Diagnosing a room with one number, volume.

[Pareto Pragmatic]

The original meaning of hacker was someone who makes furniture with an axe. Please take this thread as an example of doing something with a crude tool.

Given one quantitative interval level variable (what most would call objective) known as volume, what can you diagnose about the acoustical situation in a room?

You can use music, but no tone tests (that would give you another number). And you can use as many qualitative ordinal level measurements as you like. So, higher lower, low medium high. And you can use all of your senses. I assume hearing and touch, but if you can figure out a way to use taste and smell go for it. The close cases you won't be able to call, but for larger differences you will be able to. You give up accuracy in measurement, but it is still measurement. It will be valid (touch measures vibration for sure) and reliable (if you can feel vibration level change at a 2dB change in volume, that will be pretty constant in a test/retest check of reliability).

You can, and should, link the ordinal sense data to the volume level. Better is if you can link touch (vibrations) to hearing (vibrations) to volume. This will help you calibrate your sense data to a number.

Volume is in numbers, and 50-60 on the dial is a 10db change. You can't know the actual volume at any spot, but you can know how much the volume changes.

You can, if you wish, have another person make the same judgements. This would introduce some inter-rater reliability in your measurements. Which is nice, for validity purposes.

BTW, if you actually want to try this, it would be best to set something up in a room you have not previously measured.

I'll give one example, and hold my best example in reserve until people have a chance to chime in, should they want to.

Play Dark Side of the Moon, the start, the heart beat. Put your hand on the floor. Turn the volume up, over repeated plays. At some volume the floor will keep vibrating between heart beats, the space between beats will disappear. That volume will cause problems with bass, and for complex music likely the problem will start a bit lower in volume. Does this find the exact frequency, or level? No. Does it identify a problem in that room that starts as volume gets turned up? Yes.

You could then make some changes, like better isolating your speakers. Do the test again. Does the sustain point happen at a higher volume? Lower? The same? If the change is minor, you won't be able to tell, but if the change produces a larger effect, these measurements will surely pick it up.

What can be done with this axe? How far can a person get?

 

[Pareto Pragmatic]

I love measurement theory, and was hoping to find others not just into measuring, but the logic behind it. Like how do we make more precise instruments than we currently have? I'm still learning this place, and I seem to have been a bit off target in my approach. So let me try again, 2 quickies and one with a bit more detail.

Back in the day, this is how I would diagnose a room. Play music, turn it up until I lose "fingers on the bass strings" and the vocals started to get "veiled". That was from the room. Change things (placement, isolation, treatment, tone/eq) until I can turn it up as high as possible without losing bass and vocal detail happening to a level I notice/care about. Done.

Today, if I were going to use this method, I would use my AVR. Room correct, compare direct/no correction to corrected. I don't need to know what my AVR is doing in detail, but mostly bass stuff of course. Compare. Adjust so direct sounds more like corrected. Recalibrate AVR. Repeat as desired.

Both of those lead to better sound. Today I measure using SPL and frequency response. Measure at the listening position, because that's what matters. And I measure a lot when setting things up. Because experimental methods work pretty well.

I found I have a separate floor mode at 129hz. It is a floor mode, as it was a 65hz mode before I moved my seating area forward. I have a 100 year old VERY bouncy floor. At some point, it goes from vibration to bouncing. At that point, 129 starts to rise up, and destroys my EQ effectiveness, so I have to use a more aggressive EQ. This allows me to get louder, but not quite loud enough for where I would like to be at times. My EQ will fail at some point, since I can only push it so far before creating other problems. I do plan on getting DSP so I can use parametric EQ, and change out my budget down ported sub for a better sealed unit next summer. A thicker rug pad, better speaker isolation has helped already.

One day I was behind my seating position, and noticed the floor was vibrating, surface of feet level, which is where things still sound good. So I went to sit down, and noticed the floor in front of my seat was not just giving me vibrations on the surface of my foot, they were coming up about 1/4 the way INTO the bones of my foot. I tried other places. Then I turned it up 3dB, and now the floor in front was more like 1/3 the way into the bones, behind the seat it was about 1/4. I asked my wife to see what she noticed (without telling her more than to feel the vibrations), to get some inter-rater reliability, and she felt the same things. And THAT let me get the last piece I needed to understand my room.

I don't have a floor in the abstract, I have a lot of layers that make my floor, as most real floors are. 2x6 diagonally laid boards, under maple hardwood, under vapor barrier pad, under floating engineered hardwood, under rug pad and rugs. I have added support under the seating area across 3 joists, NOT structural support, just to keep things from sagging more over time under heavy furniture and people. So, what seems to be happening is this.

The floor vibrates. Then starts to bounce (which is where I ended up my analysis before), but it bounces in front of the seating area first. The couch and post under seem to be creating a pinch point stopping that bouncing from propagating further. A few dB higher, and the bounce JUMPS to the back and the rest of the floor is energized. That's when my EQs have been failing hard.

I am pretty sure the top layer of the floor, the floating engineered floor, is starting to move independently from the rest of the floor at higher SPL.

Would I have found this with tone sweeps? Unlikely, that's not going to energize the floor enough. I would need multitone tests at different locations at different volumes to find this effect. If I knew to look, sure, I could do that. But I already "knew" what was going on, bouncy floor is bouncy. Using feet as a qualitative measure of vibration showed the dynamics of my floor pretty clearly, and that lined up with what I had experienced of EQ collapsing at a certain point.

More accurate measures are good, don't get me wrong. I'll take more accurate every time. But analysis is more than measurement, and if your analytic method is rigorous and well calibrated, you can get a long way with "lesser" measures to figure things out and make improvements in an effective way (but likely not optimal way). You can have great measures, but if you don't get your method right you might just end up with the wrong conclusion. Like I did.

So that's how I figured out my floor was not just acting as a passive radiator tuned to 129, but a two stage passive radiator.

 

[fpitas]

Back in the day, this is how I would diagnose a room. Play music, turn it up until I lose "fingers on the bass strings" and the vocals started to get "veiled". That was from the room. Change things (placement, isolation, treatment, tone/eq) until I can turn it up as high as possible without losing bass and vocal detail happening to a level I notice/care about. Done.

I certainly hear that happen as I get things closer to perfect.

 

[RayDunzl]

Concrete Slabs are the norm here in Florida. The Texas house was on a slab, too.

They are, to my senses, devoid of vibratory dismay.

Had a basement as a kid in the early sixties, but Dad's amp was only 30W or so.

 

[Pareto Pragmatic]

Concrete Slabs are the norm here in Florida. The Texas house was on a slab, too.

They are, to my senses, devoid of vibratory dismay.

There have been times I have longed for a slab! It makes it a lot easier is a lot of ways, at least it did for me when we lived in a place on a slab.

I would guess your ceiling will be the room surface that loads up first in that situation. Not that I suggest a ladder and repeated measures to find out!

 

[Philbo King]

The original meaning of hacker was someone who makes furniture with an axe. Please take this thread as an example of doing something with a crude tool.

Given one quantitative interval level variable (what most would call objective) known as volume, what can you diagnose about the acoustical situation in a room?

You can use music, but no tone tests (that would give you another number). And you can use as many qualitative ordinal level measurements as you like. So, higher lower, low medium high. And you can use all of your senses. I assume hearing and touch, but if you can figure out a way to use taste and smell go for it. The close cases you won't be able to call, but for larger differences you will be able to. You give up accuracy in measurement, but it is still measurement. It will be valid (touch measures vibration for sure) and reliable (if you can feel vibration level change at a 2dB change in volume, that will be pretty constant in a test/retest check of reliability).

You can, and should, link the ordinal sense data to the volume level. Better is if you can link touch (vibrations) to hearing (vibrations) to volume. This will help you calibrate your sense data to a number.

Volume is in numbers, and 50-60 on the dial is a 10db change. You can't know the actual volume at any spot, but you can know how much the volume changes.

You can, if you wish, have another person make the same judgements. This would introduce some inter-rater reliability in your measurements. Which is nice, for validity purposes.

BTW, if you actually want to try this, it would be best to set something up in a room you have not previously measured.

I'll give one example, and hold my best example in reserve until people have a chance to chime in, should they want to.

Play Dark Side of the Moon, the start, the heart beat. Put your hand on the floor. Turn the volume up, over repeated plays. At some volume the floor will keep vibrating between heart beats, the space between beats will disappear. That volume will cause problems with bass, and for complex music likely the problem will start a bit lower in volume. Does this find the exact frequency, or level? No. Does it identify a problem in that room that starts as volume gets turned up? Yes.

You could then make some changes, like better isolating your speakers. Do the test again. Does the sustain point happen at a higher volume? Lower? The same? If the change is minor, you won't be able to tell, but if the change produces a larger effect, these measurements will surely pick it up.

What can be done with this axe? How far can a person get?

This will give you either 'too loud' or 'too quiet' for one particular point in the room, but not much else.

The problem is that sound in a room is multivariate. For each frequency and each 3d point, standing waves develop (at least at lower than Schroeder freq) and volume will vary with listening location and frequency.

 

[tmtomh]

If you are concerned about a single central listening position aka a sweet spot, I would imagine this kind of measurement technique could be somewhat instructive - I suppose you'd be best served by sitting in the main listening position with bare feet so as to best feel the floor vibrations. And if you wanted to feel wall vibrations, you'd have to rig up some kind of rigid but lightweight stick or pole so you could sit in the listening position and hold the pole horizontally so it was touching the wall.

In such a scenario I suppose you could also maybe detect some midrange or treble resonances or peaks, if certain musical material started to sound "boxy" (midrange/lower midrange peaking) or piercing (peak likely in the 2-5k range).

The main difficulty I would anticipate from this kind of method is that since you don't know the frequency of the issues with much precision, you can't apply EQ, and so you're left with repositioning things as you say, which is just as likely to create new or additional problems as it is to solve the ones you heard during your first test. So it might become a whack-a-mole situation.

 

[Pareto Pragmatic]

The main difficulty I would anticipate from this kind of method is that since you don't know the frequency of the issues with much precision, you can't apply EQ, and so you're left with repositioning things as you say, which is just as likely to create new or additional problems as it is to solve the ones you heard during your first test. So it might become a whack-a-mole situation.

I think you can apply mild EQ and not run into too many problems most of the time, 1-2db. But you are right there are additional problems that crop up. Too far on the 129 and I start to get suck out right above that range. I won't hear that start, but will hear it when it is -2dB too low.

Add up a bunch of 1dB errors, and you can end up with a mess.

 

[audiofooled]

Better, sealed sub can't do no harm. Also decoupling the loudspeakers from the floor, as a way of scattering of the energy which comes as a feedback from the floor bounce and introduces distortion, reducing it to where you can handle it better using EQ.

Suggestion For Amir Youtube Video - Testing Vibration Effects On Components!

https://www.amazon.com/YOKELE-Vibration-Plate-Exercise-Machine/dp/B0BV1Q824J?th=1 Big and strong enough to put one of those 90s Krell amps on And Amir can use it on his feet when typing his reviews :facepalm:

audiosciencereview.com

 

[Pareto Pragmatic]

Better, sealed sub can't do no harm.

I put an SVS sb1000 pro in, and it does help. It took out the slow run up of the floor at lower volumes, and now I am left with the floor activating at high room SPL. I need to spend today integrating it fully, so far I have only auto-set up my AVR with it. But as of now, I am getting a lot cleaner sound for a lot more of my listening range.

The old one is now in my office, and it lets me high pass to my tiny speakers. MUCH better office sound now. Which leads me to an example of how volume knob tuning can be useful for getting others to learn they have control over their sound in their rooms.

There's a guy at work, and we talked music. When I brought a small system down, he listened and was blown away (sound for price terms). Then I brought other speakers down, and after listening, he bought a small system for his office.

This person is very analytical, but not at all quantitative, not a numbers person. Think philosopher. Because that's what he is.

I asked him if I could check something to compare the acoustics of my office to his. So I put on the first track of Mingus Mingus Mingus Mingus Mingus, and did the bass analysis as described above. I turned up the sound and he listened along with me. I talked about tone, I asked and then changed his speaker tow. He heard differences. I pulled down some modes with my phone eq, he heard differences.

Then I showed him a phone based app, 20 band eq, shows spikes and dips at specific frequencies. I pointed out the shape of the curve, explained what a base boost from the room meant for sound, etc. Then I showed him an eq on my phone, 10 band. And explained "look at the numbers spiking, if we can find an eq that is close, we can deal with that specifically."

Pictures and number matching, he had no problem with that. So I told him about audio mixing frequency charts. Like this one: https://s3.amazonaws.com/ssipublic/musicfrequencycheatsheet.pdf That kind of chart puts numbers to words that people apply to music. That made sense to him.

Basically, starting with volume changes, I explained to this person everything they needed to know to take control over their sound. And I told them that. He was very appreciative. Interested. Will he start playing around? If he does, he now knows that he understands everything he needs to know to use very good tools, but by using his ears and some simple tools he can take control.

So while volume ALONE is not going to get us very far, it is a good way to lead into the idea of room reinforcement, corrections, and taking control over the sound in your room. And it can be done on the spur of the moment, with a phone.