Digital Noise: Music and Tech

Bill Gates has said that prognosticators often overestimate the amount of technological change that will happen in a year, but underestimate the changes that will take place over a decade. With the Zeroes coming to an end this week, and Steve Guttenberg's recent column questioning the viability of recorded music in 2020 as inspiration, here's my pick of 10 trends in music and technology that will shape the next decade.

Will the original iPod become an object of fetishization in 2020, like vinyl records are today?

(Credit: Apple Computer, via Wikimedia Commons)

Songs instead of albums
Musicians will always find ways to record their music--it's a fundamental drive, like painting for a painter or writing for a writer. But I agree with Guttenberg that fewer musicians will release suites of songs organized around a common theme or sound. As much as I love my long-playing records, they arose out of economics rather than art--they were a convenient way for companies to bundle multiple songs (particularly songs that might not have sold as singles) in an affordable package. With digital files already taking the place of physical recordings, there's almost no economic reason for the album to persist. By 2020, the concept of the album will be an anachronism with a few vocal adherents--like vinyl records are today--but most music will be released and consumed as songs.

Streams instead of downloads
Where did we get the idea that digital music has to be downloaded? It started with the CD and file-trading networks--content owners wouldn't sell us music in a form that could be consumed on our computers, so we ripped our own and swapped the files through Napster and its brethren. But now, every time a new song or album comes out, or we rediscover an old act, we have to rip or download the recordings, then transfer them to whichever device(s) we want to play them on. There's got to be an easier way!

If you had access to every song ever recorded, on any device, from any location with an Internet connection, wouldn't you rather pay for that service than buy a new CD or two every month? People say they want to own music, but when it's just a digital file, what do they want to own--a collection of ones and zeroes sitting on a segment of their hard drive? Why bother?

I think the real problem is that today's streaming services don't give you every song ever recorded and don't work on every device, and broadband data access--particularly wireless--is not ubiquitous. Those flaws stem from business problems (licensing, DRM, format incompatibility, and insufficient broadband infrastructure) rather than technology problems. And the business problems are gradually being resolved--look at the introduction of Rhapsody and Spotify for iPhone, and Apple's acquisition of streaming music service (and music locker) Lala. By 2020, most professionally recorded music will be consumed as on-demand streams and people won't pay by the track.

In the cloud rather than on hard drives
Some songs will never be available on demand--think of tracks from friends or obscure independent acts, or live covers (where licensing can be incredibly complicated, involving multiple performers and songwriters). But as users become accustomed to listening to more professionally recorded music on demand, they'll expect their personal collections to be available in the cloud as well. After all, who wants to spend time backing up a 120GB music collection on an external drive, or choosing particular recordings to eliminate in order to clear space on a cell phone?

This is where Apple's Lala acquisition really makes sense--imagine if iTunes served not only as an on-demand music service but also as a locker for songs you'd previously downloaded, ripped, or obtained elsewhere. Suddenly, the 16GB of storage on an entry-level iPhone would seem generous instead of paltry.

Fidelity rather than file size
Once our music lives in the cloud, we'll no longer have to worry about running out of space on our local drives or devices. Microsoft's SkyDrive already offers 25GB of online storage for free, and I could easily see that increasing one-hundred-fold by 2020. That's right: free terabytes of storage. It'll take a little bit longer, but eventually bandwidth--even wireless bandwidth--will increase to the point where streaming lossless digital files makes sense. Listeners will rediscover what they've been missing--detail in the midrange, and tons of information at the low and high ends of the spectrum--and the era of the MP3 will be looked back (and down) upon as the dark ages of audio quality.

Extras become standard
Again, with concerns over storage gradually disappearing, what's to prevent artists from packaging their music with artwork, lyric sheets, video outtakes, and even interactive applications? Today's artist-specific iPhone apps will become standard. Casual fans will stream a couple songs for free. Hardcore fans will pay to download the entire app and pore over it obsessively.

Production rather than consumption
Digital technology has already democratized the recording process--what used to take tens of thousands of dollars and a professional studio can now be accomplished with a laptop and a free program like Garage Band or Audacity. The results usually don't sound as good, but the experimentation process is fun, and sometimes a gem emerges. Digital technology and the Internet have also made promotion and distribution far easier than they were a decade ago. By 2020, music fans will spend almost as much time creating and sharing recordings with their friends as they do listening to professionally recorded music. Don't believe me? Think of this: 10 years ago, writers were a comparatively rare breed. Now, everybody's got a blog, or at least a Facebook page. In another 10 years, everybody will be a musician--or at least a recording artist.

Suggestions rather than searches
In a world of on-demand music in the cloud, search will become vitally important. Users will want to be able to find songs not only by title, album, or artist, but also by a few snippets of lyrics, or even by humming or playing part of a melody. (Imagine a combination of the voice search function available on Google Mobile with an advanced version of technology like Shazam, which can identify recorded music from a few snippets.) But search is only part of the question--once everything's available, how will users decide what to listen to? By 2020, personalized recommendation services, like those provided by Pandora, Slacker, and MOG, will become even more important than search, and will have to be integrated into any on-demand music service that hopes to survive.

Festivals rather than big concerts
Live music is already a long-tail world--with the exception of old, established acts and the very occasional pop sensation, very few bands can fill large arenas or football stadiums. This trend will accelerate as the last bands from the golden age of radio retire, labels take even fewer big promotional risks, and the market continues to fragment under the explosion in recording releases. In 2020, no single act will be able to sell 50,000 tickets at Qwest Field like U2 hopes to do this summer. Instead, the only shows that will pack large arenas will be festivals, where listeners can pick and choose among dozens of acts and classes of entertainment--just like they'll be doing online.

Spectacle rather than personality
With recording revenue plunging, bands must draw fans to their live shows in order to make a living. The common wisdom today dictates that musicians need a personal connection with their fans. They must blog, tweet, maintain their MySpace and Facebook profiles, and generally act like your next door neighbor who's always pestering you to see his band. There's a word for receiving "personal" messages from your favorite 100 bands--it's called "spam." Eventually, this cloud of self-promotional noise will dissipate, and will be replaced by old-fashioned word of mouth. Only acts that put on a great show--not just singing and playing songs, but entertaining in the old-fashioned sense of the word, with video and stagecraft and humor and spectacle--will cut through the noise. Bonus points for the first act that somehow integrates an audience-accessible game console into their act.

Retro takes on a new meaning
In 2020, the original iPod will be almost 20 years old. As the music world is overtaken by a nearly infinite selection of high-fidelity music, streamed over super-fast wireless connections to increasingly inexpensive portable devices, hardcore nostalgists will drag out their first-generation iPods and fill them with treble-heavy 120kbps MP3s. Meanwhile, grandpa will still be down in the basement with his collection of LP records and his lava lamp.

Yesterday, I compiled my list of the five most welcome products for digital audio that came out in 2009. Today, I'm following it up with my list of the year's five biggest digital audio duds.

An image from the infamous online commercial for Songsmith, Microsoft's reverse-karaoke software.

(Credit: Microsoft)

Zookz. The breathless pitch got me interested: a mysterious online service was getting ready to compete against subscription-based download service eMusic. But where eMusic limits users to a set number of downloads, this mystery service would offer unlimited music and movie downloads. How could this be? Wouldn't users just download all the material they wanted then cancel their subscriptions? How could content owners let this happen?

The trick: Zookz was based in Antigua, and according to the company, this meant it wasn't subject to those silly little things known as U.S. copyright laws and royalty rates. Unfortunately, the country of Antigua didn't agree, and days after the public beta launched, Zookz disappeared into the digital ether with a promise to refund subscribers' money.

Jango Artist Airplay. I liked Jango's online radio service back when it launched in 2007. This year, in what looked like a desperate bid for new revenue, the company launched a service called Artist Airplay, in which bands could pay for placement on appropriate Jango stations. While Jango's CEO tried to tell me this was a reasonable new marketing opportunity, I saw it as a new form of the old pay-for-play deal that beginning bands often fall for.

With regular marketing, everybody pays more or less the same amount for the same class of services and the music sinks or swims on its own merits. With pay-for-play, artists buy exposure. There's only one problem: the resulting conflict of interest drives discerning listeners--including people who might actually pay you for your music--away. Jango Artist Direct may not be as stark as those pay-to-play "showcases" and "band battles" where all the audience members are other bands and their friends, but I believe it's better for beginning artists never to start down this slippery slope. Then again, I thought users would never be ignorant enough to click on search advertisements in massive numbers, which is one reason why Sergey Brin and Larry Page are multibillionaires and I'm not.

Vevo. As long as we're talking about Google, let's talk about YouTube, which the search company owns. It's a great source for music videos, and its APIs have formed the basis for music-finding apps like Muziic and TubeRadio. Users love it. Unfortunately, the companies and artists who own the copyrights to many of those music videos don't love it--the videos are expensive to produce, and the ad revenues from YouTube and other online video sites are scanty to nonexistent. Google is also lukewarm about music videos on YouTube, finding that the cost of policing copyright and complying with take-down notices is more than the money they can earn from selling ads.

In December, two record companies--Sony and Universal--joined together with Google in a new joint venture, Vevo, to address the problem. This was supposed to be a back-end business-to-business kind of deal, where YouTube users wouldn't know (or care) that certain videos were actually being provided exclusively by Vevo, which would sell short video advertisements to run before them. Unfortunately, the glittery launch party drew undue attention to Vevo's own site, causing its servers to buckle under the load. The entire episode left music fans scratching their heads.

Songsmith. The idea wasn't all that bad. Karaoke is fun. Making music on computers is fun. So why not, reasoned some Microsoft researchers, create a program that fills in audio accompaniment as users sing. Unfortunately, the $29.95 price and unbelievably mockable promotional video turned Songsmith into an Internet laughingstock. Later videos featuring Songsmith's accompaniment to the vocal tracks of songs like Queen's "We Will Rock You" and Van Halen's "Running With the Devil" highlighted the silliness.

CMX. In August, reports broke that the four major record labels were considering a new type of "digital album" format that would include album art, lyrics, and extra content. There was just one problem: Apple was already building its own competing format, code-named Cocktail and eventually released as iTunes LP. I think the entire concept of a digital album is weird anyway: I'm not convinced that lack of album art is a big reason why users are buying singles instead of albums. (The real reason is the Chumbawamba factor, or the fact that a lot of albums contain only one or two good songs.) And iTunes LP doesn't exactly seem to be taking off, although some of the extras--outtakes and videos--are actually quite valuable. But creating a competing format that wouldn't be supported by Apple? That's just plain dumb. To be fair, we haven't heard anything about CMX since iTunes LP launched. Here's hoping this product is killed before it's ever born.

Vevo CEO and President Rio Caraeff more or less confirmed on Wednesday my suspicion that the music service was not created to serve a new need for consumers. Rather, it was built to help advertisers and content owners (including labels, artists, and music publishers) capitalize on music videos, and to help Google (YouTube's owner) offload some of the cost associated with administering rights to them. In other words, this isn't a business-to-consumer play, it's more of a business-to-business arrangement.

Vevo's launch glitches appear to have been resolved.

(Credit: Vevo)

As he put it: music videos are popular online, fans like them, and content owners think of them as premium content. But they're too widespread, appearing on YouTube, AOL, and many other sites, and the user experience is way too varied--when a user searches on a song name at YouTube, they might get multiple copies of the exact same music video, plus user-posted remixes, live versions shot with a cell phone camera, and even parody versions. More generally, music videos grew up as a promotional tool for albums, and advertisers and users have come to see them as a commodity rather than prime product. Consequently, advertisers haven't been willing to pay much to place their messages next to them, and online music videos have lost money at a "staggering" scale.

Vevo is meant to provide an online clearinghouse for label-approved music videos--the kind of professionally shot videos that often cost half a million dollars or more and used to form the backbone of MTV. Vevo will be the exclusive distributor of these videos, and will handle all licensing and ad sales, although partner Google is handling the actual video hosting and streaming. In other words, if you're running a video site and you want to post a video that's in Vevo's catalog, Vevo will be your only source. By enforcing scarcity, giving advertisers a central place to buy ads, and controlling the user experience--for example, ensuring that there aren't many copies of the same video on YouTube--Vevo believes that advertisers will be willing to pay much more to appear next to these videos. So far, this seems to be true: according to Caraeff, advertisers have been willing to pay between $25 and $40 per thousand views (CPM, in advertising parlance) for Vevo-provided videos, compared with average market rates of $3 to $8. Caraeff claimed that artists and publishers will get about 50 percent of all revenues from these ads--a much higher percentage than they earn from recordings. This is why Mariah Carey and U2 were so excited about the launch.

Interestingly, Vevo will also curate unlicensed videos. For example, if somebody creates a remix of a Beyonce song with an associated video, and it becomes a runaway hit, Vevo might try to claim the video, add it to the Vevo catalog, and handle licensing for its content owners. Caraeff claims they're not going after the home video of your dog skateboarding to your favorite song, but professional-looking videos that have never been claimed, and therefore aren't making any money for anybody. (YouTube doesn't sell ads against unclaimed content for fear of copyright liability.)

So what's in it for Google? Simple--although YouTube has tons of viewers, it also has more inventory than it can sell advertisements against. Licensing for music videos is complicated, and not in Google's core area of expertise. Google is happy to hand this task off to Vevo and accept a lower percentage of advertising dollars because it believes the cost savings and higher CPMs will eventually make business sense.

Finally, about the botched launch: As Caraeff explained, Vevo was basically a B2B play, and the company didn't expect many users to visit its site on the first day. But the publicity created by the big launch party drove massive interest, and the company got more traffic in its first hour than it expected for its entire first year. For what it's worth, the company has added 32 servers in the last 24 hours, and I'm now able to get videos to play on the site with no problem.

In addition, Vevo didn't think it was critical to launch with a full complement of content--remember, it's mainly a back-end and clearinghouse for YouTube and other sites, and if you were watching videos there yesterday, you'll still be watching those same videos there tomorrow (as long as a takedown notice hasn't been issued). So Vevo launched with only about 15,000 videos from Sony and Universal Music. In January, it will add about 30,000 more from EMI and several independent distributors.

I still don't understand why they launched Vevo.com as its own Web site, but at least I understand the thinking behind the company. It won't change my behavior--I'm still going to YouTube, and if a video happens to be provided by Vevo, I'll know that the artists are making some money from it. Fair enough.

Vevo, the new music-video site operated by Google (which owns YouTube) and co-owned by three of the four major labels (EMI, Sony, and Universal; Warner Bros. not participating), launched on Tuesday to some fanfare in New York. Big music celebs rubbed elbows with Google and label execs in the kind of self-congratulatory bash that only the entertainment industry can pull off.

This is as far as I got when I tried to play U2's video for "Even Better Than The Real Thing" on Vevo.

Maybe that's too harsh, but I visited the site on Wednesday and I quite honestly can't figure out who or what it's for. It's got music videos, but only from three of the four majors and some independent distributors, which leaves huge swaths of the entertainment landscape blank. As far as I could tell from a search of the site--and the search engine should work, given that Google's behind the site--Vevo is sadly lacking in classic rock and modern indie rock, which are the two genres I listen to most.

There's no Roger Waters or Pink Floyd. No Pixies. No Grizzly Bear. No Led Zeppelin. No Animal Collective. No Beatles. No Eric Clapton. And on and on and on. Go ahead and try your own, you'll get the idea--if you can get the site to work to work at all. (It's been plagued by glitches since launching, and my effort to play U2's "Even Better Than the Real Thing" around 1 p.m. Wednesday met in failure--the video froze around 80 percent loaded.) Apparently, if you can get a video to load, you'll probably have to watch a video advertisement before it starts.

The aforementioned artists are all over the place on YouTube--a site that everybody knows and loves and is largely free from video advertisements. And because Google is behind both sites, videos licensed for Vevo will also appear on YouTube, with Vevo getting the credit (and ad bucks) when a YouTube viewer watches a Vevo video. So why would anybody go to Vevo? Why bother building it, instead of just making it a new channel on YouTube? Who is this for?

The music industry, that's who. It wants to control the online music video experience--Universal Music Group CEO Doug Morris flat out said so. They're tired of mean old Google using its content to sell advertisements. But I honestly can't imagine why Google agreed, unless the labels held it over a barrel, refusing to license their content for YouTube unless Google agreed to help them create a music-industry answer to TV-streaming site Hulu.

Here's the thing. The big winners in the old music industry of yesteryear don't like the Internet. U2 manager Paul McGuinness has said that Internet service providers should bear part of the blame for piracy. Doug Morris earned some scorn two years ago for a Wired interview in which he revealed that his label didn't even try to come up with a digital strategy in the early days, when P2P file-trading networks first started becoming popular.

If you don't like the Internet, you're not going to be able to create an Internet service that people like. More than 15 years into this Interwebs thing, some people still don't understand that if they create an experience that users don't like, it won't get used. It's like they're still living back in 1973 when we only had three TV networks and one or two daily papers and a handful of local radio stations. We now have unlimited choice. Offer me something better than what's out there now, or please, save yourself some money and effort and get out of my way.

Hulu succeeded not only because the TV companies played hardball, refusing to license their content too broadly to other distributors, but also because it launched strong, with a big selection of desirable content. Vevo could certainly turn itself around, but its launch doesn't look very promising. I suspect it'll end up like every other entertainment industry effort that offers no clear benefit to users: on the digital scrapheap.

Here's an interesting study in contrasts. When MySpace acquired iLike back in August, MySpace left the site mostly intact. The iLike home page is still there, you can still add iLike's music-finding and sharing application to your Facebook page, and iLike is given prominent placement in Google search results for music-related queries, thanks to an October deal between MySpace and Google. (That deal also included several other companies.)

What else do you need to know?

(Credit: MySpace)

On Tuesday, MySpace completed its acquisition of Imeem, a service that used to let users upload music and videos and share playlists. In the press release announcing the finalization of the deal, MySpace noted that it will be "working as quickly as possible" to migrate "aspects of" Imeem to MySpace Music.

What does that mean? As of Tuesday, it means that Imeem's gone. Imeem.com and my personal Imeem profile both redirect to MySpace Music. The press release promises that MySpace will "be working to offer users the Imeem playlists they've created on MySpace Music," but the phrasing on the FAQ doesn't leave me much hope--"will attempt to transition" leaves a lot of wiggle room.

The free Imeem app for iPhone is still in the iTunes Store, but it's going to be inoperable until MySpace pushes an update out, and there's no telling how close the new app will be to the old one.

I always found Imeem a little confusing--especially in its early days--but I did like its user-contributed content model, which gave it a large selection of music, and I personally used it to post a lot of recordings from long-dead bands I used to play in. But I suspect that user-contributed content is one reason why MySpace is treating Imeem so differently than it treated iLike: some of that content may not have been properly licensed from content owners. In contrast, MySpace boasts that its content is "fully licensed."

Unfortunately, using MySpace Music is still a pain. It's gotten a bit better since the last time I took a close look, but I still can't find the home pages for certain artists using the MySpace search engine (Google works), the advertisements are still annoying and intrusive (text ads above search results; audio ads on a music site), and there are still no playable Led Zeppelin tracks. (Although MySpace does have The Beatles and AC/DC, which most other services don't.)

There's no free lunch, and the recent consolidation among free online music sites is beginning to look like a crackdown.

I had a fascinating conversation with MediaNet CEO Alan McGlade on Friday morning. Unless you're deeply involved in online music, you probably don't know MediaNet, but it's the back end powering a lot of music services you might have used, including MOG's subscription service that launched earlier this week, as well as Microsoft's excellent Zune Pass subscription service and iLike's online music marketplace. (MySpace acquired iLike in August, and in November, links to iLike's service began appearing directly in music-related search results on Google.)

Fox Interactive used MediaNet's technology to embed this list of Aerosmith songs in a story about the band. Readers could then listen to a sample or buy the song.

(Credit: MediaNet)

They've also got more history in online music than just about anyone. The company started off as MusicNet, with part-ownership by three of the then-Big Five major labels: BMG, EMI, and Warner. They powered RealNetworks' music initiatives before RealNetworks bought Rhapsody. They powered Yahoo Music. They powered MTV's online music store.

These early stores went nowhere. Content owners insisted on digital rights management (DRM) restrictions, which meant that content from these stores had restricted use rights and couldn't be played on every device--including, in most cases, Apple's iconic iPod. Setting up a store using MediaNet's platform often took 18 months and significant technical expertise. In the meantime, Apple focused on a dedicated online store for its own devices, and completely dominated the market for music downloads.

But the landscape has changed. Labels don't want to be beholden to Apple. They no longer insist on DRM for single-song downloads, and have realized that the more outlets there are for their digital music, the more customers they'll reach, and the more sales they'll have. (Amazing it took this long to figure out.) MediaNet is, in my opinion, incredibly well positioned to take advantage of this sea change.

In October, the company released a set of technologies called MN Open that make it almost trivially simple for companies to add a wide variety of music consumption options to their Web sites. Sure, companies can still use MediaNet to build an end-to-end service like MOG.

But say you're Fox Interactive and want to make a story about Aerosmith more engaging. Using a MediaNet component, Fox created a link for the first mention of the word Aerosmith that took users to a page with more information about the band, and links to play and buy some of their popular songs. Fox also posted Aerosmith songs in a box directly on the story page.

MediaNet handled all the heavy lifting: licensing the music, streaming the samples, and fulfilling the transaction. Fox kept its brand and design throughout the process, and users didn't have to leave the site to buy the song. Best of all for Fox, it didn't have to make any up-front payment to use MediaNet's technology. Instead, MediaNet takes the customary cut of any song purchased through the site (about 30 percent, if it's anything like Apple). The model's the same for sites that offer free ad-supported streams or subscriptions--MediaNet takes a portion of each transaction, then handles payment to the content owners.

Now imagine this kind of integration on sites for radio stations, record labels, or your favorite bands. Imagine your ISP or cell phone carrier offering you a music subscription service bundled with your Internet service or smartphone. In this world, users won't have to go to iTunes or Amazon MP3, or subscribe to Rhapsody (or MOG for that matter). Music will be available for consumption everywhere. And content owners will get paid regardless of where users buy it.

According to McGlade, it's already happening--he said MediaNet is adding about one new distributor per day, and has already got about 50 customers using the MN Open platform. One site, GetPlaylists.com, was able to add playable song samples and downloads-for-sale in only two days with MN Open, according to McGlade.

Thanks to this upsurge, the company--which is owned by a private equity firm and no longer has any direct ownership affiliation with the major labels--has recently crossed over into profitability. A rare situation indeed in today's online music landscape.

It's a great vision, and something that Microsoft, the original platform company, could have done. But Microsoft spent years pushing the Windows Media Platform, which made heavy use of Microsoft codecs and file wrappers (instead of MP3s, which were becoming the industry standard). Microsoft also spent a lot of effort trying to enable the labels' DRM demands--for example, by building a platform to enable subscription-based downloads to be transferred to portable devices. Then, just as the labels were getting ready to abandon DRM, Microsoft basically gave up pushing Windows Media as a general-purpose platform for distributors and device makers, and instead started trying to mimic Apple's end-to-end software+service+device with the Zune strategy.

Talk about an opportunity lost! Instead of struggling along with something like 2 percent of the digital media player market, Microsoft could have ended up powering the music technology on thousands of Web sites.

Another aside: while MusicNet offers a lot of flexibility for distributors--downloads, samples, free streams, or subscriptions are all supported--McGlade is most bullish on subscriptions as the digital business model of the future. He admits that old fogeys accustomed to CDs and vinyl will have a hard time giving up the concept of ownership, but suggests that today's teenagers don't care--they want music on demand from any device, any time, in any location, and don't need to have the files physically present. McGlade thinks that subscriptions will have the best chance of taking off if they're bundled with some other product, like ISP service.

Scoff all you want about subscriptions, but the concept keeps coming up: music industry expert Donald Passman also believes they're the best chance for the music industry to thrive in the future. Even Apple finally seems to be bending to the idea of streaming music with its acquisition of Lala, although Lala isn't a straight subscription service, but more of an online music locker with some free streams, plus fee-based individual streams.

MOG offered me a free trial to its subscription-based streaming music service, MOG All Access, which launched on Tuesday. The service costs five bucks a month, and gives you unlimited on-demand streams of more than six million songs from all four major labels and plenty of indies. The site is trying to differentiate itself from competitors like Rhapsody and Napster with high-quality streams--all songs are 320kbps MP3s--and some fairly sophisticated music discovery features, like playlists posted by musicians (David Byrne got the featured spot on the day of launch) and other fans with similar tastes to yours ("Moggers like me").

(Credit: Screenshot)

I've been sampling the service for about an hour, and I do appreciate the sound quality (although better volume-leveling between songs would be nice) and recommendation services. And I'd like to thank the designers personally for the ability to add any song to the already-playing queue--a feature I love on Grooveshark and my Zune HD and that I always miss whenever I use one of my iPods. But there's one big problem: song selection.

Because MOG is licensing content directly from copyright owners, there are big gaps from artists who simply don't want to participate in online music. The usual suspects like AC/DC, Led Zeppelin, and the Beatles are mostly missing in action. There are also strange gaps elsewhere. For instance, half the songs on the Pixies' "Surfer Rosa" are unplayable. Compared with Grooveshark, which relies on user-posted content, MOG has too many holes. And of course, Grooveshark remains free (although a $3/month subscription gets you a version without advertisements).

Song selection could improve over time as MOG signs more licensing deals, but I found some other related glitches as well. For instance, '70s folk artist Roy Harper, whom I often use as a test case to see how well an online service does with relatively obscure old content, has almost no playable content, but does offer a nice list of albums with links to Amazon. The only problem: when you follow the Amazon link on unplayable songs, it takes you to the Amazon Music front door--most of his songs aren't available for download there, either.

In its demonstration video, MOG touts its online radio service as a unique feature. When you're playing a particular artist, a slider lets you control how much variety you want, from "play only songs by this artist" all the way up to "play mostly songs from similar artists." It could use some fine-tuning, however. When I started a queue with a Modest Mouse song, it used Modest Mouse as the basis for its selection. Fair enough, but when I added songs by Talking Heads and Public Image Limited, the radio algorithm didn't account for those artists. It simply switched my radio playlist completely when the new songs started playing, showing all Talking Heads songs, all PIL songs, and so on. Pandora and Slacker do a much better job of creating custom stations based on multiple artists. (Although, of course, those services don't let you play individual songs on demand.)

Finally, as I wrote last week, I still think MOG's lack of mobile support is a fatal flaw, but one that could be easily remedied: Apple's approved iPhone clients for subscription services Rhapsody and Spotify, so why not MOG?

All of these flaws can be fixed, although licensing content takes time and convincing. I'm a big fan of competition, though, and MOG takes many of the best features of a lot of other services, combines them in one place, and improves on some of them. For that, the company should be commended.

For the last year or so, it's become clear that the economics of ad-supported streaming music services are not good for their creators or investors. As CNET's Greg Sandoval reported last week, the acquisition of streaming service Imeem by MySpace Music for pennies on the dollar is the latest bad news for the sector, following the bankruptcies of SpiralFrog and Ruckus and the similar fire sale of iLike to MySpace.

Offering your music via Spotify might help you fill up your piggybank.

Who's left? In the U.S., we've still got LaLa, which has the blessing of the major labels and seems to be enjoying dramatically increased traffic (as measured by Alexa) thanks to its recent deal with Google, and Grooveshark, which has kept a low profile. Neither of these services is purely ad-supported--particularly LaLa, which hopes to charge customers for downloads and "permanent" streams once they surpass a quota of 50 free streams a month.

But the service most often cited as the future of online music is Spotify. It's only available in Europe right now, but it seems like everybody who tries it loves it, myself included. Spotify offers a premium service as well, which offers portability and higher-quality streams, but the free service offers unlimited ad-supported streams, and that's the service that has everybody so excited.

But there's one small problem with the Spotify-as-savior story: it doesn't pay artists very well. According to this story in a Swedish publication, as translated and explained by the TorrentFreak blog, Spotify delivered more than one million streams of Lady Gaga's hit single "Poker Face" over five months. From these streams, she reportedly earned about 1,150 Swedish kronor--about $167--from the Swedish agency responsible for paying royalties. That's not even enough to cover the cost of four tickets to her upcoming concert in San Francisco.

If this story's true, why would any artist agree to make songs available on Spotify? With these kinds of payouts, it looks like music business expert Donald Passman is right--advertising is never going to support an online music service.

If you work in the music business, you probably already know the name Donald Passman. For the uninitiated, his book "All You Need to Know About the Music Business," which was first released in 1991 and comes out in a seventh edition today, is the book on how the music industry works. If you ever wanted to know how major and indie label deals are structured, the different types of royalties that musicians can earn and how they're calculated, what a personal manager does for a band, how much money artists make on tour, where your ticket fees go, or any of countless other nitty-gritty details, this is the book. Music industry people sometimes call it the bible, and they're not joking.

The seventh edition contains numerous updates since the last release in 2006, including the final resolution of the battle over Internet radio royalties, details of how royalties are calculated for iTunes and YouTube, and the latest developments in label deals. I had a chance to talk to Passman on the phone on Tuesday morning about the changes he sees in the music industry and where he thinks it's going.

Q: Since the last edition of the book was published in 2006, what's the biggest change to the industry?
Donald Passman: 360 deals [in which record labels get a cut of the artist's revenue from touring, merchandise, and other sources apart from record sales] are a tectonic shift in the way record deals get done. There was a smattering of them three years ago, but now it's become the norm. Early deals with Madonna and other established artists were really banking deals, where everybody knew the artist's track record and they were making a bet on what the future would be. Deals with new artists are quite different. The labels are saying that they're the only ones really willing to spend money on a new artist's career, and they can no longer make money just in the record business, so they need a cut of this other income.

We're also starting to see some industry patterns in how royalties and payments are calculated for digital music. So far, none of the business models have made a lot of money. Even iTunes doesn't make a huge profit. But part of the problem three years ago was that if you started a streaming music service, you had no idea what you were paying. Now we're starting to see industry patterns.

Do you still think it's worth an artist's time to pay or find funding for a professionally recorded demo in a big studio? Or would you recommend that artists invest that time and money in buying a computer and other home recording gear and learning how to record themselves?
Passman: There's no one-size-fits-all answer. The software that's available now is better than what professional studios used to be like 10 years ago. You can create an extraordinarily good demo in your house. The key is to create a demo that sounds like a hit so they don't have to use their imagination to hear it. People will tell you they can hear a diamond in the rough; they can't.

Do you think the rise of do-it-yourself online services like CDBaby, Tunecore, and Sonicbids will lead to a new thriving "middle class" of musicians--folks that aren't signed to a label, but spend most of their time recording and touring and make a decent living doing it?
Passman: It's a reality today. If you're a niche-type artist and you don't mind staying in your niche, you can make a perfectly good middle-class living that way. If you're more mainstream, you can use those techniques to build buzz and attract a label. Nobody's yet had a major career without a label behind them. I'm certain that's going to change, but for bands that want to be truly international and mainstream, they still need a label.

In your book, you mention that you favor unlimited music-streaming--the "celestial jukebox"--as the most likely future business model to succeed. Do you think ad-supported free services (like Imeem or Grooveshark) will ever be viable?
Passman: I'm skeptical that ad-supported music services will work in a major way. People have had a hard time monetizing music to advertisers because the music is so diverse. Advertisers don't know what type of music they're going to be associated with, so it's hard for these services to get high enough CPMs [impression-based advertising rates]. YouTube's had the same problem: you don't know what kind of video's going to pop up next to your advertisement. Free services will probably be a model, but I don't think they'll be the model.

Why do you think subscription-based services (such as Rhapsody) haven't really taken off?
Passman: They're not convenient enough, they're not truly cross-platform. For me, the ultimate would be anytime-anywhere access to any music for one subscription. On my computer, in my car, on a connected device, whether it's an iPod or something else, on an airplane when I don't have an Internet connection. Not just tied to one or two devices. I'm personally a believer in subscription services. People don't think twice about paying for cable, and when you stop paying it goes away. But with music, there's a kneejerk reaction because we're used to owning it.

We hear a lot of doom and gloom about the death of the music industry. Are you still optimistic?
Passman: I'm optimistic in the longer term. Not for the next few years. I think it'll probably get a little worse, and then we'll bump along the bottom for a while. I don't think we've hit bottom, and that's because I see the trends that are happening now. CD sales are declining and will eventually disappear, and retailers are making it happen. They're cutting back the floor space they devote to CDs, which means less CD sales, which makes them cut back more.

But the digital opportunity is huge. We can sell music to people who've never gone into a record store, people who never listen to music because they stopped listening to the radio at a certain age will now have access. Unfortunately, we're not there, technically or legally. The more pain the industry feels, the easier the legal side gets. The better the technology gets, the closer we get to delivering an experience people want.