This article is part 3.2 of the "21 Thoughts on Video Streaming in 2021"-series.
Don't forget radio!
Radio and audio is often overlooked in the streaming world. Although video streaming has migrated to safe stateless HTTP protocols, radio is still stuck with shoutcast - a protocol from the 90s (hello Nullsoft!).
With effort it is possible to reduce the amount of shoutcast listeners but some audio players still don't support HLS (such as the Google Nest speakers). On top of that, technologies for replacing or inserting targeted advertising have arrived, but the solution space for video is quite different than for audio. How for example do you carry SCTE markers in an AES67 stream? Is there a live encoder that I can use for both my video content as well as my audio content that supports a shoutcast compatible output format, ad replacement signalling, and descriptive metadata (such as song and artist name)? I don't think there is, but I am hopeful that 2021 will bring some much needed love for radio streaming!
Too often people use video streaming and media streaming interchangeably, but the same isn't true for audio streaming and media streaming.
The typical use-case for adaptive HTTP streams (like HLS and MPEG-DASH) is video, so audio is often neglected when designing solutions.
(Reminder: adaptive HTTP streams are streams where the stream quality adapts to your internet connection or device. Meaning: crappy internet = crappy visual quality.)
- When you go to a "modern" video streaming website, they'll "always" use HLS and MPEG-DASH.
- When you go to a "modern" audio streaming website, you might not see HLS or MPEG-DASH as often. Spotify doesn't use HLS/MPEG-DASH. iHeartRadio and SoundCloud uses HLS.
Many of the advertising integrations are also focussed on video ads, but not so much on audio ads. (This is changing, though!)
Audio is a fantastic medium, and I agree that it is overlooked in the "streaming world club". We focus too much on "frames", and not enough on "media".
I don't think we'll see more unity in 2021. I think audio-focussed companies will continue to focus on audio, and video-focussed companies on video – and that's a shame. Podcasts are huge. Clubhouse is all the hype. Everyone knows Spotify.
But what if Spotify wants to start delivering music videos? Will they design two streaming stacks, or design a single media streaming stack?
(If anyone ever wants to chat about their audio-related business idea or project: contact me. I'm very excited about this space.)