Follow

FAQ: What software should I use to compress/convert video?

Of all the resource types that are encountered in the online learning environment, video files tend to be the largest in terms of file-size. This is because a typical video stream can contain a large amount of data, and the higher the quality of the video—the larger the file-size.

Ultranet's Media Gallery and associated modules all have an upload limit of 30 MB per file. This is suitable for web-optimised video clips, but often the videos you'll get from digital cameras/camcorders or similar devices will cater for higher-definition targets (e.g. Full-HD TVs and monitors).

So unless you can specify on the device to record at "web quality", a post-production encoding will be required to squeeze the video down to a workable size/format.

Suitable software

Here are some software tools that handle video compression well, and are usually ready-available on school computers or freely downloadable.

  • Windows Movie Maker (Windows-only; free)

    This is one of the most commonly used video editing solutions on Windows machines, and has the advantage of offering full editing abilities as well as export/compression. The editing functions are helpful if needing to cull or trim down a video clip to just the bits you need, thereby reducing the overall file-size. It also includes handy export presets specifically targeted at 'web' or 'email', both of which are ideal for uploading into Ultranet.

     

  • iMovie (Mac-only; included on all new Macs, or in the iLife software bundle)

    This is another popular video editing solution exclusive to the Mac platform, and again offers some deep editing abilities on top of the export/compression function. iMovie is renown for its easy user interface and inclusion of fun editing features to spice up any video production. So if you use a Mac and would like to tweak your video production before compression, this is certainly the best 'first stop'.

     

  • Handbrake (Mac + Windows; free)

    This is an open-source solution that was primarily focused on ripping DVDs, but has also grown a solid reputation for its video conversion abilities. The advantage of Handbrake is its a lot more lightweight (being solely focused on just the extraction/conversion process, rather than offering full editing abilities)—so if you just want to convert videos without performing edits, then this is probably the better solution.

    The only disclaimer: Handbrake does have a potentially daunting interface, which might put off some novice users. However, this is largely due to the extensive level of control it offers users on how their videos are converted—in some cases to an even finer degree than the first two options mentioned above. If you're keen for a challenge (and comfortable reading up on their online guides for help), then Handbrake is sure to offer the power and flexibility you need.

     

  • Miro Video Converter (Mac + Windows; free)

    This is great-looking and simple drag-and-drop style video converter that relies on a 'preset' approach to conversion. This may be helpful when dealing with uncommon video file-types or wanting a dead simple approach to video conversion. However, it does come at the expense of deeper customisation—so although you'll have a more compatible video output, you may not always get a 'smaller' video file-size. Perhaps best for videos that are already under the 30 MB mark and just need to be converted to the right format.

     

  • Adapter (Mac + Windows; free)

    This is another open-source solution that has been cobbled together from various other projects to provide a 'swiss army knife' of sorts to the file-converter family. The user interface is fairly straight-forward, and has the added advantage of being able to also handle audio and image file conversion too (theoretically a perfect fit for the image/video/audio support in Ultranet!). It does offer a bit more advanced tweaking options for your file-conversions, so it may be worth checking out for the middle-ground user.

     

  • Evom (Mac-only; free)

    This is super-simple and basic converter (similar to Miro), but being 100% focused on the Mac platform only. Just like Miro, it uses a drag-and-drop approach to file conversion, and has a neat array of presets to choose from. Once again, you may find the lack of advanced tweaking options a limitation. But it may be the perfect fit if you have a large supply of FLV files downloaded from YouTube or similar and just want to get them converted to MP4 so you can view/republish them easily.

General tips on compressing video

There are no hard and fast rules on creating the perfect web-optimised video, as you will always be balancing quality vs usability (file-format, file-size, universality, etc). However, there are some general concepts surrounding web videos that are good to understand so that you'll know what to tweak to get the best results (if your chosen software offers the tweaking options!).

  • File-format. 

    All of the software solutions covered earlier will produce video formats that Ultranet understands and supports. In most cases, a video 'format' consists of the container (i.e. the package type, such as "AVI" or "MP4") and the codec (i.e. the encoding algorithm used to code/decode the video, such as "H.264"). Usually, these are a paired set, so if you just stick to the most commonly used containers for web video (MP4, MOV, MPG) then you should be fine.

    The formats that Ultranet supports in its Media Galleries are AVI, MOV, WMV, MP4, MPG, MPE, DIVX, 3GP, RM, MKV, QT and OGV.

     

  • Video resolution. 

    The overall dimensions of the video are what is known as its resolution. This is usually expressed in a paired set of numbers, respectively denoting the horizontal and vertical size in pixels (e.g. 320 x 240, 640 x 480, etc). 

    So a 320 x 240 video contains 76,800 pixels in each frame, whilst a 640 x 480 video contains 307,200 pixels per frame. As a general principle: the larger the resolution, the more pixels there are per frame of video and data the file contains—hence the larger the file size. 

    The following illustration demonstrates the exact same still frame, but at varying resolutions from smallest to largest. The largest certainly has the most detail, but it is also the largest in terms of file-size.

    Same image, different resolutions

    Video resolutions are not just important in terms of its impact on file-size, but also on the eventual output size. If you think about a typical computer screen resolution (e.g. 1024 x 768 or 1280 x 800), they also have a limit as to how many pixels are available to display visual data. If the video resolution is higher than the resolution of the device viewing the video, then scaling has to be performed to resize the pixel data into a lower ratio. But this is inefficient, because it means the video is not going to be viewed at the full resolution anyway—so why send all that extra data, only to have it be discarded?

    Full-HD videos are 1920 x 1080. This is larger than most typical computer monitors and displays, and definitely much too large for a typical web page. So if your camera records in HD (1080p, 1080i, 720p, 720i), then be aware that scaling will probably have to be performed and you're going to have massively sized output files.

    Generally speaking, the ideal resolution you should aim for with web-optimised videos in Ultranet is around the 320 x 240 or 640 x 480 mark. Make adjustments here to suit the type of footage being uploaded—if detail is important, notch it up higher; if the subject is just a talking head, and fidelity is not so important, then notch it lower. 

    Remember that your audience can also choose to view the video in fullscreen mode, so it doesn't mean that a tiny resolution video will have to be watched on a tiny screen area. It just may look a bit pixellated or boxy when maximised (see notes below on bitrate).

    Also, please note that these figures are in 4:3 aspect ratio (i.e. traditional "TV" shaped), so adjust those numbers accordingly to suit footage that is in 16:9 aspect (i.e. widescreen). Anything higher than that is probably going to be too large and too wasteful for its end purpose.

     

  • Bitrate. 

    The bitrate of a video refers to the density of data per second of video (or audio), and has the most impact on the overall quality of the image—regardless of its resolution. It is the factor that determines whether a video is perceived as 'crisp' or 'detailed' in its imagery, with low-bitrate video essentially appearing more pixellated and blocky.

    The following example juxtaposes two still images of the exact same pixel resolution, but with differing bitrates:

    Comparing bitrates.

    The image on the left has been resized from the original, but with a high bitrate, whilst the image to the right has been encoded at a lower bitrate. Whilst the subject of the image (the cat) is still discernible and not necessarily 'unacceptable' in quality for a viewing audience, the loss of detail in image fidelity is noticeable when compared to the higher bitrate image to the left.

    Although lowering the bitrate does save on file-size, there will definitely be a point where the outcome is beyond a state for 'comfortable viewing' (i.e. the video becomes unwatchable due to extremely poor image reproduction). 

    So experiment to see what level you can take it to that gets the footage to the right balance in size and quality. If the output is consistently of very low quality at the intended size, consider splitting the clip into two clips of higher bitrate each.

     

  • Framerate. 

    You may have heard the old term of 'motion pictures' being used in reference to movies. In essence, this does accurately capture what is happening when we see a video clip, in that it consists of a series of still images strung together and 'projected' at a rapid pace so as to simulate motion.

    For instance, if we look at the following sample of Eadweard Muybridge's work in capturing animal locomotion via still imagery, you will see that each step in the cat's 'motion' is constructed from individual frames.

    The still frames from Muybridge's cat motion picture

    These single frames are then strung together at a defined size, and you get a 'motion picture'. What determines how lifelike the motion appears is the framerate, which basically means how many single frames are presented within a single second of time. If you densely string together enough still frames at a high framerate, then the human eye basically can no longer distinguish or notice the individual frames being presented.

    However, as with the variables discussed above, the more frames you have per second, the more data has to be stored in the video, and hence the larger the file-size.

    You can usually drop video framerates to around 24-30 fps (frames-per-second) without any jarring sense of false motion. 24 fps is a common old standard for traditional film movies, whereas digital high-definition movies capture at a much higher rate.

    You can drop the rate lower to save more space, but again, just be mindful of the subject matter and whether the end result produces an 'unwatchable' video clip.

    The Muybridge cat, animated at a low framerate

    Remember, Muybridge's motion pictures were rudimentary and at the very birth of video capture, so their likely file-size would be minuscule in comparison to a modern video of a cat running, captured with an HD-capable camera. Modern video cameras automatically use a much higher capture rate to produce footage with high enough framerates to not even cause the audience to think twice, but their file-sizes are also not so conservative. So unless you're using vintage era technology to capture footage like Muybridge, do consider tweaking your framerates to produce smaller video file-sizes!

    Begging

Was this article helpful?
1 out of 1 found this helpful
Have more questions? Submit a request

0 Comments

Article is closed for comments.
Powered by Zendesk