Allows to 'morph' or hybridize two sounds, so the resulting one has intermediate characteristics. Control on the actual blending is provided to the user through a set of envelopes that define the evolution in time of several controls. These controls are:
Sinusoidal component amplitude: Allows to control an aspect of the timbre blending performed, concretely the amplitude of the sinusoidal component, that is the amplitude of the spectral peaks of the resulting sound expressed as a interpolation factor between the peaks of the two original sounds. The X axis is time, expressed in a parametric form, and the Y axis is the interpolation factor between the two sinusoidal amplitude envelopes. An interpolation factor of 0.0 means that the resulting sound sinusoidal envelope will match the source's one, and a factor of 1.0 means that the resulting sound sinusoidal envelope matches exactly the target's one.
Sinusoidal component frequency hybridization: Similarly to the previous one, it allows to control an aspect of the timbre blending performed, concretely the frequency of the sinusoidal component, that is the frequency of the spectral peaks of the resulting sound expressed as a interpolation factor between the peaks of the two original sounds. The X axis is time, expressed in a parametric form, and the Y axis is the interpolation factor between the two frequencies of peaks in the sinusoidal component. An interpolation factor of 0.0 means that the resulting sound sinusoidal frequencies will match the source's one, and a factor of 1.0 means that the resulting sound sinusoidal envelope matches exactly the target's one. Note that this control has rather weak effect on harmonic sounds analyzed as such. In that case the effect will be driven by the value in the pitch hybridation factor except in frames where no pitch has been found, it is then recommended that you synchronize this factor to pitch.
Pitch Hybridization: Allows to control sounds' pitches blending. Note that this involves not only to blend the fundamental frequency, but also any accompanying harmonics detected as tready pointed out in the previous control. The X axis is time, expressed in a parametric form, and the Y axis is the interpolation factor between the two sounds pitches. An interpolation factor of 0.0 means that the resulting sound harmonics will match the source's one, and a factor of 1.0 means that the resulting sound harmonics will match exactly the target's one in their frequency distributions.
Residual component amplitude hybridization: This widget allows you to control the blending of the two sounds residual component amplitude envelope. The X axis represents transformation time, in a parametric form, and the Y axis represents the interpolation factor between the two sounds hoarseness. So a interpolation factor of 0.0 means that the resulting sound noisy component will match source's one, and an interpolation factor of 1.0 means that the resulting sound residual will match target's one.
Finally, for the sake of comodity, a 'global' envelope control is provided, that allows to define morph behaviour across all dimensions. All the previously introduced controls can be synchronized to the global envelope thus becoming non editable (note that this is the default behaviour)
The option to interpolate 'intermediate' frames is given so to define SMS Morph behaviour when source and target morph sounds' lenghts differ. If it is deactivated, Tools will resample shorter sound interpolating missing frames from the neighbourghing ones. Otherwise, the resampling will be accomplished by always taking the closest frame so sometimes frames will be repeated. Transformation result will sound better with linear interpolation but also will be a bit slower.
The controls we will explain from now on could be considered as "advanced" and we cannot hope their meaning to become completely clear in this context, without some usage tutorial, we hope though that with this brief description and your experimenting, their usage can become clear
The Time Synchronization defines a temporal mapping between the source and the target sound. X axis is the normalized time for the source sound and Y is the normalized time for the target. Using this control you can achieve morphs on only some parts of the sound, define how you want to stretch the shortest sound or even reverse one of the sounds in respect to the other.
The following three parameters form a non-separable set: SinusoidalShapeW1, SinusoidalShapeW2 and SinusoidalShapeInterpolationFactor . The first two define spectral envelopes where, and for every frequency, a Y value of 0 means that the amplitude of that frequency will be taken from the source spectrum and a value of 1 means that the amplitude should be that of the target spectrum. The third control is a temporal envelope that defines how much of each of the previous spectral shapes should be taken in every instant. An instantaneous value of 0 in this temporal envelope means that SinusoidalShape1 should be used and a value of 1 means that the SinusoidalShape2 should be used.
The same happens with the controls ResidualShapeW1, ResidualShapeW2 and ResidualShapeInterpolationFactor although this spectral shape interpolation in the residual spectrum has not been implemented in the current release.