Last time, I talked about the importance of handling all the connection details when you edit remotely. It’s up to you to decide if that was the easy part. The next step was to be able to stream my session to my clients who’d be in different locations with different connections and different levels of patience.
Luckily, before this whole environment developed, I experimented with different methods of streaming. I tried out hardware encoders and software encoders. I tried an in-house streaming server and streaming providers too.
During my research, I discovered that the terminology can be pretty confusing and, in my experience, it may be improperly used. People talk about encoders and streaming and seem to imply that an encoder is a streamer and a streamer encodes, but in practice, encoding and streaming are two different things.
To clarify a bit, encoders take a video output—or in some cases, your computer’s display—and convert it into a compressed stream with video and audio. Unfortunately, most of the encoders only output a stream that needs to be sent to a server, not to a group of viewers. The encoder and the stream it produces aren’t robust enough for multiple users to view. You still need a server to deliver the compressed stream to multiple people. That server is either hardware you must set up (in-house) or a service provided by a company. It’s that in-house streaming server—or the streaming services provider—that your encoder streams to.
There are many companies that provide streaming services, and I tested several. Some are primarily geared to streaming to a lot of people, so my need to stream to just four doesn’t fit their business model. It was soon obvious that pricing and support were issues with those large-scale streaming services. They want to stream concerts and other big events.
On the other hand, there were other streaming services that could deal with smaller events. During the research phase, I also made sure I dealt with a streaming service that could handle live events. Some streaming providers don’t. It’s easy to assume streaming means live streaming, but that’s not always the case.
With any type of live streaming, latency is an issue. Latency means how much delay is introduced by the encoder and the streaming server—the whole signal path—from my edit suite to someone’s shelter-at-home office.
Latency in remote editing equates to how much of a delay there is between when I start playing my sequence and when my clients start seeing it play. The goal is to have very low latency.
This is where I discovered that “low latency” means something different to each streaming service. Some considered less than a minute to be low latency. Others thought it was 20 to 25 seconds. And there were some that didn’t even talk about latency.
Next time, my expectations for latency.