Some years back, Lidia, myself and Jean-Luc Ohl (a friend and DIYer from Paris) did a study on the effects of early reflections from side walls. We wanted to see if a near side walls had more effect on imaging than farther ones. A simulation of a source and a side-wall reflection was created and subjects (many, don't remember #) were asked to evaluate the image location. The subjects found this hard to do and the results had a lot of noise. While this noise prevented a statistically significant result, the trend was clearly that the near wall was the more detrimental to the image stability. They also indicated that the reflections pulled the image towards the reflection (not surprising.)
So, could one place a woofer at each side wall reflective point, and actively cancel out the reflection? The only active room treatment I've seen so far talked about in detail has been from @Flaesh... He describes placing woofers at the back wall to absorb the impulse there and then experimenting with peak eq's to further tailor response at the listening position.
...In other words something akind of DBA.
https://www.diyaudio.com/community/threads/double-bass-array.401001/
https://www.diyaudio.com/community/threads/double-bass-array.401001/
This is doable, sure, but I haven't seen evidence that it is a significant improvement in bass over a well done multi-sub setup. I mean done right the two should be equal in performance - in a statistical sense. I would also hazard a guess that the multisub setup would have more envelopment due to the differing source angles involved, i.e. the multisub setup would be more random. I like random at LFs since the room absorption at LFs in my room is quite high. Hence there aren't significant room modes to "cancel". Just EQing (with DSP, of course) can yield a very smooth and decay free response.
But, remember, I am talking about LFs where you can control the room response. Above Schroeder all bets are off.
Think of this as well, any noise cancelling device can place a P=0 condition at its source, but for this to be effective at room volumes it has to be large to have any wide spread effect.
But, remember, I am talking about LFs where you can control the room response. Above Schroeder all bets are off.
Think of this as well, any noise cancelling device can place a P=0 condition at its source, but for this to be effective at room volumes it has to be large to have any wide spread effect.
Last edited:
The problem with random is the number of sources required for it to work well. If you have only a few then wise coherence is almost certainly going to work better.
High LF absorption would mean unusually soft room boundaries which are not normally found in a homes?
High LF absorption would mean unusually soft room boundaries which are not normally found in a homes?
It makes a person wonder what if Schroeder was above 700Hz, maybe in a car. Does a smooth steady state response from multiple random sources remain the goal?
Putting 700hz into the modal region, along with omnidirectional presentation of 700-hz from driver.... Yes.
OK that was a really quick guess, I think you have to consider perception vs frequency. I haven't memorized each part of the spectrums role, but somewhere up there, the image probably doesn't want to appear to come from everywhere. Then again, if 700hz is in the rooms modal region it will already have issues of appearing to come from "everywhere". My point is that in the modal region, many parts of the spectrum appear to come from other places than the drivers up front, anyway, so multiple subs spaced out, no harm no foul. The early arrival of direct energy is welcome, from all angles at that point.
OK that was a really quick guess, I think you have to consider perception vs frequency. I haven't memorized each part of the spectrums role, but somewhere up there, the image probably doesn't want to appear to come from everywhere. Then again, if 700hz is in the rooms modal region it will already have issues of appearing to come from "everywhere". My point is that in the modal region, many parts of the spectrum appear to come from other places than the drivers up front, anyway, so multiple subs spaced out, no harm no foul. The early arrival of direct energy is welcome, from all angles at that point.
This weekend, I wanted to know more about a multi sub setup due to what Dr. Geddes recently stated, about randomness of low frequencies. So I moved the subs to different locations instead of stacked, and I must admit that the bass quality is substantially better than with my previous setup. Eight bandpass subs spead out along the sidewalls and in the front in different heights deliver a more uniform and enveloping bass with much less resonance. Stil need to tweak but the result is encouraging.
Last edited:
I sit close and stay on axis. Rising DI will allow a wider spectrum of direct field interaction. With constant DI, less of the field Potentially, reaches you, because;
The room is custom made with hanging CLD walls, so yes, it's not "normal".
Fs at 700 Hz would be very rare, maybe a tile bathroom, but not even a car is that high - too much absorption.
Yes, at LFs steady state is all you can hear owing to the long time constants involved.
Claude got confused and admitted that it was so... I think this is the missing aspect. As directivity is increased, spl on axis increases all else equal, so an even frequency response has to also be the goal.
It's not an easy thing to measure and at LFs it would be impossible. The reason that we use this term is to highlight a concept, about how the sound field transitions from being one thing into another. It happens for all source/rooms and understanding the concept helps to visualize it as a reason that we might hear something about this field. I know of no-one who has actually measured this, except for the situation of an anechoic chamber, where they are qualified by their being no reverb field. i.e. a stationary sound level (as opposed to 16 dB/double) with distance.
Maybe experimentation with a Mic would give some clues of approximation? A measurement taken directly in front of the source, and then continuing measurements with increased distance.... Analyze the change?
A measurement taken at 1m, at a high volume vs a low one. The change has nothing to do with the direct field? Interesting.
A measurement taken at 1m, at a high volume vs a low one. The change has nothing to do with the direct field? Interesting.