Taxonomies such as "emphasis, level and deviation" are quite obscure (especially for non technical users). Try to sobstitute them with more
clear and accessible terms . I would suggest to move towards a more perceptive taxonomy (e.g. emphsis = flanger depth, density deviation = granularity,
size deviation = size variation?)
Consider to hide some parameter in order to simplify the interface (e.g. size deviation and density deviation could be fixed parameters
not accessible to the user).
Generally, for technical labels, add number boxes with related unit of measure to display sliders values
Model comments
While playning around with the parameters it is very easy to end up with quite unrealistic sounds. Consider to rescale or limit
the range of the various parameters (see in particular the "emphasis" slide).
The RTSFX pour is quite metallic and "synthetic". A possible improvement could be related to the filtering of the higher frequency range
(low pass filter - cutoff frequency at 10000 Hz)
By listeing to the samples it is possible to note that the act of pouring a liquid is usually coupled to drop sounds (generally little size drops).
In order to make the synthesis more realistic, it might be useful to juxtapose a rain-drop synthesis to the pouring one .
By looking at the sample spectrograms it is possible to note that the main energy contents is usually constrined in the low register
(up to 1500-2000 Hz). Each of these sounds are characterised by a glissando (pitch shift) behaviour. The RTSFX synthesis generates similar pitch variations
but the various pitch shifts are spread over a too large range (up to 5000 Hz). In order to emulate the samoles, consider to constrain this glissando scope .
