You can use the MultiToMultichannel module to process channels independently; it's similar to the Replicator, but it peels off each individual channel, processes it, and outputs the processed signal on that channel. Also, instead of referring to ?VoiceNumber and ?NumberVoices, you can compute parameters using ?ChannelNumber and ?NumberChannels.
Here are some examples of using the MultiToMultichannel to process live-analyzed spectra and split-multichannel spectral read from files (using the Morph1dSpectrum which can play back split-multichannel spectra). In the example, I used a live analysis of an 8-channel source. And a file example using stereo spectra, which you can modify to process 5-channel spectra by unchecking the bypass check boxes in the MultiToMultichannel.