At the macroscopic level signals (to be) perceived by human being are analogue. Over the years the technologies to convert analogue into digital signals have greatly improved but the intrinsic high bitrate of the digital signals have remained constant. The Nyquist theorem ensures that if you have a signal whose bandwith is limited to B, you must sample it at 2B (practically to a little more than 2B) to get a perfect analogue-digital equivalence. To be digitally handled samples must be quantised: the samples of a graphic page are typically quantised at 1 bit, speech and video at 8 bits and music at 16 bits. These are the typical bits/bitrates of some digital signals
| Information type | Bits/bitrate |
| A4 page | 4 Mbit |
| Telephone speech | 64 kbit/s |
| Television | 216 Mbit/s |
| CD music | 1.4 Mbit/s |
Therefore practical use of digital signals requires the use of compression algorithms. You will find some preliminary information on compression of