Adjusting Room sound using material damping methods (not DSP)

[A.S.]

...The gunshot is described as a "clean, characterless click or crack" but that does not mean it is without pitch does it?

There is no character at all in the click, as generated. The click is an impulse, and certainly that impulse contains all frequencies in equal proportion right across the audio band. It is as pure and characterless as truly random white noise, the type of static noise you hear between stations on an FM radio.

Therefore as generated*, the impulse I've used has definitely no pitch, because it has no resonance generators and to sustain a pitch (a note). You have to have some sort of resonant system generating and sustaining the pitch. (A musical instrument is nothing more or less than a resonator in the same way that an empty wine bottle is a resonator and produces a pitch when blown across).

It is of course a trivial matter to put the click through a spectrum analyser and have a look at its spectral make-up. However - and this is important to bear in mind - you are hearing the click through your sound card and speakers/headphones as you cannot sense the click directly in your brain (as yet). So your perception of the click is the digitally generated clip mixed with the resonance characteristics of your speakers. Remember the box in the box in the box idea? The clean, dry click takes on the character of the speaker box (and drive units/crossover) and is being reproduced in your listening room - another box. So it is entirely possible that to your ears, with your speakers in your room that the click has some sonic character, but this is not inherent in the click itself for sure. So the very evident twang in the reverberation tail in Clips 11, 11A and 12 is solely that of the room modifying the characterless click.

* Note: when the repetition rate of the click starts to become very frequent, that is, very many times per second, it may well take on a pitch, but this is not relevant at this point. I've shown that even when there is a gap of a second or two between the clicks, there is still a definite audible character in the reverberation.

[A.S.]

... It [the click] is as pure and characterless as truly random white noise, the type of static noise you hear between stations on an FM radio. Therefore as generated*, the impulse I've used has definitely no pitch, because it has no resonance generators and to sustain a pitch (a note). ...

The advantage of the gunshot click is that is discrete. That means that it is not continuous (like a machine gun) but occurs just every second or two, as we generate it. But whilst the energy in the click is all brought together into one instant, we can generate the same audio spectrum from white noise.

To remind ourselves here is the click again:

(Clip #9A) Dry gunshot, gain increased to 100%

And here is exactly the same frequency spectra but spread out in time:

(Clip #13) Completely random white noise (FM inter-station radio hiss)

Incidentally, we can use the pulse or the random noise interchangeably for making acoustic measurements on speakers, microphones, rooms, amplifiers CD player or whatever. Under controlled conditions, the results would be identical. But, as already noted, the click have very little acoustic energy in it so if we were making acoustic measurements with a microphone we would have to ensure that the environment was extremely quiet or even the slightest ambient noise would be picked up by the microphone and would garble the measurement. So for practical acoustic (environmental) measurements, the pulse is not really much use. White noise would be a better bet because we can push so much more power into the test, so that we can swamp ambient noise like doors shutting or cars and planes passing by.

I've taken the white noise Clip 13 and added the hall reverberation that gave us Clip 10A, 11 & 12. Now we can't hear the twang but it is definitely there buried under the random white hiss. It sounds like this now:

(Clip #14) White noise with hall reverberation added

Although we can't immediately identify it, the twang in Clip 14 is the same as in Clip 12 here:

(Clip #12) Filtered echo component of gunshot in room (gain increased)

The point to make is that the listening room may well have a serious sonic character. That may be masked by the type of music being played or it may be exposed by the music. It may irritate you immediately or it may take hours, days or weeks until your subconscious alerts your conscious to the presence of a drone from the room. But once you are sensitised to the twang, you just can't get it out of your mind and your enjoyment of music is seriously impaired.

[STHLS5]

...... the impulse I've used has definitely no pitch, because it has no resonance generators and to sustain a pitch (a note). ...

What is "pitch" that we are referring to in this subject matter? I am not sure that I really understood the meaning of pitch.

ST

{Moderator's comment: Suggest internet search for 'pitch' or try here. Pitch is the basic component of music.}

[honmanm]

Just catching up with this before bedtime - have been rediscovering old favourites via Spotify so computer speakers are at music-listening level (and those white noise clips are pretty loud). So when I hastily stopped the replay of the basic white noise clip, it was suprising (but not too much) to hear a "raaa" echo once it had stopped. BTW the gunshot doesn't excite the room in the same way.

{Moderators comment: the white noise is actually at a much lower peak level than the click. But because the white noise is continuous it seems louder.}

[A.S.]

It occurred to me overnight that there is an obvious way to expose the twang that is masked under our random white noise clip.

Here again is the white noise with hidden reverberation element with its own sonic character (you need to read all previous posts to understand why):

(Clip #14 again) White noise with hall reverberation added but buried under noise

but if we abruptly truncate (sharply switch off) the white noise generator, we can expose the hidden reverberation here:

(Clip #15) Truncated white noise with hall reverberation exposed

Now if we pass that reverberation tail through our 500-700Hz filter as we did with Clip 11 and apply some boost so we can clearly hear the reverberation we have clearly exposed the room twang from under the random noise:

(Clip #16) Filtered echo component of white noise, abruptly stopped, in room (reverberation gain increased to hear it better)

and we can compare that sound with the twang of the click when exciting the room resonance that we heard before:

(Clip #12 again) Filtered echo component of gunshot in room (gain increased)

================================================== ====================

Conclusion at this step:

1. Regardless of whether we use a single-pulse click or gun shot or continuous noise (or music) we can expose the characteristic sonic signature of the room by analysing what the microphone in the room hears
2. Regardless of the stimulus, the signature of the room has exactly the same pitch and importantly ...
3. Only one single click is required to energise the room, to set it reverberating and expose its characteristic twang

The implication of 3. is that from the very first note of the very first bar of the music leaving the speakers and radiating into the room, the room is given enough energy to set it into resonance, and once the characteristic twang is established in that first 0.1 second, every successive note will re-energise and re-stimulate the room's character until some time after the speakers have fallen silent after the final musical note. Therefore, an untreated, undamped room will add its sonic signature as a (continuous) drone over every note and every bar of the music. That is anathema to high fidelity sound reproduction. We cannot expect high fidelity sound reproduction is untreated, undamped rooms. And furthermore, it must be extremely obvious that gadgets and gizmos of small surface area proportionate to the room's surface area cannot and will not make any improvement whatsoever to the room's acoustics, and if they are bell-like in shape, material and construction may well degrade the acoustics even more at their resonant frequency.

So where do we start analysing the room and making a treatment plan to ameliorate uneven reverberation - that is, reverberation that has a strong sonic signature. And more pertinently, how can we reliably 'test' the room without any test equipment?

[STHLS5]

.....And furthermore, it must be extremely obvious that gadgets and gizmos of small surface area proportionate to the room's surface area cannot and will not make any improvement whatsoever to the room's acoustics, and if they are bell-like in shape, material and construction may well degrade the acoustics even more at their resonant frequency............So where do we start analysing the room and making a treatment plan to ameliorate uneven reverberation - that is, reverberation that has a strong sonic signature. And more pertinently, how can we reliably 'test' the room without any test equipment?

What should we use? Clapping hands, knocking a glass with a spoon, two soup spoons by knocking their bottom making a nice "tuk" tuk" sound. I used all these and including test tone CD with a single speaker.

ST

[honmanm]

So where do we start analysing the room and making a treatment plan to ameliorate uneven reverberation - that is, reverberation that has a strong sonic signature. And more pertinently, how can we reliably 'test' the room without any test equipment?

Since the reverberation is a physical effect resulting from physical properties of the room, is there some way of getting from those properties to the "problem" frequency band? (this is discounting reverberations in the structure of the room).

It may be a bit early to ask the question, but it's intriguing to think about what physical process is at play here. A first guess would be interference between room modes... but that's probably barking up the wrong tree...

[A.S.]

... Clapping hands, knocking a glass with a spoon, two soup spoons by knocking their bottom making a nice "tuk" tuk" sound....

As we discovered when listening to the impulse (the click) there really isn't much energy in it. So I'd forget about the two spoons and the wine glass. We really need to put more power into our test impulse to be able to hear anything significant in the echo. So as you suggest, clapped hands are a better option.

But exactly how to clap the hands? We want to generate as clean, sharp characterless crack as we can. We can try two approaches: Generating a clap as we would when applauding, whereby we bring together the cavities of our hands into a clasp and that produces a lot of low-frequency energy as the air is squeezed out from the void formed by the palms. Alternatively, we can keep the two palms flat and parallel and bring them smartly together. The sound will be different so the acoustic spectra will be different. But for sure, it's far less painful to bring the cupped palms together than suffer the sting from the underside of the fingers brought together! Either way, we need to repeat the clap as loud as we can about once every two or three seconds, and we need to walk around the room carefully listening to the echo and defocusing on anything in the room. Our total attention should be to the character and direction of the reverberation.

So the next step is to actually do some walk-clap-listening (WCL) in your listening room (or in any building) and to build-up some expertise for yourselves as to how different rooms respond to the WCL test. At first, don't expect to be able to hear much in character in the echo - but with a little more experience and comparing the sound in different parts of the room, you'll really see begin to appreciate how very influential the room acoustics are to the reproduction quality of high fidelity sound at home. If you want a quick comparison of extremes, try the WCL test in your bedroom and then your bathroom. And just what factors make these to extremes behave as they do acoustically?

The listening room's acoustics is the hidden devil in the hi-fi reproduction chain. Unless this devil is tamed, the listener is just wasting his money on ever more sophisticated equipment.

[HUG-1]

We are looking forward to some input from the HUG membership concerning your room clap-test.

[A.S.]

I'm going to have to assume that you're with me on this and move on. I think the audio examples are clear enough.

We've looked at the general reverberation in the concert hall and/or listening room. We've seen that if the general acoustic takes on a sonic character, that that character overlays reproduced sound in the room. We've seen from the white noise (random noise) clips that even when the room drone is quite marked, we may not be able to hear it consciously. We may only latch onto that acoustic issue when the music abruptly ceases or between the notes under ideal conditions. We can be absolutely certain that the room drone is not part of the composer's score, nor is it a characteristic of the speaker, and to spend serious money on hi-fi equipment just to to dump it into an untreated listening room and expect great sound is just daft. Great painters need natural, untainted daylight to reproduce reality: great audio needs natural, clean acoustics in which to perform.

Aside from the drone or honk or twang we've heard in the room's reverberation there are other listening room (or studio) acoustic issues which are at least as serious, and may be even more irritating. One of these is sonic bounce of flutter. Sound waves behave just like a tennis ball, and given just the right circumstances they will efficiently bounce backwards and forwards between (generally opposing) walls. With each bounce they'll lose a little energy (just as the ball would) until finally they are too weak to make the final journey and the echo ceases. But the sonic character of the bounce is really irritating because it takes on a hardness of tone which may be unrelated to the music. So it is more audible.

Here is the gun shot outside again:

[Clip #9A again] Dry gun shot

And next is the gun shot (click) inside and where it excites a bounce echo, probably between two opposite walls. Note, I have not increased the echoes level relative to the click and you can hear that the echo is really prominent, and has the same sort of sharp wide-band tonality as the gun itself. Fatiguing to listen to.

[Clip #17] Gun shot inside, bounce echo

First the music as before, recorded dry, in the anechoic chamber:

[Clip #4 again] (Dry music, recorded in dead room)

Now we add the same bounce echo used to make Clip 17 to the music and we have this:

[Clip #18] (Bounce echo added to dry music)

There are three things to listen for in Clip #18.

• The bounce echo make the string tone brighter and more toppy through
• In the middle of the clip, between the beats, you can hear a distinct echo as if there are a greater number of string players, but slightly out of step
• The final seconds have a definite wah-wah modulation effect and a 'dwaaannnggg' due to an unwelcome coincidence between the tempo and the timing of the echo and its character.

All in all, a horrible, fatiguing sound fairly typical (although this is extreme) of home listening rooms.

Now we can compare Clip 18 above with its bounce echo with Clip 19, where I've only added the previous hall echo what gave the coloration from clip 10 onwards.

[Clip #19] (Hall echo added to dry music)

What sounds worst? Clip 18 or Clip 19? Clip 19 is of the music played in the great hall or Clip 18's bounce echo of the type that we'd find in a typical domestic room? I'd say Clip 18 was far more fatiguing to listen to. At least we are generally familiar with and untroubled by the (relatively) long reverberation of large halls. But strong, intense echoes rapidly following the notes are disturbing.

[A.S.]

...Aside from the drone or honk or twang we've heard in the room's reverberation there are other listening room (or studio) acoustic issues which are at least as serious, and may be even more irritating. One of these is sonic bounce of flutter. Sound waves behave just like a tennis ball, and given just the right circumstances they will efficiently bounce backwards and forwards between (generally opposing) walls. With each bounce they'll lose a little energy (just as the ball would)...

Another thread has started recently which is covering part of the same general subject of room acoustics that we are discussing here, specifically the Left-Right balance as perceived in the listening seat at the sweet spot. The appreciation of room acoustics as I'm explaining in this thread (will) completely explain why the balance of sound could indeed seem to shift left or right and once that's understood yield the inescapable remedial conclusion. It may be more fun to get hands-on but personally, I'm at an age now where if I can think through a solution without moving from my seat, I'd much rather do so.

So, assuming that we are all travelling together on this room reflection issue, and assuming that you have listened to the previous sound clips, we should be in agreement that there are two (obviously) interlinked echo characteristics in the listening room:

• Frequency related room characteristics such as a note or notes (a drone) which adds energy in one or more frequency bands but probably not across the wide audio band (example Clip 12) and/or
• Time related room characteristics where an echo results in a mirror image of the wideband audio sound being splattered around the room (example Clip 17) and can effect many frequencies or all

Time and frequency are two sides of the same coin and are intimately connected, so these two are indivisible, and classifying an echo as only a time or frequency issue doesn't make sense. So we are going to have to sidestep that academically and just look at how in the real world listening rooms these two characteristic echoes manifest themselves.

To save me the trouble of dreaming up a room, can anyone upload a line drawing of their listening room viewed from above showing the position of their furniture and speakers with approx. dimensions?

[STHLS5]

...To save me the trouble of dreaming up a room, can anyone upload a line drawing of their listening room viewed from above showing the position of their furniture and speakers with approx. dimensions?

Sorry, I am still experimenting with WCL, meanwhile I have attached my room plan which is too basic. Anyway...

Room dimension is about 370cm x 440cm.

Speakers (tweeter) to front wall = 116cm, Speaker to side wall = 102cm. Sitting position from rear wall from head to wall = 106cm. The sketch did not include other room treatment materials and two small CD rack next to the speakers. There are no windows, except for two wooden door one left and right.

ST

[A.S.]

Excellent. Many thanks. It's better to work from a real-world room rather than me draw an entirely imaginary one. I need to 'scan' your drawing to a CAD program - can you confirm how it is scaled? As a cross check, can you tell me what size you drew the speaker boxes at and what the scale factor is to real life? Ditto the room itself. With this plot and your confirmation of scales, I'm limbering up to move on to some simulations of the situation in your room over Christmas.

[STHLS5]

.......I need to 'scan' your drawing to a CAD program - can you confirm how it is scaled? As a cross check, can you tell me what size you drew the speaker boxes at and what the scale factor is to real life? ......As a matter of interest, have you tried the hand clap test throughout the room? Are there any zones where you can hear anything 'in' the echo as it dies away?

I am reattaching the plan drawn to scale. The scale is 1:27.54. Speaker height is 107cm. Room height is 286cm. The table is flat wooden board with the PC LCD monitor. And a small bare wooden chair (not stool). One side table that I forgot to add.

OK, a Merry Christmas to all.


ST

[Sebastien]

... The appreciation of room acoustics as I'm explaining in this thread (will) completely explain why the balance of sound could indeed seem to shift left or right and once that's understood yield the inescapable remedial conclusion.

I'll look forward on this subject since I like to improve the sound quality in my room. This is my project for 2011.

Sebastien

[A.S.]

I am reattaching the plan drawn to scale. The scale is 1:27.54.

Thanks. I've imported your PDF via another program into AutoCAD and confirm the scaling as 1:2754. That means, from the speaker baffle to your ear, seated, is about 2.0m.

Now a thought to leave you with about echoes. Suppose you stand with your nose literally touching a wide, tall brick wall, perhaps the side wall of an apartment block, sports centre or town hall. You clap your hands. What do you hear? You step back 5m, still facing the wall. You clap again. What do you hear now? And another 5m back. And another until you are at least 15m or more mtrs. away from the wall. Let us know what you discover; listen very carefully to the echo.

[STHLS5]

Thanks. I've imported your PDF via another program into AutoCAD and confirm the scaling as 1:2754. That means, from the speaker baffle to your ear, seated, is about 2.0m.

It is 2.36m. I have re-measured and the PDF scales is accurate with a margin of error plus minus 3cm though now I realised that I missed out the carpet and the equipment racks.

Now a thought to leave you with about echoes. Suppose you stand with your nose literally touching a wide, tall brick wall, perhaps the side wall of an apartment block, sports centre or town hall. You clap your hands. What do you hear? You step back 5m, still facing the wall. You clap again. What do you hear now? And another 5m back. And another until you are at least 15m or more mtrs. away from the wall. Let us know what you discover; listen very carefully to the echo.

What I hear in the emergency stair case (a perfect echo chamber) of my building is there will be a lingering "twang" or "mmmmmm" or slap when you shout or clap near the wall. The further you are the 'mmmmm" becomes distinct, i.e you hear the second and third sound of the original. I mean echo.

I am glad I took this topic seriously even though I was thinking I know enough of room acoustics to set up my system. Your slow and methodically teaching now taught me look at echoes and reverberation more critically and it teaches me distinguish pure sound and coloured sound and to know what causes the colouring. It would be nice if more could join in to provide feedbacks.

ST

{Moderator's comment: It does feel very lonely here sometimes with the paucity of feedback. I'll leave Alan to comment to your echoes.}

[A.S.]

...What I hear ... The further you are the 'mmmmm" becomes distinct, i.e you hear the second and third sound of the original. I mean echo. ...

Excellent observation! Couldn't have asked for a better response. Perhaps you don't appreciate yet just how significant your discovery is. Let me repeat what you said in slightly different language ...

'When you are really close to the wall, you cannot hear the echo even though you know for certain that there is an echo being created. As you move back from the wall, you can begin to hear the echo (or echoes) as a separate, distinct sound source in addition to the direct sound from your hand clap to your ear.'

Did you know that there is a name for this phenomena, named after the researcher who first discovered it? I attended a presentation concerning the King's Cross Station fire at which there was tremendous loss of life. One factor was the unfortunate communication difficulties with the public due to the hard tiled-wall acoustics underground (easy to clean) and the PA system. So this issue of fidelity in the presence of echoes is not just a scientific curiosity - it has great importance in the design of safe public spaces.

I'll slightly re-scale your existing room plot. No greater precision is required as any conclusions we draw cannot be better than educated guesses. We don't know the precise nature of your absorptive surfaces (walls, floor etc.) so we can never do better than a crude approximation to the real world. Even so, we may be able to tease out something useful, and applicable to all listening rooms with a little more work.

Next we need to have a look at the nature of human sound perception relating to echoes.

[Pluto]

This phenomenon is called the Haas effect which is well known by public address engineers everywhere, or should be! What this quantifies, in principle, is the inability of the human ear to separate two similar sounds if they arrive within a short time of each other (if I remember correctly, up to about 40mS).

When digital delay systems started becoming affordable, this principle had a significant impact on the intelligibility of the sound reinforcement in some, suitable, venues. At some audience seats it became possible to time-align the arrival of the reinforced sound (electronic transmission) with the residual stage sound (acoustic transmission) which could create a massive improvement in intelligibility, and in turn meant that the system could be run rather quieter than was otherwise possible, with obvious advantages. However, setting up such delay systems is not for the feint-hearted or unskilled novice, for it can truly open a huge can of worms. Think about what might happen should a residual of the delayed, reproduced sound be audible on the stage by the performers!

[Pluto]

Actually, having read the foregoing postings a bit more carefully, I wonder if you are actually describing the Precedence effect, of which the Haas effect is a special case.