Norway has started switching off (almost) every FM transmitter, replacing it with DAB+. But why aren't they just streaming everything? I do some tedious maths. And here's the maths written down, because, heavens, this was a difficult podcast to make even approachably interesting.   This week, FM broadcasts in one area of Norway finally fall silent. The majority of radio stations move over to broadcast on DAB+, alongside online and DTV. The whole country will have turned off FM by the end of the year. (Some small local stations continue for now on FM, but stations that get 95% of listening will move). DAB+ is broadcast radio, just like FM. You need a new receiver, but it works exactly the same: broadcasting from a tower, being received on an antenna. But why bother with broadcast radio, people are asking, when streaming is all we need? Let’s do the maths. In many countries, you rent your FM (or DAB) broadcast equipment from a transmission provider, who deal with maintenance and the electricity bills. A station I picked at random in the UK spends US$6,000 a month on their FM transmission to cover 2,500,000 adults. It has a market share of 5%, and at their peak time, 43,000 listeners are tuned-in. Now, with FM (and DAB), it doesn’t matter how many people are tuned in: it makes no difference to the cost of the transmitter. But with internet streaming, you pay for enough capacity to cope (servers and bandwidth). Now, 43,000 is a LOT of concurrent listeners for internet streaming. Most stations won’t be doing anywhere near that - I know of a few running about 15,000 peak concurrent listeners, though most are in the hundreds. But, I spoke to an internet streaming company which specializes in radio streaming for large companies. Their rough costs are $55 per month for 100 concurrent streams - a maximum of 100 listeners at the same time. To turn off FM and have to stream a station to 43,000 listeners at peak would cost $23,650 a month: internet is four times more expensive than an FM transmitter. The station has 11m total listening hours a month, by the way, if you want to go and do some other rough costs. In bandwidth costs alone, one CDN I checked would charge $65,000 a month for the 343TB a month your audience would use. So the proper radio streamers I spoke to earlier have a good deal. There’s then the cost to the consumer. That station is listened-to for 32 hours a month by each listener, which is about 1GB of data per month. That isn’t entirely free. There’s the reliability of a data signal for live streaming, too; and streaming uses seven times more battery than listening to FM on the same device. And music rights. And so it goes on. In Bodø, the capital of Nordland, the county in Norway where they’re first to switch, a typical FM listener might get ten radio stations. When they turn on their DAB+ radio, they’ll find 30 stations there: a significant increase in choice. Perhaps that explains the global trends. Live internet streaming for radio is growing very slowly: far slower than DAB take-up in most countries, including Norway. So, even when people are being forced to change their radio sets, they don’t start using streaming: they stick to broadcast. And why wouldn’t they? It’s free, reliable, and it works. Internet broadcasting is great. It’s definitely part of radio’s future. But it isn’t - can’t be - a replacement for broadcast. Not yet, anyway.    Support the show.This podcast uses the following third-party services for analysis: OP3 - https://op3.dev/privacy