Quote:
Originally Posted by spyboy /forum/post/16966067
Hi Mark
Can you comment on the importance of the time domain correction in the newer (than the SMS-1) EQs? Around here it seems that more than a few people consider the SMS-1 outdated because it cannot correct in the time domain.
Thanks In Advance
There does seem to be a lot of speculation around here about that.
I can understand the attention being given to efforts to look more at the lingering energy at low frequencies. Meridian and Lexicon both have given this attention to this in their room EQ efforts from a few years back.
I haven't had time to get into the exact details of what the Audyssey system is doing, and I expect they prefer to keep some of the finer details to themselves. What I have seen suggests to me that they are measuring and defining the EQ efforts based on the LF energy vs. time. I have not seen any indication that they are direct correction vs. time such as with impulse response correction where functions can be created that are variable over time.
My understanding, which Ed Mullen or Chris from Audyssey could probably correct or confirm, is that they are looking at both the LF decay times and the total energy at multiple locations, with which they use their proprietary algorithms to determine a target amplitude (and possibly phase) response which they then implement using FIR based filtering which does allow for zero-phase response EQ, as well as targeting a given phase/group delay response.
The big difference here should be in the ability to manipulate magnitude and phase/group delay separately, ie changing one while not affecting the other if you choose to. More conventional response EQ emulate analog filters (assuming DSP based) using IIR / minimum phase filters. Such filters have direct correlations between changes in magnitude vs. phase.
Of course some have argued (ie Welti & others at Harman) that at low frequencies room issues we can or should be correcting for are already minimum phase in nature, where accurate response EQ does greatly improve the time domain behavior. I do know that not every blip at low frequencies is minimum phase, but those cases also tend to be ones which vary greatly from one listening location to another.