I'm thinking about using Disperser on midrange-y sounds to unmask the bass component, so I can cut them higher using an hi-pass filter and keep (or even enhance) the impact of those sounds, while making headroom for the low-end... It's just an idea I want to experiment with, I'm not sure of the results I'll get.
In relation to the higher peak values using Disperser, I had mixed results while trying the demo version; anyway I think this could be explained because of a property of all pass filters: they make asymmetric waveforms more symmetric, but they also make symmetric waveforms more asymmetric. I think this is a good article on the subject (have a look in particular at the "failure" chapter):
http://www.tonnesoftware.com/appnotes/a ... lpass.html
The key passage is:
Well... what if your sound is a square wave (or any sound with a spectrum dominated by odd harmonics)?If the waveform to be corrected has no even-order components (unlikely in practice) and so is symmetrical top-to-bottom in the first place, then the use of an allpass network might not be beneficial
Sometimes you can indeed get less headroom by using an all pass filter, so I think it may not be always the best tool for this purpose if we are talking about synths...
However that's just one of the possible uses of an all pass filter; I also won't mind too much a little increasing in the crest factor if that's the compromise required to unmask some frequencies better, I could need a little bit more of compression on that sound (or I may just lower it, depending on the context) but it could be still the best solution for that sound.
I think Disperser is a useful tool.
