A video sequence by itself is made up of a number of images, or frames per second in a specified resolution. For example, 1920 x 1080 at 25 fps – the standard HD television size. A codec is a mathematical algorithm that defines how the image information is stored in the the same digitized video sequence.
The original or master files above can use codecs such as:
- Apple ProRes
- Avid DNxHD
- IMX/ XDCAM / XDCAMHD ( mpeg2)
- AVC INTRA 50/100
- DV/DVC PRO
- ARRI RAW
These files are quite big, but it is not the file size we should focus on, but rather the bitrate – the amount of data per second. Remember that we are looking at a number of frames each second. Each of these frames has a data size – and after watching a second, we have 25 frames X each frames data size – this makes up the bitrate per second.
The bitrate is one of the challenges we need to address in a production system. When ten editors start playing back media sequences from the central storage, the storage needs read capability to deliver media simultaneously to all ten editors. Depending on your codec of choice, the same media can be defined with different bitrates. For these reasons, a production house might decide to start a media transcode process to convert the original media files from one codec to another codec with a lower bitrate or complexity. This conversion, also known as, a media transcoding process, will enable a faster and more efficient editing workflow. Also, if the material needs to be available on the cloud for web viewing, the media file's bitrate needs to correspond directly with the available bandwidth of the web viewer.
Let's take one of the codecs above, for example. The Apple ProRes HQ 1920 x 1080 has a native bitrate of 220 Mbit / second. This theoretically means that you need to have more than 220 Mbit available bandwidth to watch this media file in real-time. Otherwise, you will have to wait for a download. Not optimal, right? The bitrate per second is also one of the challenges we need to address on a media supply chain's distribution side.
Media transcoding for distribution
To be able to solve the issue described above, we need to invoke a new media transcode process. In the media transcoding process, we will now use a much more suitable codec for this distribution purpose.
There are several codec alternatives for distribution today. In common for these codecs is the use of a compression technique where pixels are “reused“ within the same picture (spatial compression) and over time between pictures (temporal compression). The result is compressed video for distribution at a maybe 1/ 20 or less of the bitrate and size of the original file – still visually looking the same.
Codecs for distribution:
H264 / AVC
H265 / HEVC
We will use the H264 codec – essentially doing media transcode from the original 220 Mbit Pro Rres HQ 1920 x 1080 file to an H264 file for web distribution.