Improving the intelligibility
of SSB transmissions
What is important when processing speech?
200 to 400Hz frequency range and this, coupled with my ‘Mancunian’ pronunciation of vowels, doesn’t aid communication. In Western languages, human speech has components that fall into three main groups: vowels, consonants and sibilants. The vowels, A, E, I, O and U contain most of the energy in the human voice and generally occur in the frequency range below 500Hz. Consonants such as B, K, T and L convey the majority of information in speech and occupy the frequency range 500Hz to 3000Hz, but at much lower energy levels than vowels – in some cases they can be 30dB lower. Strongly emphasised S, Sh, Ch, Z and J sounds are termed sibilants and are found at frequencies above 3000Hz.
The vowels help define who is speaking and give clues to what is being spoken, but the consonants are the components of speech that actually convey useful information. Without the vowels it is difficult to identify who is talking, but without the consonants it is difficult to understand what is being said. The sibilants help to differentiate between words, and their absence makes it hard to distinguish between ‘F’ and ‘S’ or ‘D’ and ‘T’ sounds, however the majority of sibilants are removed when SSB is used as the signal is filtered to achieve a channel bandwidth of around 2400Hz. There is not much we can do about this, but fortunately the information conveyed by sibilants can usually be derived from the context in which words are used in sentences, making them slightly less important for good intelligibility.
POWER LEVELS. In an SSB transmission, the amount of transmitted power is defined by the level of speech energy being fed to the modulator. In a linear system, the majority of the speech power contained in the vowel sounds drives the transmitter output up to full power. However, this level may be considerably greater than the power generated by the consonants that are actually conveying the majority of the information in the transmission.
If speech compression or ALC is used to increase the average transmitted power level then further problems may occur due to the high energy, low frequency vowel sounds modulating the speech envelope that contain the high frequency, low level consonants.
MORE ..... Microphone equaliser - SSB TX audio