Opalchip: A microphone picks up whatever is fed into it, both direct and reflected energy. It can't discern if the primary or reflected signal should should dominate as it can't differentiate between arrival times and their individual intensities. In effect, it becomes a recorder of acoustic activity at that specific point in time and space based on the specific pick-up patterns of the mic being used.
The Walsh driver simply re-radiates the energy that was captured at the mic as a point source and re-radiates it into the listening environment as a point source. The fact that the original ambient sounds heard during the recording could be heard at every point in the room, and are pretty much preserved and reproduced due to the pseudo-omni radiation characteristics of the Walsh design, is one of the most endearing properties of these speakers. The fact that there is only one driver acting as a point source for each channel reduces time / phase distortions to a minimum, hence the preservation of natural harmonic structure. This too is a very endearing quality of this speaker design. The effects of binaural recordings as heard on these speakers is pretty amazing.
Other than that, each musical note has a primary frequency and multiple harmonic frequencies. These harmonics vary in spectrum and intensity. Any device that tries to separate the audio spectrum into different segments will introduce distortions into each note reproduced. That's because the time, phase & amplitude of the primary note vs that of the harmonics will not remain cohesive in presentation.
As a case in point, the specific device that you mention is capable of expanding multiple different frequency regions at different rates. When doing this, it means that a harmonically rich instrument ( like a Cello ) that is centered in one specific frequency section may be expanded at a different rate than the harmonics, which might fall into one or two different frequency spectrums. As such, each spectrum is / can be expanded at different rates, which in turn varies the amplitude of the harmonics in respect to the amplitude of the primary notes.
The reverse of that is also true. That is, an instrument that covers a very wide range of the audio spectrum ( like a piano ) can have different levels of expansion applied to it across the entire band due to the spectrum segmentation that the device does as part normal processing. This would take place on both the primary note and the harmonics.
As such, expanding a compressed recording could only be done optimally if the algorythms used during recording and playback were exactly the same. Given that this is next to impossible given the differences in recording, mixing and processing techniques, the end results of attempting to expand a compressed recording can be very "interesting" to say the least. I will agree that "expanded" music sounds noticeably more dynamic and "punchy", but at the same time, it also has a certain "artificiallity" to it. On top of that, quite a bit of electronically generated music IS compressed, even when played live. Electric guitars, bass guitars, electronic keyboards, etc... are often processed in a certain manner with the musician specifically desiring certain sonic attributes that compression / clipping bring along with them. Trying to "undo" what was meant to be, both live and on the recording is nothing more than a distortion. These distortions may be pleasant on certain recordings, but it all boils down to a matter of personal preference vs articulate preservation of what is on the recording. Sean
>
