Low phase growth tune

Chris A

Member
Joined
May 22, 2017
Messages
47
Location
Arlington, Tx
More  
Preamp, Processor or Receiver
Emotiva XMC-1 AVP, Xilica XP-8080 & miniDSP 2x4 HD
Main Amp
First Watt F3, Crown D75-As (5 total) bi-amping
Additional Amp
Crown XTi-1000 for subwoofers (2)
DAC
Topping D10 Balanced (stereo only mode)
Universal / Blu-ray / CD Player
LG UBK90, Oppo BDP-103, Laptop
Front Speakers
Klipsch Jubilees (TAD TD-4002 compression drivers)
Center Channel Speaker
K-402-Multiple Entry Horn (full range)
Surround Speakers
Klipsch Belle bass bins with ESS AMT-1, bi-amped
Surround Back Speakers
-
Front Height Speakers
-
Rear Height Speakers
-
Subwoofers
DIY SPUD Tapped Horn (2) behind fronts
Screen
LG OLED 77"
Remote Control
Logitech Harmony One
Ron,

I think that the last time I tried using Dirac (the full version) on my Emotiva AVP, it appeared to be doing some sort of FIR filtering flattening of the midbass response (but not a lot). If it can do that, my guess is that it will try to do phase flattening at higher frequencies. You can think of that as "automatic", but there are some real downsides to "room correction" firmware, notably in the midbass region where your room starts to become the limiting dimension in reproduction of acoustic waves (i.e., the so-called Schroeder frequency of your listening room). These automatic packages always seem to screw up that portion of the spectrum, resulting in a thin sound overall.

There are other solutions available to use FIR filters to flatten phase, most notably RePhase as found mostly on diyAudio, etc., but there are other tools available. The problem is that you need some sort of DSP app or FIR-capable crossover box to flatten phase. And it's definitely not "automatic".

The processes that I use are all focused on eliminating all the all-pass phase growth through the crossover interference bands, but I live with the minimum phase response of the individual drivers in each way of the loudspeakers themselves. This is a very low-cost way (from a computing standpoint) that seems to give a 99% solution (i.e., within ±55 degrees of total phase swing for our MEHs) that seems to be below the threshold of audibility without having to resort to FIR filtering, with all its computing requirements and insertion delays.

YMMV.

Chris
 
Top Bottom