Digital video is made up of the raw video signal (video stream) and the way it is encoded into a file format (the codec). The file format itself (the container) also contains the encoded audio stream, which has its own codec. Compatibility issues with playback of video files can be due to a mismatch between the video and audio codecs being used, resulting in a video signal without sound or sound without a video signal. Other video encoding issues arising from incorrect settings can result in a highly distorted or unplayable video signal.
Before starting to learn about recording, editing, and saving digital video for long-term use, it can be useful to be familiar with digital audio and the processes of using codecs, discussed in the Audio section of this guide.
Video signal standards primarily relate to the number of images (or frames) created per second, the image ratio calculated by the length in pixels and number of horizontal lines (the definition) in an image scan, and whether the image is recorded using an interlaced or a progressive scan. A progressive scan is generally higher quality than an interlaced scan, which records two fields by separately scanning alternating horizontal lines of an image, and then interlacing them to make a whole frame.
Cinema-oriented digital video is recorded at 24 frames per second, at a picture aspect ratio of 2048 pixels x 1080 lines (2K) or 4096 pixels x 2160 lines (4K), using a progressive scan. Currently most recording devices at these resolutions are only affordable by professional cinematographers, and require high-end computer hardware and storage to process and edit the resulting video.
Television-oriented digital video is recorded at 25 or 50 (PAL) frames per second, or at 30 or 60 (NTSC) frames per second. Due to a combination of legacy and emerging technologies, multiple definition standards exist. For digital video these can be most simply grouped into standard definition (TV and DVD), and high definition (HDTV).
[insert video clips that demonstrate this difference]
Standard definition is now largely a legacy standard, but can still be found in cheaper digital camera devices that capture either an interlaced 576 (PAL) or 480 (NTSC) line scan at a picture aspect ratio of 4:3 or 16:9 anamorphic. These are television studio hardware standards set within a broadcasting industry standard called ITU-R BT.601-4. Unlike high definition video and graphics, the pixels used in these standards are not square (meaning their width is not the same as their height). This can cause problems (such as 'squishing' or stretching a picture) when converting an analogue video to digital, converting between PAL and NTSC video, or inserting a standard-definition video clip into a high-definition video. Further complications may occur if dealing with legacy digital video, especially that created or edited in the 1990s, as standards for a time were often wrongly applied or not followed.
High-definition video standards fix a number of problems that existed in standard definition, in particular by using square pixels. The industry standard, SMPTE 274M-1995, sets out three groupings of high-definition systems: 1080 progressive line scan (1080p), 1080 interlaced line scan (1080i), and 720 progressive line scan (720p). The groupings may use different frame or field rates for their scans (24, 25, 30, 50 or 60 frames/fields scanned per second). Knowing the settings used for the source video is important for editing and outputting high-definition video into different formats. While the settings are usually picked up correctly by editing software, if something goes wrong it is useful to know that not all sources may be correctly labelled.
Comparison of progressive and interlaced scans
Source: Wikimedia Commons
Encoding high-definition video always involves a trade-off between picture detail and file size. The bitstream of video data can be reduced to limit file size, but, as in all lossy reduction processes, the information that is removed can never be recovered. As a general rule, maintaining the highest bitstream version both preserves picture detail and your ability to edit and re-encode the file into new copies multiple times. This has significant storage and back-up implications for anyone interested in archiving high-definition video.
Almost all supported video codecs use lossy reduction to encode video signals, although some aim to be 'visually lossless' in that the lossy reduction is optimised to have no real visual impact on the image. One lossless codec, Motion JPEG 2000, while being an open standard and having future potential, at present places very high demands on hardware and can have limitations in audio support. In practice this means in 2010 all but the most state-of-the-art users need to choose lossy codecs to create, edit and archive their digital video with. Each video codec will also require a compatible audio codec for the soundtrack, which can be lossless (such as Linear PCM used by WAV, AIFF and CD audio) or lossy (such as AAC).
Lossy codecs commonly used for standard-definition encoding tend to be targeted at specific devices or recording formats, such as MPEG-2 (DVD players), MPEG-4 (Quicktime for computers, DivX and Xvid encoding for computers and DVD players, and FFmpeg for computers), WMV (Windows Media Video for computers), and DV (used in camera recorders). MPEG-2 and Quicktime are published ISO standards that can be subject to patents, while Xvid and FFmpeg are free and open-source standards. WMV consists of multiple proprietary standards and the open WMV 9 standard DV is a published industry standard, but has many proprietary modifications.
In high definition there are three popular lossy codecs in use:
Due to its efficient compression, H.264 is used in a large number of HD video cameras, and for a growing range of internet-based streaming and downloadable video. While some cameras remain in a manufacturer's proprietary encoding or older MPEG-2 encoding, most recognised brand-name consumer and semi-professional cameras now encode directly to H.264. All three codecs are approved to be part of the Blu-ray disc storage format.
[insert image or videos illustrating the difference between these formats]
It's important to know that the video encoding determines the choice of format (or container). This can be confusing, as most codecs allow you to save your video into different formats, meaning you cannot determine the compatibility of a video file with your software or hardware by looking at the file extension. For instance, the WMV 9 and DivX codecs can both save video into .AVI format, while .MOV and .MP4 can both be used to store MPEG-4 encoded video. Adobe's proprietary Flash format, .FLV, is also able to hold several different types of video, including the high-definition H.264.
HD digital video cameras may store their video stream in a variety of formats, some of which are proprietary to the manufacturer. One industry standard format that is becoming more widely used by camera manufacturers is AVCHD, developed by Sony and Panasonic.
Similar to CDs, DVDs and Blu-Ray discs have their own formats for video storage, known as VOB and BDAV respectively. DVDs and Blu-Ray discs need to be "authored" through a process or organising a number of audio, video, and data files in a standard sequence that can be read by players.
There are currently no satisfactory open video standards for video creation, encoding, and editing. However, we can provide the following guidance based on what we consider to be the best compromises and the general industry trends:
[insert video of someone choosing a camera, choosing their settings, editing and storing video, and talking through what they're doing and why]
Back to Creating Digital Content