Have you ever heard of these "codecs" before?

So I stumbled upon an article about the use of codecs and what they were and in all honesty, I have never heard of these things until just now. I guess I always just figured codecs were the files like .mov or like .m4p and all that, when in reality those are "containers". If you would like to give that article a read or if you would just like to watch the 43 minute video, I will put the link here for you guys to go see.

Basically codecs stands for compressor/decompressor. They are used when you're trying to export a file in a smaller or larger file form. At least that was my prediction. Codecs are not the same as a container, codecs are stuff such as h.264, PRO-RES, and DNxHD. While containers are things such as .mov or .m4p and that is the file form. Think of a container as a bucket and codecs as things, you put codecs (the things) into the container (bucket).

You would want to use these codecs when you are doing what I call the 4 step process which will be, capture (what your camera is recording), edit (use to edit), delivery (exporting and final product (file put online, client, or burn to dvd)), and archival (back up of final copy to make better quality). For when you are on the capture form, you're gonna want to use h.264 and a 4:4:4 chroma-sub-sampling. Edit, you will use a small form of DNxHD. When you go through the delivery process, you will use the highest for of h.264 and finally when you go to archive it, you will use a super high res form of DNxHD to get the really big file and make it easier to compress if you need it again.

Ah, now we're here to bit depth, and basically the deeper your bit depth, the less "bars" or "lines" of different color you will see, and it will just be one giant gradient. An 8-bit depth is a real world type of bit depth. Basically that is saying two to the eighth power, or 256, is made up of different values for R, G, and, B. so two to the tenth, 1,024, would be the amount of different values for a 10-bit depth. 8-bit is a perfectly fine and acceptable way to work with bit depth.

Chroma-sub-sampling is something that is a little more difficult to understand and I'm still not sure what it is, so I'll explain what I get. There are three different types of sub-sampling, and those are basically 4:4:4 (no sub-sampling), 4:2:2 (some sub-sampling), and, 4:2:0 (a lot of sub-sampling). So basically imagine 4 pixels two on top, two on bottom. for 4:2:0, 3/4 of the pixels lose their own color and use only the one's color that is left, all of them keep there color. 4:2:2 is basically two of the pixels lose the color, and take the colors from the one next to them and applies their brightness.

A visual diagram of Chroma-sub-sampling

Spatial Compression is simple and basically an image is cut into boxes and thinks "hey, these pixels are all the same color, we can save space by coping and pasting!" Those are the little boxes you see in images at some point.

Up to here all the other Codecs have been intra-frame compression.

Temporal Compression or Inter-frame, is a compression over time or a compression over multiple frames. This uses a technique that effects multiple frames at once. This is also pretty short and sweet to understand. Basically it will only cut out the different parts of an image, and keeps saving the rest over the frames of the shot. So if you have a shot of a head and a guy talking, the only difference in the frame would be a mouth moving slightly, the background should stay the same. So what it does is basically make a box and put it around the stuff that changes, and makes that the difference, it copies and pastes the rest, and makes then puts the difference on top, this just saves file room per shot.

Up to here we've been cutting up data, well with the Lossy codec, it basically makes a file larger, and you will need it probably less than 1/100th of a time.

A bit rate is simply how much data a codec uses. We measure bit rate as bits/second so stuff like kbps or Mbps. Now 1 Byte, is equal to eight bits. So if I have a bit rate of 8Mbps, and I want to know how many bytes I'll get in 5 minutes, you basically do:
-Take 8 Mbps divided by 8 and then you get 1 MBps (mega byte)
-Then take 1MBps times 60 and get 60 MBp Minute then 5 times 60 to get 300 MB in five minutes.
Basically the higher the bit rate, the better the quality. And you can't compare two different bit rates with different codecs.

RAW is just as it sounds, the original untouched footage.

And that wraps up my little clarification on how codecs work.