What is Adaptive Bitrate Streaming and why is it so difficult to measure?
Consumers want to watch online video anywhere, anytime on any device
And they want the videos to be streamed without interruption or glitches in the highest resolution possible
That’s easier said than done
Video looks great when streamed at the highest resolution.
But, when bandwidth conditions suddenly
worsen — like when going from wi-fi
coverage to the 3G network — the
demands of high resolution video can
overload a viewer’s connection.
This results in buffering and other delays that negatively affect a consumer’s quality of experience (QoE)
To address this, content delivery networks (CDNs) and video providers are turning to
adaptive bitrate (ABR) streaming technologies.
Adaptive bitrate streaming detects a user's available bandwidth in real time and calibrates the video stream accordingly, always delivering the best possible picture quality.
ABR dynamically shifts
bandwidth to higher and lower
levels based on availability
While the concept is generally the same, there are several
different flavors of adaptive bitrate streaming technology.
HTTP Dynamic Streaming
HTTP Live Streaming
For Adaptive Bitrate Streaming to work, a video is first encoded at different bitrates to accommodate varying bandwidth connections. Each bitrate version is sliced up into tiny fragments — typically 2 to 10 seconds.
The video player then pulls fragments from the different encodings and inserts them into the stream as bandwidth dictates.
This results in faster video start times
and a continuous, uninterrupted
...which makes for happier viewers
and content providers
It also results in a major
measurement challenge, however
Not long ago, an entire movie could
be measured as a single file.
Now, with Adaptive Bitrate Streaming, measuring
that same movie in aggregate over a large viewing
audience can require analyzing tens of thousands
of separate files.
Then consider the sheer volume of ABR files coursing
over a network with 150 channels.
The number of fragment files over this network could be in the tens of millions for a single bitrate.
Add several more bitrates into the equation and the number of
Adaptive Bitrate Streaming files could jump into the hundreds of millions!
This level of fragmentation turns the
measurement data into a lot of unusable noise.
Skytide Insight for CDNs separates the
signal from the noise by identifying,
re-aggregating and reporting on
the actual content assets.
Skytide Insight for CDNs can handle the volume and
complexity that adaptive bitrate streaming demands
and transform it into meaningful reports in minutes.