Why Does Video Compression/Rendering Require So Much Power?

[Struzinski]

I like to ask questions It's something we've all done at some point; whether it be through WMM, Sony Vegas, Adobe Premiere, AfterEffects, or Handbrake. But something I've always wondered was why it takes up so much time and CPU/GPU power to render HD videos? I use VirtualDub to compress Fraps 720p 45fps gaming streams, and all I have is a dual core i7 w/o a dedicated GPU. Obviously, my processor is taking an asswhooping everytime I do this.

What causes the asswhooping to occur? I mean visually, all I see is that the video file is being shrunken in size by using MPEG4 codecs for video and MP3 codecs for audio. I just don't understand what is processing in the computer during this procedure.

 

[iFargle]

Well.. Not sure about the others, but for compression the rules are pretty much always the same.

Lossy compression takes a set of pixels and instead of each having their own individual value they are essentially mapped and compared to one another. If two pixels are nearly identical, their values are changed to something that's a "best fit". That's why you get those artifacts and fuzziness in dark colors.

Lossless compression, on the other hand, will only look for exact matches, which is why lossy videos are generally much smaller than lossless.

And that's why it takes such a toll on your processor. It has to scan the entire video frame by frame, pixel by pixel, to find matches.

 

[Struzinski]

Makes sense to me. Thanks for the reply!