@erik_squires wrote: "Duke, Having a pro in the thread is like cheating. :)"
"If you’re not cheating, you’re not trying hard enough." - Gene LeBell (off topic, but he’s the old guy who choked out Steven Seagal. Twice. In the same day.)
Erik: "Can we define good though? I mean, I agree with the on/off axis description, but! What about imaging and detail?
"How would omnis compare to dispersion limited speakers like big ESL’s, line arrays or horns with narrow beam pattern? Pro acousticians I’ve read say that the better the dispersion control is, the less room treatment is required.
"The Omni story flies in the face of this, unless we don’t care about detail."
EXCELLENT points!
Literally EVERYTHING you have said here is why I don’t do omnis myself. (I use constant-directivity waveguide-style horns.)
Experiments with varying the level of spectrally-correct reverberant energy convinced me that there is a "sweet spot" above which the reverberant energy is arguably "too loud" and clarity starts to be degraded. The "sweet spot" level for the reverberant energy is lower than one would normally get from an omni or from a dipole.
The arrival time of our spectrally-correct reverberant energy matters as well, and imo a dipolar or bipolar radiation pattern is generally preferable to an omni pattern because of geometry: The path length for a bounce off the wall behind the speakers is usually longer than the path length for a bounce off the near-side wall. Imo we want to minimize the reflections arriving within the first 10 milliseconds or so (for the sake of clarity and imaging), but then after that spectrally-correct reverberant energy is generally beneficial (for the sake of timbre and envelopment). Credit to Earl Geddes for that 10 millisecond figure. Siegfried Linkwitz says something similar regarding reflection arrival times, though the figure he arrived at is 6 milliseconds.
The approach I use might be called a polydirectional (credit to the late great Richard Shahinian for that term). I use fairly directional main array and then a secondary, similarly directional array aimed up-and-back such that its energy bounces off the wall and then off the ceiling before reaching the listening area, to maximize the time delay without requiring as much distance from the wall as a dipole or bipole would. The level and spectral balance of this secondary array are user-adjustable, so that different room acoustic situation can be adapted to.
Duke
"If you’re not cheating, you’re not trying hard enough." - Gene LeBell (off topic, but he’s the old guy who choked out Steven Seagal. Twice. In the same day.)
Erik: "Can we define good though? I mean, I agree with the on/off axis description, but! What about imaging and detail?
"How would omnis compare to dispersion limited speakers like big ESL’s, line arrays or horns with narrow beam pattern? Pro acousticians I’ve read say that the better the dispersion control is, the less room treatment is required.
"The Omni story flies in the face of this, unless we don’t care about detail."
EXCELLENT points!
Literally EVERYTHING you have said here is why I don’t do omnis myself. (I use constant-directivity waveguide-style horns.)
Experiments with varying the level of spectrally-correct reverberant energy convinced me that there is a "sweet spot" above which the reverberant energy is arguably "too loud" and clarity starts to be degraded. The "sweet spot" level for the reverberant energy is lower than one would normally get from an omni or from a dipole.
The arrival time of our spectrally-correct reverberant energy matters as well, and imo a dipolar or bipolar radiation pattern is generally preferable to an omni pattern because of geometry: The path length for a bounce off the wall behind the speakers is usually longer than the path length for a bounce off the near-side wall. Imo we want to minimize the reflections arriving within the first 10 milliseconds or so (for the sake of clarity and imaging), but then after that spectrally-correct reverberant energy is generally beneficial (for the sake of timbre and envelopment). Credit to Earl Geddes for that 10 millisecond figure. Siegfried Linkwitz says something similar regarding reflection arrival times, though the figure he arrived at is 6 milliseconds.
The approach I use might be called a polydirectional (credit to the late great Richard Shahinian for that term). I use fairly directional main array and then a secondary, similarly directional array aimed up-and-back such that its energy bounces off the wall and then off the ceiling before reaching the listening area, to maximize the time delay without requiring as much distance from the wall as a dipole or bipole would. The level and spectral balance of this secondary array are user-adjustable, so that different room acoustic situation can be adapted to.
Duke