Lossless Video Compression: What is it and Why Should I Care?
One thing we take a lot of pride in at Archival Works is the fact that we do true, lossless video encoding, lossless video compression, and lossless capture to perform our video tape to digital file conversion. But what does this term “lossless” mean anyway? And why should you care?
In short, lossless means exactly what it sounds like. It means “without loss”. If something doesn’t have loss, then it is loss-less. Makes sense, but what is it we’re losing? Or in the case of Archival Works, what is it we are NOT losing? The answer? Picture quality.
Full quality and uncompressed video files are massive. Far larger than most people think. A 60-minute standard definition video (like that from a VHS tape) is a massive 70 GB. A two hour long 4k UHD movie like you might find on Netflix would be 3400 GB—That’s huge!
These files are so large they can be difficult or even impossible to watch or to work with. To solve this problem we use lossless video compression (sometimes also called “lossless video encoding”) to make video files smaller. There are two types of compression: lossless compression and lossy compression. There’s the term again- “lossless”… But “lossy”? What does that mean? Lossy, is the opposite of lossless; that is to say, picture quality HAS been lost.
Lossless video compression is exactly that. It is compressing the video file to make it smaller, but doing so without actually losing any quality. This is done through clever mathematical trickery and is what we use here at Archival Works. Lossless compression is great because it makes the video files smaller and easier to work with, but does so without damaging picture quality in anyway. This has the added bonus of being reversible; you could reverse the process and end up with the same file you started with. However, the one flaw of lossless compression is that it might not make the files small enough for a particular workflow. In my experience, a two hour Hi8, Video8 or other 8mm tape is around 140 to 200 GB uncompressed, but can be shrunk down to around 50 GB with the most advanced lossless video compression technology available. 50 GB is still pretty big, but not too big for us at Archival Works. However, most of our competitors aren’t able to work with files that big, so they resort to lossy compression and damage the quality of your videos in the name of convenience for themselves.
Lossy compression is used very widely, and to be clear, it isn’t all bad. It does serve a purpose. Lossy compression is great for making files small enough to share online or to watch on a smartphone. Lossy compression is designed to be used AFTER a video file is completed and isn’t being worked on anymore. Lossy compression is meant to be used to “lock” your video into place.
Lossy compression makes video files smaller by reducing the quality of the picture. Once this happens, there is no going back. That is to say, you can’t get that lost quality back again. It’s gone. That is why it should only ever be used as the final step. You do all of your work with lossless files, and then you save a lossy file for publishing on the internet. This will result in the best possible lossy file at the end of the process. If you use lossy compression more than once, you damage the video multiple times. This also results in an unnecessarily low quality file because the video file didn’t have to be damaged by lossy compression more than once.
For example, let’s say you have a 50 GB lossless file sourced from a VHS-C tape. You want to shrink that file down to 3 GB so you can put it on your iPhone. If you go directly from the 50 GB file down to the 3 GB one, you will end up with something that looks really good. That’s because lossy compression was only used once. This is a good way to use lossy compression, and at Archival Works, whenever customers order lossy files, this is what we do.
Let’s now look at a bad example. One of our competitors doesn’t like to be inconvenienced by that 50 GB file. So, they use lossy compression to shrink it down to around 9 GB. From here, they carry out with the rest of their processing. Then, they shrink it again with even more lossy compression down to .75 GB. This will give you a file that makes your tape look pretty bad. This happens for two reasons:
Compressing down to .75 GB is a lot and forces the computer to throw away more picture quality than if the file had been shrunk less. .75 GB is four times less than the 3 GB used by us. That’s a massive amount of data to have to lose for little to no real world benefit as 3 GB is still small enough to share online or to watch on your phone or smart TV.
The file went through two rounds of lossy compression. Even if they ended with a file that was the same 3 GB size as ours, theirs would still look worse because of that first 9 GB compression they did for their convenience. Every time you use lossy compression, quality is lost.
There could be other highly technical factors at play here such as lossy video codecs used, compression efficiency, chroma subsampling, bit depth, and more, but those are mostly outside the scope of this post. With that said, if all of those things are kept equal, the 3 GB file will have four times more data than the .75 GB file, and thus, look significantly better.
At the end of the day, lossless video compression is key in keeping videos looking their best. And while you might want a lossy video file to watch on your phone or to share online, that lossy file should be sourced from a lossless one to help it look its best. At Archival Works, we use lossless compression in every step of our process. The competition doesn’t.
For you video professionals out there: if you are trying to make your own video files, for video tape capture or otherwise, consider these codecs:
Lossless Video Encoding (Uncompressed):
v210 (Can be contained in Quicktime MOV)
v410 (Can be contained in Quicktime MOV)
Uncompressed UYVY (Can be contained in Quicktime MOV or AVI)
Uncompressed YUY2 (Can be contained in Quicktime MOV or AVI)
Lossless Video Encoding (Compressed):
FFV1 (Fun fact, FFV1 is created by the developers of FFmpeg)
HuffYUV
Lagarith
H.264 Lossless
Apple Animation
OpenEXR (Technically this an image sequence format, but it’s almost exclusively used for motion picture)
Lossy Video Encoding:
H.264
HEVC (Also called H.265)
VP9
ProRes
DNxHD and DNxHR
AV1 (Still experimental at the time of writing)
Note: ProRes and DNxHD/DNxHR are sometimes falsely called lossless or “visually lossless”. They are very high quality, but not lossless.