Maybe this is why corner placement sounds good

[Greg Oshiro]

<snip>If you can hear the difference between a 500 Hz "suck-out" due to near-field reflections and a room mode at 500 Hz, you have outstanding hearing. <snip>

One of these days I'll have to set up such a reflection and listen. Here's what I think it would take: a 2ft X 2ft (about one wavelength at 500 Hz) piece of 3/4" thick plywood placed just outside the direct line-of-sound from speaker to listener; angle the plywood using the mirror-on-the surface trick we usually use to find reflecting surfaces when applying fuzzy stuff to the room; measure the two path lengths and adjust the reflector for a path length 1mSec longer (13.5"). Put some fuzzy stuff on the plywood to attenuate the reflection by 20*LOG10(0.5x0.9) <the factor from the original example> and give it a listen.

Just visualizing this, it's a pathological arrangement. I think it would corrupt the direct sound from the speaker and be clearly audible.

I *think* the room modes require more time to be heard. If the sound has to travel from the speaker to the listener (time 1) and then (to a reflecting surface and back to the listener)=time 2 in order to cancel, which is the null in a room mode. Time 2 might be ~10 mSec if there's 6ft to the wall behind the listener. So there's 10 mSec of direct sound only for the listener perceive before the room sound (including modes) comes into play.

IIRC, the recording studio control room designer folk want the reflections to arrive at the "sweet spot" no earlier than 20 mSec after the direct sound *and* they want the reflections to be diffuse, hence the exotic back wall treatments in control rooms.

<snip> You might also want to place your microphone close to your listening position so that you're measuring something that resembles what your ears are receiving... <snip>

One of these days... I'm still trying to get the anechoic output of the speaker straightened out. In larger performance venues the strategy of flattening the anechoic response of the speaker is known to work. In domestic listening, Floyd Toole's book says the same strategy works. I'm still trying to figure out how to get anechoic data given the tools and room available. If I had speakers that didn't need a corner, it would be a no-brainer; drag them out to a large flat open area (concrete parking lot) and make time-windowed ground-plane measurements. Getting anechoic data from K-horns in a 12ft X 17ft room requires thinking that may exceed my capabilities.... The TEF25 measuring system can "ignore" reflections, but at the expense of frequency resolution. The more reflections are ignored, the less detail there is in the data, meaning peaks/dips get smoothed out, reducing the value of the data.

<snip> ...the dark green trace is the final EQed response at my listening position...<snip>

Hey! That doesn't suck! <This is an accolade among some professional sound folk I know>

I'm going to have to read up on REW. It probably doesn't measure the same way as TEF, so data interpretation will be different than what I'm used to.

[Greg Oshiro]

<snip> I've really been enjoying your tests and posts....keep them coming.

mike tn

Thank you. I have a known tendency to spiral off into engineering-speak. Please let me know if I get out of control....