Getting to Know Video
Compression and Evaluation of Video Formats, CODECS, and Containers
If a picture is worth 1,000 words, then a video is worth 1 million. It’s no secret that videos are incredibly popular on the internet, and not just the YouTube kind. Videos are used for entertainment but also for product reviews, case studies/testimonials, how-to/tutorials, live talks/presentations, Q&A, announcements, etc. As a designer, you’ll probably be incorporating videos into a website, either for yourself or a client, at some point. It’s important to know how the types of video files and formats as well as what compression does to them so you optimize video for your site. I spent some time researching and testing different files so you don’t have to, this article contains my results.
File Formats
The first thing you need to know about video files is that they all contain two parts: the CODEC and the container. The word CODEC stands for “compression and decompression”, so it’s shouldn’t come as a surprise that the CODEC’s job is to compress and decompress data contained in a video file. Videos create some of the biggest files on the internet so compression is especially necessary for this media. The second part of a video file is the container. There are 9 containers that are considered the most common. These include:
. MP4, .M4P, .M4V
A digital multimedia container originally created in 2001, then relaunched in 2003. It’s most commonly used for downloading and streaming videos from the internet. It also stores audio, subtitles, and still images. It is considered the international standard for audio-visual coding. With the high level of compression used with the MP4 format, it allows the file sizes to be much smaller than other video formats without losing too much of the quality. Almost all video players can support MP4 and Windows and Mac users can open this file without having to use any third party software.
. AVI
AVI is a commonly used multimedia container format that was developed in 1992. It is based on the Resource Interchange File Format (RIFF) released in 1991. It stores audio and video using a variety of codecs. The file format separates all of the information into blocks that can be tagged as AVI. AVI uses less compression (sometimes none) compared to other formats, making it larger than most files sometimes as big as 2–3 GB per minute of video. It allows for playback without the use of any codecs.
. WMV
WMV offers the smallest available video file size. It was originally designed for internet streaming applications and is very nice for sharing through message or email. It is lossy and therefore offers the lowest quality video. It is compressed with Windows Media Compression. Its quality is maintained best with Windows video format, if you are not using windows you can use converters.
. MOV/QT
MOV is an MPEG4 video container format created by Apple in 1998. It is a format that is designed with embedded with cross compatibility support, so it can be viewed with both Mac and Windows but will open with different programs for each. For Windows, MOV will open with programs such as Quicktime Player, Roxio Creator, PowerDirector, PowerDVD, and Adobe. For Mac it will open with programs such as Quicktime Player, Roxio Toast, Roxio Popcorn, and Adobe Flash. It can contain multiple tracks that store different kinds of media data.
. FLV/SWF
This was a very popular format used by nearly all embedded video on the internet, it was used to deliver content using Adobe Flash Player. Since this file format does not support iOS devices it is quickly becoming obsolete and dropped in favor of HTML 5. The data within FLV files are encoded like SWF files.
.WebM
WEBM is a video file format that was initially released in 2010 as a royalty free alternative to be used in conjunction with HTML 5. It is primarily used for videos on websites as its file sizes are small enough to offer fast loading times. It is a lossy format, so it will lose some quality with compression.
. AVCHD
AVCHD was designed for high-definition digital video cameras to record in HD with high-efficiency compression coding technology. AVCHD is a container format that is very popular among videographers due its high quality. It also supports 3D video. It allows large files to be captured and stored on digital media such as hard disk drives and SD memory cards. It records at resolutions including 1080i, 1080p, 720p. It uses 8cm DVD media as its recording medium and is also blu-ray compatible. AVCHD uses MPEG-4 AVC/H.264 format for highly compressed HD recording. This format also includes some presentation features such as subtitles and menu navigation.
Bitrate, Frame Rate, Field Order
Videos have other components as well, these include bitrate, frames, and field order. The bitrate is the amount of bits per second encoded in your video. Basically the higher the bitrate, the higher the quality of your video. But since there is more data in the video it also means the file size will be larger.
The frames of your video are individual pictures that play as a sequence of images. The FPS, frames per second, measure the amount frames that are played in one second of your video. So the more frames per second, the higher quality of the video. But of course this also means more data is stored in the video so the file size will be larger.
Each frame in a video is spilt into two fields. Each field contains lines of horizontal resolution running across the screen. With interlaced videos each half of the video contains the total lines of resolution while progressive-scanned video format lines are drawn one at a time, from the top of the screen to the bottom.
Experiment
I wanted to learn more about how changing, compressing, and exporting the frame rate affected a video. So I conducted a little experiment. I took four different video clips that possessed different qualities and tested them in the file formats and presets shown below:
I used both Adobe Premier Pro and Adobe Media Encoder to manipulate the frame rate and compress the files. I tested the first three video with different frame rates, I used 10, 15, and 50 FPS. Then for the last video I took the best presets from each frame rate to compare against each other. From each of the 72 video clips, I decided to include only the 6 files I felt had I highest contrast and quality. These are the video clips I used:
10 Frames Per Second
The first frame rate I tested was 10FPS. The first clip (shown below) is a 480ppi standard definition H.2.65. As you can see the image quality is blurry and we can very distinctly see aliasing happening by noticing the individual pixels in the picture. The playback speed was choppy and overall this file was pretty painful to watch. The second clip had a higher pixel count at 1080ppi. This clip was high definition in a H.265 container. The playback speed was just as choppy, but the image quality was significantly better than the 480ppi clip. The third clip was a file format deigned for mobile phones with 2160ppi. As you can see the picture quality doesn’t differ too much from the previous clip, which I found interesting, but the playback speed was just as choppy. The fourth clip resulted in a less pixelation in the image quality, though it’s difficult to tell unless you really zoom into the picture. The fifth and sixth clips were the two biggest file sizes and as you can see really didn’t differ much in image quality but they were by the far the most delayed and laggy in playback speed.
15 Frames Per Second
I tested the second video at 15FPS. Since this is a video that is very saturated with light I was curious to see what the different settings would do to the clips. Clip four had the lowest quality and file size, the picture was washed out quite a bit and had low. Clip one, H.265, it is less washed out though still fairly pixelated. These files had the fastest playback speed, I imagine this is because the lower resolution removes pixels, allowing for the video to have to display less. The videos get more delayed as the sizes increase. While video six is the highest picture quality, it has the laggiest play back speed and doesn’t differ too much from lower resolution videos.
50 Frames Per Second
For the third video, I tested each file with 50FPS. The playback speed was normal for clips one and two, and very slow for for clips five and six. The nature of this video has bright moving light, which turns to fuzzy balls of light no matter the resolution. When you zoom in however, you can see aliasing starting to be much more obvious in the low resolution videos. For a video like this, it may be worth using a lower resolution video for a smaller file size, since the picture quality isn’t much different and it will provide you with a smoother playback speed.
10 vs 15 vs 50 Frames Per Second
For the last video I tested various FPS. Clips one and two are 10FPS in both H.264 and H.265 file formats. The second video clip has much better picture quality and both were equally as delayed in playback speed. Clips three and four were tested at 15FPS, and while clip four has the same quality as clip one it’s playback speed was much smoother. Clips five and six were tested at 50FPS and had the crispest picture quality but the worst lag. There were only what seemed like three frames displayed throughout the clips, this could be due to poor read speed from my hard disk drive since the file sizes are so massive.
What Did I Learn?
From this experiment I learned a couple of things. I learned what frames per second means, which is the speed at which images are shown. So if I use a 10FPS setting, I will be getting 10 distinct still images each second. It makes sense then that this provides a choppier playback because there is less information than that of a 50fps file. I found that the reason why the most common FPS is 24 is because it sits in the FPS sweet spot, high enough in quality and low enough in file size. I found 10FPS to be much too low, resulting in missing info, or jumping scenes. It makes the video choppy and hard to watch. On the other hand though, bigger is not always better. 50FPS is so much information per second that the result is a delayed or lagging effect. The files I tested that had the most normal play back speed was that of 15FPS which is just enough info to play things smoothly.
Which FPS should you use?
Just like different strokes are for different folks, different rates yield different results. Even though 24FPS is the most common FPS, that doesn’t necessarily make it the best. If you want to know which frame rate you should use the answer is: it depends.
If you have software that can support it, a higher frame rate will give you a crisper image quality and smoother playback. There is some debate around using the highest frame rate because some feel that it’s not realistic. Naturally our eyes view moving objects with a certain amount of blur, having no blur can make the video feel dramatic which may not be an effect you want in your video. 60FPS is used for things like sports footage when we want to view things at super fast speeds while TV productions are usually shot at 30FPS. If you’re sending a video to your grandma who can’t see anything anyway, you may want to send the lowest frame rate for the smallest file size. It’s also important to consider what device you’re recording for because certain presets for certain devices work well with certain frame rates.
In Conclusion
Finding the best frame rate for your video will often come down to sacrificing detail and playback speed or low files sizes because videos file can get huge. The best way to find out what file type, CODEC, or frame is best for your video is by testing it different ways. Though it can be tricky to find the optimized frame rate for your video, it’s worth figuring out because frame rate greatly impacts viewing experience and feeling of a video.
Jessica Ames is a student in the Digital Media program at Utah Valley University, Orem Utah, studying Interaction & Design. The following article relates to (Video Compression and Evaluation) in the (DGM 2341 Course) and representative of the skills learned.
