The human voice is a consolidation of the waves of electrical energy carried across the given channel capacity. Humans generate a combination of amplitude and frequency changes in a continuing flow. If the changes are held constant, then the conversation becomes monotone, highly unacceptable for the average conversation. Indeed, if everything were held constant, the recipient of the information would be lulled to sleep. We use the variations in our voices to reflect deviations in emotion, accentuation, or articulation and emphasis on certain points. Every voice generates a different pattern of amplitude and frequency changes. This is what gives every individual a unique and recognizable speech tone.
It is interesting to note that the female voice pattern typically generates more changes in amplitude than in frequency. This accounts for the higher pitch in the female voice. Conversely, the male vocal pattern generates more frequency changes than amplitude shifts. This accounts for the low pitched, more grainy tones of the male. These are averages; every human is different and the norm can be deviated from at any time. Suffice it to say that when someone is highly emotional, there are definite shifts from that person’s normal voice patterns. The pitch might go up by varying degrees, showing the results of stress or anger.
The telephone links are designed to handle this widely varying pattern of shifts. However, from time to time the voice pattern might exceed the differences in the allotted bandwidth. Then the frequencies in the voice go above or below the normal range. At this point, the band pass filters start to remove the excesses. What can result is a pattern of tinny or flattened speech conversations. This can be also heard if someone uses the letters F and S where the frequency ranges on these two sounds might exceed the band pass ranges; they get flattened.
Other Communications Services: Data, Facsimiles, Images, Video
Because the telephone network was built to carry the analog equivalent of human speech patterns, the other services that we wish to communicate, such as data, facsimile, images, and video must be transmitted within the same constraints as voice calls. The network allows any telephone set to contact any other telephone set across a 3-kHz bandwidth. If you want to communicate anything else, such as data, it must be accommodated on this same 3-kHz bandwidth. This invokes a limitation on the speed of our data communications channel capacities. Although modems can transmit signals more quickly, they must be constrained into the size of the pipe. A video conference transmission will also reflect non-real-time motion because of the channel capacities on a dial-up basis.
If you need to move more information across the channels, you have two choices. One option is to dedicate a high-speed line between the two or more end points we need to communicate with. However, this approach will eliminate the possibility of “any-to-any” connection and will require planning well in advance of communicating with the end points. Leased lines between the points might be underutilized and could be more expensive.
Another option is to have the band pass filters moved out to (for example) 8 kHz between the end points. Unfortunately, however, there are similar limitations here. We would have to know every location that we would be communicating with, losing the benefits of the any-to-any connection. As well, there is no guarantee that the switching systems will route the call over the same path on a switched dial-up connection every time. This would force us back into a 3-kHz channel along the route selected if it is different every time, eliminating the benefits of the wider channel capacity.
