Another counterintuitive thing with digital video systems is this:

The worse the quality of the source- the harder it is to compress.

For example if you source something from VHS - its much harder to make it compress than if the source is digi-beta.

This is because there is a lot of noise and distortion in VHS images.
The compression algorithm has to compress this noise as well.
The data rate of a VHS image - compressed with a codec - will be higher than a digibeta image.

So higher 'quality' images are actually easier to deal with - at greater compression rates- which means you can play them back without raid arrays.