63 percent of streaming video viewers in 2017 cited buffering issues as the most serious problem they encounter. That number is an increase from the 2016 numbers. This points to a very serious issue with online video: providing a good quality video streaming experience to all your users. One of the best ways to deal with this is by using video transcoding.
Video transcoding is the process of creating multiple versions of the same video. These different versions have optimizations for different users. Ideally, this solution allows you to deliver high-quality videos to users with fast internet, and lower-resolution videos to those with slow internet. The result: no buffering.
This article will dive more deeply into video transcoding. We’ll discuss how transcoding works, how to maximize QoE for viewers, transcoding locally vs in the cloud, and more. We’ll finish the article by looking at how to deliver video in multiple bitrates to ensure that everyone gets the best possible experience.
What is video transcoding?
Video files are very large. They contain a lot of data, which means they use up a lot of memory on devices and bandwidth during transmission over the internet.
To address the large file sizes of video, we compress video files using technology called “codecs.” The most common video codec today is called H.264. There are many other codecs as well. They all basically work in the same way: namely, they remove extraneous data from the video feed to reduce file size while maintaining as much quality as possible.
Codecs use complex algorithms to shrink video files. As computers and digital processing devices have become more powerful, more effective codecs have allowed us to further shrink down video files. The next-generation of video codecs like AV1 and H.265 are now in use on a limited scale.
Technically, video transcoding refers to the process of taking an existing video file (or ongoing stream) and re-encoding it using a different codec or different settings. Usually, when a video recording is made, the resulting file is called a “master” file.
This file is the best quality recording that exists. Any subsequent versions are usually encoded down—that is, they are transcoded into a lower resolution and/or bitrate. Up-encoding (to a higher bitrate) is impossible, although limited up-rezzing (increasing resolution) is possible—with a loss of quality.
Codecs, bitrate, resolution, and file size
As I’ve stated, usually video transcoding helps to provide smaller versions of the original file. For example, let’s say your master file is encoded using the following settings:
- Codec: H.264
- Resolution: 1920 x 1080 (full HD)
- Bitrate: 6 Mbps
- Audio: AAC, 44.1 KHz, 164 kbps
From this original file/stream, we would usually recommend creating a few different lower-bitrate, lower-resolution versions. These can provide a better viewing experience by allowing people with slow internet to watch your stream.
- 720p version at 4 Mbps
- 480p version at 1.5 Mbps
- 360p version at 1000 kbps
Maximizing QoE (Quality of Experience) for viewers
When they watch a poor quality stream, viewers are upset before they even realize that they’re upset. Their happiness level falls. After an experience like this, they’re much more likely to leave, and to view your brand negatively. One survey showed that 62 percent of viewers view a brand negatively if they publish a poor quality video.
This points to the importance of quality. Transcoding, once again, allows us to address this issue.
In detail, that’s because the internet speed of a given viewer must be able to accommodate the bitrate of the video stream they are trying to watch.
For example, my current internet speed is roughly 16 Mbps (note: that’s megabits, not the more common megabytes. Divide by 8 for a more common measure; 16 Mbps is equal to roughly 2 Megabytes per second). With a connection speed like that, I can watch 1080p HD streams. However, if I leave the house and watch on my cell phone, I may need to receive a smaller video—especially if reception is poor.
If there smaller, transcoded versions of the stream are not available, I will experience buffering, lagging, slow startup, or even failure to play. These problems are collectively known as QoE (Quality of Experience) problems. Maximizing QoE is essential for any video broadcaster. Transcoding can help.
Transcoding locally vs. in the cloud
Generally speaking, there are two ways to transcode your video: locally, or in the cloud. There are benefits and drawbacks to each approach. Generally, local transcoding is slower and results in multiple files that you must upload and manage separately. However, it’s free. Cloud transcoding may have specific pricing, or limited transcoding may be included with an online video platform, but it’s fast and convenient.
For example, local transcoding usually involves using a video editing or compression software (like Avid, Final Cut X, Compressor, or Adobe Premier) to transcode a video file on your computer.
If you’re using a cloud video hosting solution, such as DaCast, you upload your “master file” to the DaCast platform. Then, you can click through to the “Multi-bitrate” options for that file. This screen will allow you to select various resolutions and transcode your file into those versions.
You can also select “auto-encoding” when uploading a video, which automatically transcodes your VOD into several different renditions.
What formats does DaCast use for video transcoding?
DaCast uses the most common file formats for video transcoding. The settings used are:
- Video codec: H.264
- Audio codec: AAC
- Video file format: MP4
- Resolutions supported: 1080p, 720p, 576p, 480p, 240p
Any video files that you transcode manually, or that auto-encoding automatically transcodes, will use these settings.
Using an adaptive video player to deliver in multiple bitrates
Once you have completed your video transcoding into multiple bitrates, you need a way to deliver it to your viewers. This delivery happens via a technology called “adaptive video players.”
An adaptive video player will automatically detect the internet speed of a given viewer, and attempt to deliver to them the corresponding quality video from the available options. These video players (most of them nowadays are native html5 player) also allow the viewer to manually select the quality they would like to receive using a menu in the player window.
Given that online videos with a start-up time exceeding even two seconds have significantly higher streaming video abandonment rates, using an adaptive video player with transcoded videos is essential. In fact, each incremental second propels another 6% of viewers to tune out.
Live streaming vs. video-on-demand in multiple bitrates
Video transcoding is easy for video-on-demand files. For live streaming the process is a little more complex. However, the same effect can be created via multi-bitrate live streaming.
This is done by setting up multiple different outputs in your live stream encoding software or hardware. Each output should correspond to a given bitrate/resolution that you wish to live stream. The encoder will create versions of the stream corresponding to each setting, then send all these versions to your online video platform simultaneously.
The only drawback of this method is that it requires more processing power and more upload bandwidth. Nevertheless, we recommend that everyone do multi-bitrate live streaming.
Hopefully, this article has helped introduce you to the concepts around video transcoding for online video and live streaming. As these topics determine QoE, this is an important subject for any broadcaster to understand.
Do you already have a video streaming solution? If not, you should check out DaCast. It’s a powerful, versatile live streaming and video-on-demand hosting platform with tools like analytics, video monetization, security and privacy settings, and more!
Why not take advantage of our free 30-day trial (no credit card required)? You can try all the DaCast features yourself to see if our platform is a good fit for you!
For exclusive offers and regular live streaming tips, join our LinkedIn group. Do you have questions or feedback on this article? We’d love to hear from you in the comments below!