Do Codecs Really Matter?
If you’ve followed recent news in the streaming industry, you’ve probably heard about the latest skirmish in the codec wars. For those of us who have been in the digital media world for awhile, it may seem like codec decisions come about as frequently as presidential elections—and with almost as much polarization.
Whether it was Real versus Microsoft; MPEG-1 versus MPEG-2; Apple versus Microsoft; or even On2 VP8 versus H.264 (a.k.a. Advanced Video Coding, or MPEG-4 Part 10), content producers have often been forced to pick a side—or play the field, by choosing both—when it came to reaching the biggest possible audience.
In early 2018, the choice seems to be between maintaining H.264 or moving to VP9. Then there’s the added decision point of whether to adopt H.265 (High Efficiency Video Coding, or HEVC) or wait for the upcoming AV1 codec from the Alliance for Open Media (AOM).
AV1 appeared poised to win this battle—at least, until a few weeks ago. Currently in a code freeze, it appears AV1’s tight correlation to VP9 use cases will offer a number of advantages in a royalty-free package when it’s released later this year. This stands in sharp contrast to HEVC, which has long suffered from complicated licensing structures at the hands of several HEVC licensing pools, while royalty-free codecs such as VP8 and VP9 have grown, unencumbered by patents.
However, one of the HEVC licensing pools—HEVC Advance—recently announced an effort to reduce or eliminate royalties for patent holders. This is welcome news indeed to broadcasters in over-the-top (OTT) streaming; over the air (OTA) local television; cable; and even satellite distribution, though well-informed industry observers are still scratching their heads as to why it took so long to get to a reasonable royalty landscape.
Even with the licensing issue becoming less of a decision point, the wide range of codec choices is enough to make content producers and publishers throw up their hands in disgust and pray for a single, universal codec—one without patent royalties or playback limitations on particular devices. Ideally, it would also be one that already has hardware support and can handle compressing videos beautifully for any use case: from low-bandwidth, cellular-only encoding to 4K Ultra HD (UHD) living-room delivery.
Here’s the good news in all this: The codec you choose doesn’t really matter.
Why? Here’s the quick answer: Most media servers will handle the conversion between codecs (known as transcoding) just fine. The focus should be on getting the highest-quality stream ingested into the media server and allowing it to do what it does best.
A few media servers will do even more, repackaging content from a classic RTMP stream into HTTP segment-based delivery options, such as Apple HLS and the MPEG-DASH solution. Both of these offer adaptive bitrate (ABR) options, so end-user devices can choose which one of several data rates or screen resolutions to watch at any given moment, based on current network conditions. For example, Wowza Streaming Engine™ media server software offers all these features.
There are a few more tricks up the media server’s sleeve: Some offer network-based digital video recorder (DVR) capability, allowing content to be watched later via the inherent time-shifting nature of DVRs, without clogging up a user’s mobile phone or laptop storage. Network DVRs have the added benefit of allowing content to be watched just a few minutes later—a feature I’ve put to use with clients who need to stagger viewing start times for corporate all-hands meetings, or even for multi-campus houses of worship.
One final point on media servers: Not all servers offer equivalent functionality between their cloud solutions or their on-premises (installable software) solutions. Be sure to check the codec options for both ingest and output to make sure your desired workflow is supported.
Then get out there and start producing content—and rest assured that you won’t have to know anything more about codecs, except maybe for their acronym, at your next tech discussion.
Additional Resources