The music industry has posted surprisingly good numbers lately. But is it time to breathe easy?

Of course, the steady arrival of new consumers is still good news for the industry.

The long-ailing music industry hasn't so much found new channels as it has found new consumers.

The market has done surprisingly well. On June 30 this year, Jay Chou's new album Jay Chou's Bedtime Stories went on sale on QQ Music at 20 yuan a copy. Ten days later, Tencent vice president Peng Jiaxin told a press conference that the album had already grossed more than 30 million yuan — that is, 1.5 million copies sold.

1.5 million copies is a number the Chinese music industry hadn't seen in a long time. Only two years earlier, when Jay Chou sold his first digital album, Aiyo, Not Bad, on QQ Music, the figure was just 150,000. A report from the time described a slightly awkward scene: "When the host showed Jay Chou the sales figure of 150,000, Chou looked blank and asked: 'Is that good or bad?'"

Americans seem to be breathing easier too. In the first half of 2016, sales at BMG, a music-rights management company, rose 4.7%. Zach Katz, who runs the company's U.S. operations, said: "It looks like, after many years of crisis and contraction, this market is finally recovering slowly ... this is definitely a step in the right direction."

A Bloomberg story ran under the headline "The Music Industry Is Finally Making Money From Streaming." Citing figures from the Recording Industry Association of America, Bloomberg noted that this was the first time since 1999 that the U.S. music industry had grown two years running — the industry, it said, was in a "fragile recovery."

The report put the recovery down entirely to streaming, but the word "finally" in that headline is a reminder that streaming and paid subscriptions have both existed for at least a decade. So the better question is: why did channels established so long ago only now let the whole music industry exhale?

The answer, as noted at the start, is not that the music industry found new channels but that those channels finally got the new consumers they had been waiting for. Better technology, better products and a shift in what consumers are willing to pay for have given streaming — born ten years ago — a user base large enough, a decade on, to become the music industry's savior.

Changes driven by technology are always the most basic. Just as there would be no car industry without steel, streaming is built on top of bandwidth and smartphones.

Download Bedtime Stories on QQ Music and you get four audio-quality options — Smooth, Standard, HQ and SQ — with the album file ranging from 14.1 MB to 257 MB. Only when network bandwidth is good enough to cut the time an SQ-quality album takes to download can consumers be persuaded to pay for the album itself.

Seen this way, Spotify, born in 2006, and Pandora, which came earlier still, were both ahead of their time. Steve Jobs did not unveil the smartphone until 2007, and even 2G on a phone was something of a luxury then. Broadband, too, was stuck in the JPG era rather than the MP4 era — never mind the whole payment system that paid streaming requires.

Now that smartphones are everywhere and bandwidth keeps growing, streaming itself keeps getting better to use. In a sense it is now the only channel that delivers music straight to consumers.

Record stores are slowly disappearing; CD players and Walkmans are almost impossible to find. Downloading MP3s on a computer is a hassle. Only the phone stays with you around the clock — open an app, tap a few times, and the music plays. That convenience is what has made streaming the most important way consumers get music.

Technology aside, big companies have also brought the music industry plenty of new consumers — Apple, for one, which has charged back into the music business after iTunes.

Since Apple Music launched on June 30, 2015, it has gathered 17 million paying subscribers. A Bloomberg report quoted a record-label employee saying that most Apple Music users are new users rather than converts from Spotify. That means Apple Music has turned up 17 million people who had not used streaming and had no habit of paying for music.

To many, that result only just matches Apple's clout. There are close to 500 million iPhone users worldwide; Apple Music arrived as a built-in iOS app; and Apple already had a payment system in place through the App Store. Against all that, 17 million paying users is adequate at best.

Still, just as Apple upended the record business once before with iTunes, the music industry pins high hopes on the company. The future is unpredictable, but Apple has again been cast as something of a savior.

Apple is not the only big company re-entering streaming. Amazon recently launched its online music subscription service, Music Unlimited. It is separate from Amazon Prime and its companion service Prime Music, and is billed on its own at $9.99 a month. Users who already subscribe to Prime pay a discounted $6.99.

With 69 million Prime members, Amazon will probably bring the music industry a wave of new users, much as Apple has.

The digital downloads Apple once sold through iTunes and the CDs Amazon shipped both trained Americans to pay for music. In China, getting consumers to pay still takes an extra lure — the pull of an idol, for instance.

The best-selling run of albums on QQ Music came from Big Bang. Between May and August last year, the Korean idol group released four digital mini-albums on QQ Music, one a month, selling more than 4.1 million copies in total. Yet in July this year Wu Weilin, general manager of Tencent's digital music division, said QQ Music's cumulative digital album sales came to only 20 million copies.

As Gao Xiaosong put it in an interview after Alibaba Music launched Alibaba Planet: "In China, ask a fan to pay for music and they won't part with 8 yuan; ask them to pay for their idol and they'll spend 8,000. That's how it is here, so of course a company goes after the most important piece first."

One thing may hold true in both China and the U.S. NetEase Cloud Music said in a report that nearly 70% of its users were born in the 1990s or the 2000s. At the end of 2015, Spotify said 47% of its users were under 24, and more than 21% under 18. The new consumers streaming — digital music — has picked up are all young.

But can the music industry breathe easy? In China it is far too soon: the market as a whole is simply too small.

U.S. music-industry retail revenue in the first half of 2016 was $3.4 billion, while a report from the International Federation of the Phonographic Industry puts China's 2015 music sales at just $169 million.

Part of the gap is what each customer spends. Since it started selling digital albums, QQ Music has sold 20 million of them, and revenue has only just passed 100 million yuan. At 5 yuan a copy, a digital album really is dirt cheap.

Chinese users are more willing to pay than they were, but even QQ Music, with 100 million daily active users, has only 10 million paying ones. Close to a third of Spotify's users pay.

The Chinese music industry is small not only next to America's, but next to other Chinese industries.

$169 million is about 1.14 billion yuan — exactly the box office of the film Operation Mekong. True, from the first day of shooting to its release Operation Mekong also took more than a year and a great deal of money and manpower. But when an entire industry's annual output is worth no more than one film's box office, there is nothing to celebrate.

Output that low means not enough capital flows back into making music. When less money reaches creators, the industry holds little appeal for newcomers. Keep that up and exhaustion may be the fate this industry cannot escape.

Any industry, in other words, has to earn enough from the product itself to pay its creators. Only that virtuous circle lets an industry thrive. Look at the global music industry that way and it is clear that even with new consumers, it is nowhere near time to breathe easy.

In its 2015 annual report, Spotify lost money again: a net loss of $194 million, 6.8% worse than the $181 million it lost in 2014. Pandora reported third-quarter results not long ago; better than a year earlier, but still a loss of $61.5 million. Its stock fell more than 7% in the trading that followed.

Losses are now normal in streaming. Streaming already accounts for the largest share of music-industry revenue worldwide, yet not one streaming product makes money. These companies are not yet strong enough to build a virtuous circle inside the music industry.

Apple and Amazon may not have to worry about losses, but as one piece of a large company, how much attention their music businesses can command is a question of its own.

A month ago, Apple CEO Tim Cook personally congratulated Drake when his new album passed one billion streams on Apple Music. Apple is looking for its next source of growth for the day iPhone sales stall, and internet services — music chief among them — may be that future.

But Apple is betting on many things, cars and watches among them, and music is only one of many internet services. That makes streaming far less important to Apple, and in future, when a more profitable business comes along, music could be the one sacrificed.

In short, none of the key companies in streaming stand on firm ground. And although many see streaming as the music industry's future, its output still does not count for much against the industry as a whole. In the first half of 2016, U.S. paying subscribers spent $1.013 billion, while the whole music industry's retail revenue over the same period was $3.4 billion.

All of which blurs streaming's future within the larger music industry.

In fact this is less streaming's predicament than the whole music industry's. Cassettes, CDs, vinyl, streaming — all of them exist to meet one need: listening to music.

As technology advanced, the cost of making, copying and distributing music fell sharply, and piracy flourished. Pirated music meets that same need to listen, and it is free. That hurts the whole industry and shrinks its output badly.

Fighting piracy is one workable way to save the music industry. Beyond that, the industry has to find and cultivate new demand among consumers, and supply the products to match.

Getting people to buy CDs and artist merchandise in the name of "topping up your faith" is one approach; live shows, which cannot easily be copied or passed around, are another.

The first is the fashionable idea of the "fan economy." Open Alibaba Planet, launched this March, and the first thing you see is a page called Fan Playground, where you can join a star's fan circle and follow their news. But how to run a fan-economy business well is something the whole music industry is still working out — in the past a record label only had to sell CDs.

The live-performance model is far more mature by comparison, and that market is growing fast.

Between 2010 and 2014, annual revenue at Live Nation, the largest live-events company in the U.S., grew from $5 billion to $7.6 billion. In China, there were at most four or five music festivals in the whole country in 2007; by 2014 there may have been 200, under brands including Midi, Strawberry, West Lake and Hot Wave.

The whole music industry is being rebuilt as a result. Record labels used to handle everything — signing musicians, training and packaging them, pushing them into the market and making money from CDs. That is all still true, but the value chain has grown much longer.

Music itself is no longer the final product of that chain. There is far more to build around music and musicians: advertising, tours, merchandise, long-term rights management. None of this is new; it has simply never mattered this much to the music industry before.

However the music value chain changes, one thing does not: music and musicians remain the core of this industry. However strong the downstream machinery, it still takes a breakout musician and a body of breakout music to set it running. Li Yuchun broke through on Super Girl in 2005 and is still influential today. Lady Gaga was pushed to her peak by two songs, "Poker Face" and "Bad Romance."

Where new stars and new music will come from, and how, may be what this industry should worry about most.

The signs are already there. Can you name the last Chinese-language singer to really blow up? Xue Zhiqian? Set aside that he became famous as a teller of jokes — he was competing on a talent show back in 2005. As for the fresh-faced idols returning from Korea, EXO debuted four years ago.

Listening habits bear this out. The market-research firm Nielsen splits albums in two: anything released within 18 months counts as new, anything older as catalog. In 2005, new albums sold 389 million copies and catalog albums 229 million — new outsold catalog by 1.7 times. By 2015, new-album sales had fallen to 118 million, short of catalog's 122 million.

People are listening to fewer new songs, which means the music industry has less and less idea where its next stars and next music will come from.

The traditional star-making machine still centers on record labels. Their A&R departments scout, train and develop singers and artists, and A&R can eat up a third or more of a label's total revenue.

But as the record business shrank, labels spent less on A&R. Over the ten years from 2005 to 2014, Warner Music's A&R spending fell 21.7%. Adjusted for inflation over that decade, the drop widens to 35.4%.

Nothing has replaced it. Talent shows, music variety programs and internet platforms cannot take on the job of developing new artists that record labels once did. On that front, the music industry is nowhere near able to breathe easy.

For now, of course, rising music-industry revenue, pulled up by streaming, is still the more important good news. Money coming in can at least keep a gasping industry alive — and the luckier part is that these new consumers are the right age.