Serving on a video project for a church many years ago, I was faced with a technical decision that was being made by a nontechnical person.
When considering how to deliver streaming video, several important technical decisions must be made.
That job? To produce a promotional piece for a community event.
In the process of editing, the creative pastor, who functioned as the primary stakeholder for that project, said to me, “When you’re done, please make sure this is in Flash format.”
While dating myself a bit, at the time, Flash was a buzzword in the video and web application world. We all fondly remember the litany of websites, full of effects and features, powered then by the Flash Player plugin. The issue I had then, is that the person mandating Flash, was doing it because it was a popular buzzword in the industry.
As the editor, I’d have probably delivered the video in H.264, using a standard MOV (Quicktime file) container. It irritated me that I was being “told” to use Flash, with no technical reason to justify it.
In my experience, this sort of thing happens all the time with technology, particularly when a nonexpert or nontechnical people influence a technical decision. In this scenario, you can count on a well-intended, but less-than-informed person, to force their nonexpert opinion about a technology, because it’s simply popular or commonly known.
Fortunately, the most popular technologies tend to work very well, but wouldn’t you rather be more informed in your technical decision?
When considering how to deliver streaming video, several important technical decisions must be made. Primarily, an ample supply of network bandwidth is needed to transport, and in some cases, scale multiple video streams to a group of users. Hardware including cameras, video switchers, encoders, media servers, and content delivery networks (CDN’s) come into play.
Even more important are how each one of the mentioned elements are configured to meet the demands of a use case. Streaming involves a unique choreography of technologies and services to arrive reliably, creatively, and within budget.
A lot of us technical folk tend to be drawn to equipment and services that are fun to use, look at, and talk about at parties. If you happen to attend a broadcasting trade show, companies like Blackmagic Design, Sony, Crestron and Samsung tend to have the largest, most sensational booths.
As a comparison, nonprofit technology firms, who exist to evangelize and represent a codec or protocol, will by comparison, have a very small presence or are listed as a feature as part of a more prominent firm.
Let’s face it, at a pool party, people would rather talk about their new Blackmagic ATEM unit than how they’re leveraging a particular technical configuration to optimize their video streaming workflow!
Ironically, the technologies used to deliver the content, most often produced by attractive, recognizable equipment, are critical, and can’t be assumed or overlooked as part of the overall solution.
Regardless of scale or investment, all streaming content gets encoded into some sort of stream which is transported in various forms, typically across the internet, where it is eventually processed and distributed for viewers.
In the streaming world, the process details of encoding, choosing the right codec and tuning the best protocol can literally make or break a streaming broadcast. This literally means that broadcast from an iPhone may have more chance of success than a poorly-configured broadcast from a $300,000 production suite.
Choosing the best codec, bit rate, and broadcast standard should be supported by people with experience, not relied upon brand influence or reputation.
What we’re really talking about is video engineering, which is a variety of settings that are established in a video encoder, tuned using a media server, and distributed across a content delivery network.
This is a thankless job.
It’s something can’t be more important in determining the reliability, resilience, and overall success of a live streaming workflow.
Not all content, networks, and use cases are the same, but there are a few major considerations that should be given proper - sometimes constant - attention, to be sure you’re set up for success.
The video encoder is where the most important settings are established. This device or software captures and compresses the video signal, applies a specified codec, and transmits the data to a destination (media server and/or CDN), using a specified protocol. In this process, the resolution and complexity of the video are defined resulting in what we commonly known as, but never limited to, 1080p or 720p video.
Deeper in the configuration are settings like frame rate, keyframe interval, and more. This is where you want to make an informed decision.
Mastering these settings generally requires a trained video engineer, with years of video experience.
The good news is that most video streaming engineers share their knowledge in all sorts of forums, across the internet.
Most churches will never staff a video engineer, but if you’re streaming, somebody is doing the work. My advice is to make sure your video engineer has performed proper research and testing, to be sure your streaming workflow is optimized.
If you’re using a media service, like Livestream, Vimeo, or Wowza Streaming Cloud, each service will provide recommended encoder settings that are proven to work well with their service.
As a general rule, for streaming media to large audiences, set an H.264 or x264 codec, transmitted using RTMP or SRT, using a bit rate between 2-6 Mbps, at 30 frames per second, with a two-second key frame interval for 1080p or 720p. This will enable effortless transcoding and packaging of media, for just about any player.
It’s also possible to skip the video engineering altogether and opt for an encoder appliance that will dynamically choose the best encoding settings, automatically. The ClearCaster, from Wowza Media Systems, does exactly that, even going so far to reduce bit rate and complexity when network conditions degrade.
In situations where you have control of how and where video is being consumed, more aggressive settings should be applied. For example, for a point-to-point broadcast, broadcasting to a distant campus or audience, an encoder-decoder combination can leverage a fast, high-bandwidth network (like fiber or multiple cellular bonding), where you can use a codec like HEVC/H.265 to carry high resolution video, using a protocol like SRT. Teradek provides excellent encode-decode technologies that do exactly this.
The best advice I can give is to research, test, collaborate, refine and repeat. Inversely, until you have perfected your workflow, never neglect your video encoder settings.
Take the time to experiment with different codecs, stream configurations, and network solutions. Don’t rest, until you have it just right.
Video engineering is a quiet, scientific role that enables content to stream to a global potential. Building expertise in this area will pay you back for many years to come.