You get practically the same wayform, that could be either in the time domain or frequency domain.
The frequency domain speech coders and adaptive transform coders, whereas the time domain coders are non differential coders, are the pulse code modulation, just now I have described. Then you have the differential pulse-code modulation or differential code-modulation schemes, where it could be Delta-modulation (Adaptive)Delta modulation or continuous various slope delta modulation.
On the one hand an DPCM, that is the differential PCM adaptive differential PCM, and adaptive predicting coding are the other forms of the differential codings in the wayform kind of coders. You could have the source coding using the techniques of linear predictive coding, or you could have vocoders there working in the frequency domain and the famous vocoders are channel, formant, cepstum or voice excited vocoders, whereas in the linear predictive coding, under that heading you could have multipulse excited LPC, regular pulse excited LPC, code excited LPC, or residual excited linear prediction. You could as well have vector sum excited linear prediction or a linear prediction coder developed by the Qualcom.
In fact the cellular mobile communication primarily uses the linear predictive coding. The reason being that it gives moderate bit rate on the one hand, and gives tolerable quality on the other hand. Therefore one goes for the LPC kind of coding.
