Video compression makes everything possible. Not everything, I suppose, but pretty much everything I blog about, anyway. Video compression is the technology that makes it possible for you to get a whole movie on a disc, multiple movies on your phone, hundreds of channels on your satellite dish, thousands of different videos through your streaming box, and millions of cat videos on the internet. It’s all made possible due to compression.
What does compression do?
Compression takes a large signal and makes it smaller. At its simplest, it’s sort of like those vacuum bags where you suck all the air out and all of a sudden a pillow only takes up the same space as a sheet. It gets a lot more complex though.
Good compression does more than just suck the empty space out of a signal. It also tries to make intelligent decisions about how much of the signal can actually be thrown away without you noticing.
Through complex math, compression hardware will eliminate parts of the picture that are too dark for you to see. It will move numbers around just slightly so that the whole signal can be compressed even more. It will even mess with the colors just slightly to see if that makes the whole thing more “compressible.”
What problems can happen with compression?
Compressing a picture comes with risks. You can use the 4:2:2 color space instead of 4:4:4 and most of the time no one can tell the difference. You can mess around with numbers and even with very simple hardware get a 97% drop in size. However, you need to be careful. If you overcompress you’ll run into mosquito noise, bitstarving, and macroblocking. If you take away too much of the shadow information using gamut compression, the whole picture will look grey and lifeless.
The best way to avoid these problems is to use one of two technologies.
Improving the codec
“Codec” stands for “compressor/decompressor.” It is the software that actually does the hard work. These codecs are designed around standards so that a signal that is compressed at one place can be decompressed at another. There have been several generations of compression, each better than the last.
ZIP Compression
ZIP compression is still used to make files smaller but it’s nowhere near smart enough to be used for video. On the upside though it’s completely lossless, meaning that you can compress a file and it can be uncompressed with complete accuracy. This is the compression to use if you’re sending bank records, but it’s not so helpful for video.
JPEG Compression
JPEG was one of the earliest forms of “lossy” compression, which tries to make tiny changes to a digital file to make it smaller. It works fairly well for images but isn’t really good for video because it’s not very efficient.
MPEG-2 Compression
MPEG compression takes a lot of the same logic used in JPEG but also adds the idea of a “key frame.” The codec looks at the different frames as time goes by and determines when a camera shot changes or when something big happens. Then all it does is provide the areas of the picture that change. MPEG-2 is what made it possible for DVDs to work and for DIRECTV and DISH to fit 600 channels of standard definition on one satellite.
H.264 Compression
H.264 is also known as MPEG-4 AVC and is a standardized form of MPEG compression that is designed to do very well with high definition sources. As a result of more complex math and better decision making in the software, H.264 files can be 33% smaller than MPEG-2 files and still retain the same quality. It’s what makes Blu-ray discs work and it’s the standard used by DIRECTV and DISH.
H.265 Compression
H.265, also known as HEVC, is a form of compression that seeks to make 4K video small enough to send over the internet. Files compressed with H.265 aren’t necessarily smaller than H.264 but they often have a lot better fidelity at the same size. H.265 is the fastest codec for 4K, which is important if you’re doing real-time broadcasts.
VP9 Compression
VP9 is a competitor to H.265, pushed mostly by Google and YouTube. This is an extremely efficient codec that decompresses easily and produces very small files. However it can be slower than H.265 so it is less useful for real-time broadcasts.
Improving the process
If you’re not worried about getting a signal out in real time, you can do very well by running a video file through a codec multiple times. This is known as “multiple pass encoding.” Sending a compressed file back through a codec can make additional changes that are really small but can continue to shrink the files down a lot. However, it’s practically impossible to do it on-the-fly without a lot of hardware and there’s going to be some sort of lag somewhere.
The real improvements…
The math found in codecs gets more and more complex, especially on the encoding side. Today’s encoders have to be a lot more powerful than ones just a few years ago. They’re required to deliver smooth, high quality 4K files without any noticeable flaws. Luckily computer chips continue to improve and those chips are finding their way into encoders that do things which seemed impossible just a few years ago. These advances in technology make it possible for DIRECTV and DISH to provide better quality. They can also fit even more channels onto their existing satellites. It’s also making it possible for services like Netflix to deliver 4K that is slowly, (and I mean slowly) getting to the point where is looks better than HD. Still a lot of work to be done there, but it’s getting better.
Get the most out of your TV experience
Generally you as a consumer don’t have to worry too much about video compression. There are some things you might want to worry about. Think about replacing cables and upgrading your home entertainment system. For these needs, you might want to check out the great selection available at solidsignal.com. You’ll find everything you need to make sure your video experience is the best it can be.
