from phasing to phase alignment

In the recent article about audio signal coloration I’ve already talked about the importance of the signals phase response in respect to the perceived tonal spectrum and today I’m going straight ahead towards phase alignment and how a signal delay relates to the phase response. But first let’s have a look at some nice youtube stuff showing Jonathan Little on demonstrating his Little Labs IBP phase alignment tool.

Having a closer look to what the IBP actually does, it turns out that it basically offers some different methods and configurations on how to “align” already shifted signals back into phase. The main method combines two 1st order all-pass filters in a serial fashion altogether with some additional twist. The basic phase shifting shows up as in the following diagram (90/180 phase adjustment engaged):

example phase shift in the IBP

There are some further alignment options such as the typical polarity switch (also referred to as phase inversion) or a sligthly more sophisticated phase frequency center switch. The interesting thing about the IBP is that even if some rather constant phase shifting approaches (which are possible in the digital domain) might appear to be more reasonable or accurate in theory the IBP approach still maintains a more interesting or  musical compensation at the end (as mentioned by some users and judged by their hearing).

Such alignments can be used for rather “artistic” mixing treatments and audio signal coloring as well. As a simple example, adjusting it properly to the upper frequency range one can easily achieve some serious amounts of transient smearing or, applied to the lower frequency part, the bass range can be decoupled to some extend (and so some bass enhancers are taking advantage of such effects as well).

The digital implementation of the IBP offers yet another method for phase alignment which is just a plain digital delay w/o any further bells and whistles. The ultra short delays (basically below around 4ms where they are indistinguishable from the source by human hearing) are adding a more drastic and different phasing effect. Technically speaking, one could easily create  an allpass filter by utilizing a simple delay line too and so both effects appears to be just like two sides of a coin.

In this way all the well-known time shifting based effects (such as chorus and phaser just two name the two) are implying heavy phase distortion but I’m not going to follow that specific lane any further here. Instead, in an upcoming article I’ll gonna show you that there is just a little gap (circuit wise) on how to extend the already discussed concepts into a quiet simple but already musical sounding delay unit.