Synthesizer V lets the “strength” of each phonemes to be set customly. Yet this notion doesn’t seem to be phonetically justified.
Currently, the phonemes that are affected the most by this notion are sibilants. I verified this by trying modifying the strength of the English phoneme
s. But when this parameter is set to the minimum, the
ch phoneme sounds quite buggy, depending on the voicebank (Kasane Teto for example). I consider this a bug.
Furthermore, this parameter seems to affect plosives rather inconsistently. For the English phoneme
p, setting the parameter to the maximum lets the said phoneme to become tenuis in some voicebanks (Asterian for example), and doesn’t seem to be affecting it much in some other voicebanks (Ryo for example). Likewise for setting the parameter to the minimum. What is much more confusing is, the inconsistency depends not only on different voicebanks, but also on different phonemes. For Ryo as an example, the parameter seems to be affecting his English
k phoneme for its aspiration.
My proposed solution is like this: For fricatives, retain the strength parameter, but rename it to “frication”. For plosives and affricates, let there be two strength parameters; one for the duration of aspiration, and the other for the frication during the aspiration. And for the others (nasals, liquids, and vowels), drop the strength parameter. After all, there is no phonetic notion of “strength” for them.