If music subscription services were easy, everybody would be doing them and millions of Americans would be paying. Numbers from RealNetworks’ latest earnings show subscriptions are still one of music’s greatest paradox: so much potential but so few paying customers.

Rhapsody finished Q1 2010 with 650,000 subscribers, according to its earnings release last week, a 3.7% decline from 675,000 at the end of Q4 2009 and down 18.8% from 800,000 in Q1 2009.

It’s a familiar refrain. Napster was losing subscribers before it was acquired by Best Buy in September 2008 and hundreds of thousands more were lost when AOL shut down its subscription service. (Napster paid for AOL’s 350,000 subscribers in January 2007, bringing its total to about 900,000. Since Napster’s subscriber count stood at just over 700,000 in June 2008, it can be reasoned most of them didn’t stick around.)

Not long ago, Rhapsody was gaining subscribers. At the end of Q4 2007, according to a RealNetworks SEC filing, Rhapsody had 775,000 subscribers after adding 150,000 net new subscribers in Q3 and 25,000 in Q4. In 2008, the company launched a multi-million-dollar advertising campaign around its Music Without Limits initiative that included a new MP3 store, a partnership with Verizon (VCast) and full-song previews at iLike. By the end of Q3 2008, Rhapsody had competed a one-time migration of customers from Yahoo! Music’s shuttered subscription service.

Now, media darling Spotify has 300,000 paying subscribers and over seven million users of its free service in six markets. It’s a good start, but nothing more. To put it in perspective, Spotify has fewer paying customers than Rhapsody and Napster have lost in recent years. The game-changing gains have been made by only one company: Pandora.

The timing of Rhapsody’s Music Without Limits campaign couldn’t be more coincidental. In the same month, Pandora launched its hugely successful iPhone app. It can’t boast eight million on-demand tracks, but it obviously has enough music for a large section of the market. Most impressively, Pandora achieved a rare feat by the end of 2009, less than a year and a half after it launched its iPhone app: it turned a profit. In contrast, competitors are struggling to acquire users to scale to profitability.

The final verdict on the current subscription model has not been delivered, but its outlook is grim. New competitors are needed in the U.S. market to breathe life into a staid situation and, for a change, excite consumers. Given consumers ambivalence about today’s subscription market, it’s no wonder labels and publishers are desperately hopeful that partnerships with ISPs and device manufacturers will bring new life to subscriptions.

Pandora, however, provides reason for caution on subscriptions. The runaway success of a service with a small catalog and no ability to grant on-demand access – the exact opposite of the services most favored by content owners – shows people may be overestimating the demand for a celestial jukebox.