Phonemic and phonetic input

There is however a drawback with this approach. The problem with phone representations was discussed previously in Section 7.3.2 where we stated that although phone descriptions could be very useful in certain circumstances, they should be treated with caution as they are a discrete representation enforced on a continuous space and as such only crudely reflect the actual processes going on. To give an example, let us consider again an example of place of articulation assimilation that occurs with words such as 

---

Perhaps unsurprisingly, the consensus as to where this point should be has shifted over the years. When more traditional systems were developed, memory was very tight and hence the number of base types had to be kept low regardless of any errors. In more recent years technological developments have eased the pressure on memory making more abstract representations possible. Given this, there is more choice over where exaetly the ideal representation should lie. In fact, as we shall see in Chapter 16, the most successfiil systems adopt a quite phonemic representation and avoid any rewriting to a phonetic space if at all possible. Because of this, the pronunciation component in modem systems is in fact much simpler than was perhaps the case in older systems, and quite often the input to the synthesiser is simply canonical forms themselves, direct from the lexicon. 

Some systems use the output from the lexical lookup process and/or G2P conversion directly as the input to the synthesizer. In some systems, however, an additional step often known as post-lexical processing is performed, which modifies the output in some way. The most significant factor in determining how much post-lexical processing is performed is the specific form of the input to the synthesizer. This was discussed in Section 8.1.3, where we saw that we could choose high-level phonemic representation or a low-level detailed phonetic representation. A low-level representation will require significant processing, for example to turn an /n/ into an /m/ in a labial context, but, because... 

---