Kvålsvoll Design Audio Forum

EnglishEnglish text EN    NorskNorsk text NO
Please or Register to create posts and topics.

Room Acoustics / Romakustikk

Now: A concept for rooms that actually works

Fast forward to now, after experimenting with different solutions for acoustics treatment, in combination with working on the loudspeaker systems, it is apparent that some things work, some things are stupid, and there are indeed universally applicable metrics for acoustic performance.

Acoustic performance can be measured, and put into quantifiable numbers to be used as a specification. When the room measures according to those metrics, it will sound good.

And proper acoustic treatment can bring any room to perform in-line with those metrics.

So we have two stages of development of the solution; first, how it should be acoustically, end then how to practically get this performance in the room.

One important realization is that the decay attenuation must be much faster and stronger than usually seen recommended. Early decay must be fast and large. Then the decay profile of the room must be fairly flat across the frequency range, so that high frequencies have a similar rate of decay compared to lower frequencies.

Then it is desirable to preserve later reverb, so you get a decay profile vs time that attenuates quite fast early on, and then slow down later in time.

This means the room must have a lot of absorption, while some reflections must still be preserved. This is achieved by installing absorption with reflective surfaces on most surfaces of the room, and finally tuning with diffusors.

More pictures from the post #6 room. Poly diffusors on side walls and ceiling. I was a bit skeptic about all the black color, but it turned out very nice.

Now a good room acoustically for sound, very nice both for performance and visual impression, because everything was planned and there is a consistency in the overall design.

The customer with the Mini-Acoustics Project room had 2x T140 installed, and during the calibration process, I noticed he had fixed the room already, and that it worked very well, a bit towards dead for music perhaps, overall quite good. How can I tell?

I look at decay, with 20ms slice interval, 1ms rise time. How much does the sound drop in level during first 20ms:

Around 20dB down at 20ms, and that is very good.

Then I look at frequency response slope of the decay lines, want a flat, smooth response that does not rise much in the low mids, and does not drop off in the highs:

Reasonably good. Now this is a quite small room, so some deviations from flat is to be expected. Note that the first line, the first direct sound arrival, does not necessarily show a correct frequency response due to to the 1ms rise time, because it may be too short.

Then I look at the decay in low mid in approx. 100Hz - 1KHz range, to see if all resonances are gone and decay is controlled and does not rise:

A room with insufficient damping at lower frequencies tend to show an increase in decay towards low mid frequencies.

Other graphs can be nice to look at, but they do not necessarily show much more information about acoustics that already verified in the decay analysis.

GD- group delay, note scale 120ms, NOT 1.0s or something similar that would only show a perfectly flat line:

IR, impulse response, is popular to look into, tells something about reflections towards higher frequencies:

IR: Good rooms tend to have a fast, huge initial drop, and then a bit later reflection level rises, before it continues to drop.

Spectrogram is a nice way to visualize what happens in time and also see room decay frequency distribution, note scale time range is 100ms NOT 1s or even more, scale range is 40dB:

RT60 measures like this, and tells nothing about the acoustic performance of this room, so please stop using RT60 numbers as specification or performance indicator for acoustic performance of small rooms:

"..  I realized that you can get the same thing with acoustic treatment."

This visitor was interested in Bacch sound processing. Experiencing the sound in a properly treated room made him realize that good acoustics gives similar improvements as the Bacch spatializer is supposed to do. And it does. The overall presentation is larger, much more immersive, with envelopment, and instruments rendered with solidity and presence and separation.

Is it the same, as this presumably very excellent sound processor? No. The Bacch is a spatializer using cross-talk elimination, implemented in dsp. It will sound different, and can still be used in a treated room. But the improvements to sound presentation with regards to holographic rendering is similar.

A widely accepted misconception is that a treated room will be too dry with too little contribution from room reflections, and thus will create a sound scene with no envelopment, you loose the sense of being inside the acoustic venue of the recording. But reality is quite the opposite; a good treated room will create a more immersive experience, where the sound fills the listening room .

 

Membrane or Helmholtz bass absorbers useful?

Room examples shown does not show any membrane or Helmholtz absorbers, does that mean they are not useful, or do they not provide any advantage over velocity-based absorbers?

Depends. A membrane bass absorber can indeed give better absorption, at low frequencies, over a limited frequency range, compared to similar sized velocity absorption. Where it is mainly the depth that is different, say a 10cm deep membrane absorber can be made to absorb at 40Hz, where 10cm velocity absorption will not do much at all.

I do not know how well this can work, but I can find out. Set up a simulation, put in a membrane absorber, see what happens.

As explained before, efficient bass absorption is impractical and in many cases not needed. This is because typical walls already have some bass absorption, they flex. It will not be perfect, still modes and resonances and cancellations, but equalization can bring most rooms up to a quite reasonable performance. But concrete walls can be problematic, with very high q on resonances and modes. And it does not improve that much with inner walls, because the wall panels are too light in combination with the small amount of free space behind, to be able to stop the bass. Then other means, such as membrane based absorbers, can be a viable alternative.

Let me know if there is any interest in doing a simulation on this, that could give some useful information.