple’s site suggests that the numbers are in the right ballpark, and the shape of the curve is documented. It is an extrapolation though, which is always a bit dangerous.
EDIT: In the comments, Mark Chang writes:
The spectrum is there. It all depends on regulation and deployment costs in a country like the US. For instance, in Korea, SK Telecom has had MelOn, a streaming music service, since 2004. AFAIK, you can subscribe and stream all you want to your phone….
I would say that storage is free, bandwidth is where you pay. And you end up paying the provider of said bandwidth plus the fees to cover the licensing of the material. Since in the US, the copyright complexities and general craziness of the recording industry, and the propriety-stuff-first thinking of the wireless carriers, it becomes orders of magnitude harder. But only from a business perspective. This is all easy from a technology standpoint, and really, advantageous from a cost perspective.
zed equals zee 
Something that doesn’t seem to be addressed by this but that I think is an important side point is tracking the physical size of hard drives over the same period. When you can spend the same amount of money on 5GB you can attach to a keyring, 250GB you can carry in your pocket, or 1TB that can sit on your desk, those tradeoffs impact (I would argue significantly) the consumer decisions that go into the makeup of the “average capacity of regular consumer harddrives.”
Ah – I glossed over that. Kryder’s Law is actually about the areal density of storage (how many bits can be crammed into a square inch, say), so it’s really about how much storage you can fit into the same physical size.
But Kryder makes a related point in the SciAm article: incremental change in storage capacity can lead to transformative change in the user experience. “Who would have predicted the success of hand-held digital audio players? We completely missed seeing the iPod coming….Today the density of information we can get on a hard drive is much more important to enabling new applications than advances in semiconductors.”
The spectrum is there. It all depends on regulation and deployment costs in a country like the US. For instance, in Korea, SK Telecom has had MelOn, a streaming music service, since 2004. AFAIK, you can subscribe and stream all you want to your phone. Options include streaming to your other devices (like a PC), purchasing songs, ringtones, and a rental model for individual songs instead of the subscription model.
It is popular, and I’m sure there are other models in tech-forward countries like Korea and Japan. The awesome comes with the songs that released either exclusively or early using the service.
Of course, Korea is not the same mobile landscape as the US. With higher population densities and government subsidies (okay, the US has plenty of those), coverage is ubiquitous.
I would say that storage is free, bandwidth is where you pay. And you end up paying the provider of said bandwidth plus the fees to cover the licensing of the material. Since in the US, the copyright complexities and general craziness of the recording industry, and the propriety-stuff-first thinking of the wireless carriers, it becomes orders of magnitude harder. But only from a business perspective. This is all easy from a technology standpoint, and really, advantageous from a cost perspective.
Thanks, Mark! I especially appreciate the non-USian perspective.
But it’s still a little opaque to me: if ‘storage is free, bandwidth is where you pay,’ why is it cheaper/better to pay a provider for streaming rather than storing it locally?
Good question. I see it a couple ways.
If a large enough collection is put in the cloud, you don’t need to store any of it. No more managing your music. Just on-demand it! So convenient. The music is stored somewhere already, so it is just a matter of duplicating it. At some point, storage becomes cheap *enough*. Of course, this doesn’t cover every single song ever. So for those, it would make sense to store it on your own. But that service can be easily synchronized into the cloud, and then you would pay the same provider for that privilege. Small technical hurdle.
Another way to think about it is in efficiencies. If you and I store the same 50 songs, it would be more efficient to store them with the service provider, so they can be in charge or resampling or reencoding it so that it best utilizes their bandwidth and keeps up with encoding technologies. It is in their best interest (if they want happy customers) to keep lots of copies around in different bitrates so that they can maximize their pipes and deliver your content. If they let you try and stream 320kbps OGG, a) it might not work on your phone, b) the bitrate might be non-optimal for the network.
From the provider’s perspective, if they get all customers to commit to not having music at all, but rather subscribe to it, they are happier. Control of content, control of distribution (oops, that’s a monopoly, but whatever), and (somewhat) happy customers. Not the question you asked, but I thought it was worth mentioning.
Finally, and I put it in the cases above, it would likely be technically easier to centralize the storage with the provider, rather than deal with everyone’s network issues. So if you store your MP3s at home, you have to have a machine on with the storage live, have a network connection up, have software to manage that network connection and share the music and handle credentials, and then, the networks must talk. If it’s all one network, it makes it easier. And easier, ideally, means a better experience for the user.
Now, this doesn’t even scratch the surface of content management. But hey, you asked about storage/networking :).
[…] Streaming vs downloading: Do we really want a jukebox in the sky? Or does it make more sense to hold it in our hands? […]
[…] I do find it intolerable that there are 200 channels and barely anything worth listening to). As streaming becomes an increasingly viable alternative to downloading, is something similar going to happen […]
[…] power increases exponentially. Hard drive storage increases exponentially. And the energy density in batteries? Decidedly not exponential. Lithium ion batteries are thought […]