We get a lot of questions about bit rate from clients because it’s crucial for video translation, but is also a little tricky to understand – you’ll see exactly why in our post. Bit rate affects everything in video production from quality and file size, to compression time. While localization professionals don’t need to be bit rate experts, a little knowledge is a must to reign in production budgets and timelines.
In today’s post we’ll look at what bit rate is, how it affects video output, and what you need to know about it for localization.
[Average read time: 5 minutes]
What is bit rate?
Bit rate measures how many bits – the basic units of information in any digital media – are in a video stream. You already count bits every time you measure a file size – for example, a file that is 2 MB (megabytes) is composed of approximately 11 million bits. While file size measures the total number of bits, bit rate measures the number of bits per second of video.
For example, video might have a bit rate of 1 megabit-per-second, or 1 Mbps – that means that each second of that video contains 1,000,000 bits. Likewise, the bit rate could be 1 kilobit per second, or 1 kbps – that’s 1,000 bits for every second of video.
Why is bit rate important for video production?
Bit rate directly affects the quality of a video. Higher bit rates produce sharper videos, while lower ones produce blurrier videos, because the bit rate controls how much information – how much data – is in each video. The following e-Learning scenario clip shows the same scene compressed at different bit rates. Note the difference in quality (it really helps if you view it full-screen):
It’s even easier to see in the following split-screen comparison – note that in the higher bit rate, you can see the texture of the hair and suit, and make out each fingernail clearly:
It would seem, then, that higher bit rates are always preferable, right? Not quite.
Bit rate also affects the final video file size. In the example above, the video with the higher bit rate has a size of 2.35 MB, while the one with the lower bit rate is only 893 KB. In effect, the larger video is three times the size of the smaller one – which lines up with the differences in bit rate as well.
Video image quality vs. file size
Video producers have to contend with this conflict all the time. On the one hand, they want to deliver high-quality videos that look great. On the other hand, each bit of video costs money.
For example, an e-Learning developer whose clients stream videos from its server pays for every single bit of video that gets streamed. That developer has a very strong financial incentive to lower its videos’ bit rate as much as possible. Likewise, its customers want videos that can stream without interruption over internet connections – which means that they too want the bit rate that will give them good quality and uninterrupted viewing.
Another good example of this is YouTube – with millions of videos on their site, they have a financial incentive to keep their server usage low. They also want to make sure that videos stream well for most users, who will have different internet bandwidths. When you upload a video to YouTube, the site automatically compresses it to bring down the bit rate, so that videos can play uninterrupted.
Finally, media like DVDs and Blu-rays have strict data limits (4.7 GB and 25 GB respectively), and their players have bit rate limitations – thus, videos for these media have to be compressed in very specific ways. Blu-rays are higher quality specifically because they can take larger file sizes and higher bit rates.
So why does it matter for video translation?
Clients spend a lot of development time figuring out the compression “sweet spot” that provides them with the best quality and lowest costs possible. If their localized videos are larger, the client loses money; if they’re smaller (and don’t look as good), the client may lose customers. In fact, bit rate settings sometimes are hard-coded into websites, learning management systems, or e-Learning course authoring platforms, so that videos that don’t conform will be rejected.
Therefore, it’s crucial to match client bit rate specifications. This sounds relatively easy – one should be able to just match the bit rate of the videos, right? Well, sort of. There are three complicating factors:
- Video bit rate can be variable – that is to say, a video encoding program will compress sequences that are simple (for example, just titles over black) at lower rates than complex ones (for example, with a lot of movement), all in an effort to keep file sizes low while maximizing quality. This means that getting an accurate bit rate from the final compressed videos can sometimes be tricky – and not nearly as accurate as getting it from the client’s specifications.
- Localization work is done in source files that are uncompressed, or compressed at higher bit rates – that is to say, video files that don’t match the final delivery specs. This is especially true of dubbing and subtitles, since it’s a best practice to lay subtitles back to the highest image quality available, and mix lip-sync audio to the highest-quality audio – and then compress. Unclear or incomplete specs can cause issues with deliverables integration in these cases.
- Localized videos require different bit rates. This isn’t as common, but sometimes locales have different bandwidth requirements – for example, if a video is going to a country with intermittent service, the client may want to compress at a lower bit rate. In this case, it’s good to do a few compression tests with the client to find the new “sweet spot” bit rate.
Compression – the final step of localization, in which the deliverable videos are output in the editing studio – is intensely time-consuming. For example, compressing a 10-minute After Effects green-screen video (common for e-Learning scenarios) can take as much as 30 minutes. Imagine doing that in 20 languages – you’re looking at 10 hours of machine time. Because of this, you want to get compression right the first time, to avoid costly re-compression costs. Or, even worse, if the videos make it to the client, they’ll cost more in server and bandwidth fees. As with all localization, the only way to avoid budget and timeline overruns is to get the work done right the first time.