Hello fellow scientists!
I've got a bit of a problem with noise in my listening/measurement room.
From 100 Hz up, I have about 60-70 dB range between noise floor and finger snapping at 0.5m. Not sure if my usb mic is the limiting factor here, or if something else is...
Below 100 Hz, we have a measurable bit of noise around 60Hz and 30Hz, as well as audible intermodulation of the noise around these 2 frequencies with about a 0.5 Hz "fundamental". I still have about 40dB of range down here, and when music is playing, it's completely inaudible.
However, sometimes I like to sit quietly between songs and just relax. The room is so quiet that the tiny amount of bass feels like it is overwhelming my entire sense of sound, and the slow intermodular components are a bit nauseating.
At the PE tent sale I grabbed 300 5.25" woofers, and I would like to use them to cancel noise below 100Hz inside my room.
I'm rather familiar with noise canceling techniques and I figure I have two options:
1. "Passive" noise canceling with discrete circuitry for each location to be canceled (like this: Analog Noise Cancelling Headphones: 10 Steps )
2. "Active" noise canceling with a multichannel input/output from a computer with a predictive algorithm
I like using MAXMSP, and I shouldn't have a problem building the system around sound cards I already have... I figure 15-20 channels input with 15-20 zones output for bass should be enough.
However, I would like the system to be as low maintenance as possible, and a computer is certainly not the simplest path forwards.
I am proficient with Sigma Studio, and I can implement acoustic echo cancellation if anyone has a guess at to whether that would work, but I personally think that it wouldn't.
Thoughts? Criticisms? Snide remarks for the good of the cause? Let me know below!
bob
I've got a bit of a problem with noise in my listening/measurement room.
From 100 Hz up, I have about 60-70 dB range between noise floor and finger snapping at 0.5m. Not sure if my usb mic is the limiting factor here, or if something else is...
Below 100 Hz, we have a measurable bit of noise around 60Hz and 30Hz, as well as audible intermodulation of the noise around these 2 frequencies with about a 0.5 Hz "fundamental". I still have about 40dB of range down here, and when music is playing, it's completely inaudible.
However, sometimes I like to sit quietly between songs and just relax. The room is so quiet that the tiny amount of bass feels like it is overwhelming my entire sense of sound, and the slow intermodular components are a bit nauseating.
At the PE tent sale I grabbed 300 5.25" woofers, and I would like to use them to cancel noise below 100Hz inside my room.
I'm rather familiar with noise canceling techniques and I figure I have two options:
1. "Passive" noise canceling with discrete circuitry for each location to be canceled (like this: Analog Noise Cancelling Headphones: 10 Steps )
2. "Active" noise canceling with a multichannel input/output from a computer with a predictive algorithm
I like using MAXMSP, and I shouldn't have a problem building the system around sound cards I already have... I figure 15-20 channels input with 15-20 zones output for bass should be enough.
However, I would like the system to be as low maintenance as possible, and a computer is certainly not the simplest path forwards.
I am proficient with Sigma Studio, and I can implement acoustic echo cancellation if anyone has a guess at to whether that would work, but I personally think that it wouldn't.
Thoughts? Criticisms? Snide remarks for the good of the cause? Let me know below!
bob
Last edited:
Snide remark #2: 300 5.25" woofers, sounds like a weak moment.
Concern: how do you plan on targeting the noise only, so it will not interfere with music reproduction? If the ac unit is located centrale and distributed by a duct system, it would be manageable and there are solutions for that kind of problem. If the ac unit is located in the room, how can you record it properly?
Concern: how do you plan on targeting the noise only, so it will not interfere with music reproduction? If the ac unit is located centrale and distributed by a duct system, it would be manageable and there are solutions for that kind of problem. If the ac unit is located in the room, how can you record it properly?
Could have sworn I already replied.... Dang it
Yeah the woofers were a steal, but PE will certainly get their money from me on the amplification, etc.
Noise vs Music... It would "sink" both unless I have an active system with programmed noise management as a semi-static output.
I've attached a plot of the noise in my room taken with a calibrated mic, as well as music (Norah Jones) at a comfortable listening level (1m measured)
I tried using a friend's Bag End E-Trap, but the noise is a bit too quiet to register for the E-Trap, and I got feedback before I could control the AC noise...
Yeah the woofers were a steal, but PE will certainly get their money from me on the amplification, etc.
Noise vs Music... It would "sink" both unless I have an active system with programmed noise management as a semi-static output.
I've attached a plot of the noise in my room taken with a calibrated mic, as well as music (Norah Jones) at a comfortable listening level (1m measured)
An externally hosted image should be here but it was not working when we last tested it.
I tried using a friend's Bag End E-Trap, but the noise is a bit too quiet to register for the E-Trap, and I got feedback before I could control the AC noise...
Music vs Noise — imgbb.com
^^ Link to photo - it's not showing for me in my post above
Edit - here is a smoothed version: 1 48th smoothed — imgbb.com
And here is the Umik-1 Noise floor as a mic - https://www.minidsp.com/media/kunena/attachments/670/umik-101.jpg
^^ Link to photo - it's not showing for me in my post above
Edit - here is a smoothed version: 1 48th smoothed — imgbb.com
And here is the Umik-1 Noise floor as a mic - https://www.minidsp.com/media/kunena/attachments/670/umik-101.jpg
Last edited:
Bob,
The unsmoothed version shows that the pipe resonances are very narrow band, which means they have a good chance of being killed without affecting music much at all.
The problem will be the phase and amplitude of those narrow peaks will be different all over the room, cancelling other than at the listening position would require an overly complex multiple mic multiple amp multiple inverter set up.
If you are "lucky", the pipe resonances are loudest at the plenum- try measuring there and see if there is an inverse distance gain in level compared to where you measured before.
If the plenum exit is the "hot spot", the inverse phase sub could be located there, drivers surrounding the plenum - "down, down, down, that ring of fire..." (apologies to Johnny Cash), otherwise you might have some success with "Texas Headphones", (inverse phase transducers either side of your head).
Art
you could build an ADAU1701 with some TPA3116 on a pcb tot fit inside the sub. just need a mike amp to pick up a reference, using a low cost mems analog mike.
not sure if the 1701 has the memory and flops to do the work. some nc algorithms in sigma studio were greyed out for the 1701. but that is one of the cheaper solutions ( around usd 20 per board, 2x 50W. w/o supply. a 19V laptop supply would most likely provide enough power for various channels
I once played with the idea of silencing a fan duct leaking sound from a live music bar. treating the duct, and even adding a silencing chamber in front ofd the duct helps, except for the lowest frequencies, and that is where active noise cancelling comes in, where the driver could be placed in that same silencing chamber near the exit duct. and even thought of adding an vertical array of drivers to the room between the two door serial entrance, that is obligatory in many music bars now.
if you need less than 2W, the WM8281 packs all the stuff you need, but programming may not be as easy as sigma.
not sure if the 1701 has the memory and flops to do the work. some nc algorithms in sigma studio were greyed out for the 1701. but that is one of the cheaper solutions ( around usd 20 per board, 2x 50W. w/o supply. a 19V laptop supply would most likely provide enough power for various channels
I once played with the idea of silencing a fan duct leaking sound from a live music bar. treating the duct, and even adding a silencing chamber in front ofd the duct helps, except for the lowest frequencies, and that is where active noise cancelling comes in, where the driver could be placed in that same silencing chamber near the exit duct. and even thought of adding an vertical array of drivers to the room between the two door serial entrance, that is obligatory in many music bars now.
if you need less than 2W, the WM8281 packs all the stuff you need, but programming may not be as easy as sigma.
I should have added more mass to the drywall when I had the chance. It's quite too late now though.
The current wall layout is like this:
Drywall
Corning Pressed Fiberglass 3/4"
1/2" acoustic foam decoupler medium compliance
Corning Pressed Fiberglass 3/4"
1/4" acoustic foam decoupler high compliance
Mass-loaded Vinyl 1/4", 2lb per sq ft
3" acoustic foam barrier low compliance
1/2" sound diffuser NRC 1.0
While there was a HUGE difference before/after installation, with about 42dB reduction in broadband noise from 200Hz to 20kHz, the low frequency noise, which had been masked before, is now audible.
I'm sure I didn't do a perfect job installing all of the absorbers, but they do an exceptional job... I have never seen such clean wavelets from clapping in a room.
The noise in the room is very different depending on location, and it is nearly square so modes are a total mess. The 120hz spike I believe to be driven by room modes. Not sure to be honest.
The current wall layout is like this:
Drywall
Corning Pressed Fiberglass 3/4"
1/2" acoustic foam decoupler medium compliance
Corning Pressed Fiberglass 3/4"
1/4" acoustic foam decoupler high compliance
Mass-loaded Vinyl 1/4", 2lb per sq ft
3" acoustic foam barrier low compliance
1/2" sound diffuser NRC 1.0
While there was a HUGE difference before/after installation, with about 42dB reduction in broadband noise from 200Hz to 20kHz, the low frequency noise, which had been masked before, is now audible.
I'm sure I didn't do a perfect job installing all of the absorbers, but they do an exceptional job... I have never seen such clean wavelets from clapping in a room.
The noise in the room is very different depending on location, and it is nearly square so modes are a total mess. The 120hz spike I believe to be driven by room modes. Not sure to be honest.
... would a small duct "trap" be able to help sligthly? Would probably be easier to null noise in a duct so it does not exit into the room. And what about stiffening the structure where the ac is attached, and at the same time see what can be done with anti-vibration engine mounts. Make it "float" in a more rigid structure, sort of.
Hi,
You ve got typical structure borne transmission problem.
The only efficient way to deal with that is a box into a box (floating system) but as you said this is too late for what i understand from the room side of the problem.
Usually you build the box inside the room using decoupling pad for the ground, then you build CLD wall to have absorbtion for the low end, then the ceiling is suspended using decoupling support. This is known as 'isophonie' in french, sound isolation i think in english.
Then you'll isolate the source of structural born noise same way.
The only way you have now to deal with it is to isolate your AC: putting it into abox then suspend the box or mount it using isolating/decoupling pad. Then you'll have to use some 'soft' socket/ tube to connect it to existing dispatch tubing.
If the low end is a problem into the room then you probably have too much absorbtion into it (once you have dealt with structural transmission then you'll have to control the room acoustics by itself) . It is usual to have about 1/3 absorption, 1/3 reflection, 1/3 diffusion to have something coherent as a result. But given the dimension of your room you'll end up with issues below schroeder frequency inside your room ( is suspect from 250hz and lower you talked about size of it in your WFS thread iirc). You may need some serious bass trap.
You ve got typical structure borne transmission problem.
The only efficient way to deal with that is a box into a box (floating system) but as you said this is too late for what i understand from the room side of the problem.
Usually you build the box inside the room using decoupling pad for the ground, then you build CLD wall to have absorbtion for the low end, then the ceiling is suspended using decoupling support. This is known as 'isophonie' in french, sound isolation i think in english.
Then you'll isolate the source of structural born noise same way.
The only way you have now to deal with it is to isolate your AC: putting it into abox then suspend the box or mount it using isolating/decoupling pad. Then you'll have to use some 'soft' socket/ tube to connect it to existing dispatch tubing.
If the low end is a problem into the room then you probably have too much absorbtion into it (once you have dealt with structural transmission then you'll have to control the room acoustics by itself) . It is usual to have about 1/3 absorption, 1/3 reflection, 1/3 diffusion to have something coherent as a result. But given the dimension of your room you'll end up with issues below schroeder frequency inside your room ( is suspect from 250hz and lower you talked about size of it in your WFS thread iirc). You may need some serious bass trap.
Last edited:
An externally hosted image should be here but it was not working when we last tested it.
The above picture []imgur: the simple 404 page are my first 14 room modes, and it seems to correspond well to what I'm seeing in noise internal to the room, except for the 15Hz and 30Hz sounds.
A friend of mine has a Type 1 sound level meter, and I'll try to get some input from him on fixing the noise intrusion.
To be clear, the goal has never been a proper listening room - it's only about maximum isolation. In some ways, I feel I have been successful with creating proper isolation - it's so quiet and dead in the room that leaving it is disorienting. One of my good friends lost his balance and fell over immediately after walking into the room. I consider that a huge success.
Your observation on the schroeder frequency issue is spot-on. Below about 180Hz, sounds are harder to localize (harder than normal, anyways) and there are huge nulls and peaks internal to the room. Above 250Hz, everything sounds completely dead and dry, which is kind of what I've been going for.
I figure that a WFS or CABS approach to bass will give me more even coverage, and I'm prepared to account for room resonances with DSP... Noise intrusion on the other hand, I will not allow.
As far as decoupling goes, there is very little I have not done except for the floor, and it seems that floor treatment is the next step (I'm not on top of a concrete bed).
How thick should I plan for a properly decoupled floor to be? Any examples of products I should look to for inspiration?
I think you don't get my point, for decoupling you don't need thick or other thing acoustical treatment, you need floating the room on top of the structural foundation ( or the source of noise, or both).
In fact you face two different problem in my view: the first one i already talked about ( structural born contamination whose answer is floating the room or your AC), the second one is more sneaky (because it is related to the lack of floating): as you created a quasi anechoic condition for frequency past 200hz, you'll face inverse 'masking effect'.
I don't know if you are familiar about it so here is a quick explanation: if you have low frequency at let's say 50hz every multiple of this frequency have the potentiality to be 'masked' if the 50hz is changing it's volume ( overpowering the multiple frequency).
This is this psycho acoustic principle at use in the mp3 encoding: every frequency multiple of the lower one is 'decimated' (because your brain is fooled thinking the high freq isn't present).
During mix it is often used a great example is the intro of " power of equality" from RHCP 'blood sugar sex magic' where guitar sound skinny until the bass come in. Once the bass is present they sound 'fuller' but there is no eq engaged on them, the bass frequency content make your brain recreate the missing freq in the guitar and as a result they sound thicker.
What you face into your room is the inverse: as you don't have reflection in the mid and upper end the only things remaining is the mode from the room & the offending freq from your AC. What is sneaky is that if you had med or hi freq reflection in the room you probably won't notice the AC freq as the balance of freq between them would make them much less obvious.
This is the same situation regularly seen in semi pro studio where they built over absorbent vocal booth where this is always the same outcome: voice sound dull and boomy ( you hear only the mode). The answer is to introduce diffusion in order to reintroduce a bit of med/hi energy into the room.
In your case i suppose as it is your goal you'll have to live with the limitation as the only way i see would be to have a greater volume for the room ( in order tho lower the schroeder frequency and have wider spread of modes ) which is obviously impossible to achieve.
For the AC i repeat as other have pointed the only way to deal with it is to decouple it from the structure to limit structural born transmission. There is pad with resonant freq below 10hz, it'll greatly reduce the 60hz and probably the 30hz tone but not make it disappear.
For it you'll have to decouple the whole room from the structural of the build. The fact that you don't have concrete or hard rock wall is two edged sword: in a way it could make decoupling a bit more difficult than with concrete ( but not impossible) but the room is 'leaky' in the low end so you won't have same bass build up as in a concrete shell ( in EU we have more problem about that because built technics are differents).
In fact you face two different problem in my view: the first one i already talked about ( structural born contamination whose answer is floating the room or your AC), the second one is more sneaky (because it is related to the lack of floating): as you created a quasi anechoic condition for frequency past 200hz, you'll face inverse 'masking effect'.
I don't know if you are familiar about it so here is a quick explanation: if you have low frequency at let's say 50hz every multiple of this frequency have the potentiality to be 'masked' if the 50hz is changing it's volume ( overpowering the multiple frequency).
This is this psycho acoustic principle at use in the mp3 encoding: every frequency multiple of the lower one is 'decimated' (because your brain is fooled thinking the high freq isn't present).
During mix it is often used a great example is the intro of " power of equality" from RHCP 'blood sugar sex magic' where guitar sound skinny until the bass come in. Once the bass is present they sound 'fuller' but there is no eq engaged on them, the bass frequency content make your brain recreate the missing freq in the guitar and as a result they sound thicker.
What you face into your room is the inverse: as you don't have reflection in the mid and upper end the only things remaining is the mode from the room & the offending freq from your AC. What is sneaky is that if you had med or hi freq reflection in the room you probably won't notice the AC freq as the balance of freq between them would make them much less obvious.
This is the same situation regularly seen in semi pro studio where they built over absorbent vocal booth where this is always the same outcome: voice sound dull and boomy ( you hear only the mode). The answer is to introduce diffusion in order to reintroduce a bit of med/hi energy into the room.
In your case i suppose as it is your goal you'll have to live with the limitation as the only way i see would be to have a greater volume for the room ( in order tho lower the schroeder frequency and have wider spread of modes ) which is obviously impossible to achieve.
For the AC i repeat as other have pointed the only way to deal with it is to decouple it from the structure to limit structural born transmission. There is pad with resonant freq below 10hz, it'll greatly reduce the 60hz and probably the 30hz tone but not make it disappear.
For it you'll have to decouple the whole room from the structural of the build. The fact that you don't have concrete or hard rock wall is two edged sword: in a way it could make decoupling a bit more difficult than with concrete ( but not impossible) but the room is 'leaky' in the low end so you won't have same bass build up as in a concrete shell ( in EU we have more problem about that because built technics are differents).
Last edited:
Bobthedespot,
( i like your avatar name)
Have you determined from where the most of offending sound travel? Are the ducts the most offending culprit? If this is i've seen some kind of helmotz resonator used to mitigate the effect in some older studios. It is placed right at the output of the AC. In there is some kind of labyrinth path with lead foam lining the interior.
I know it is forbidden in EU now ( because risk of bacteriological hazard) but if you make the thing accessible and cleanable maybe that could be part of the answer too.
Contact me via pm i may have something for you.
( i like your avatar name)
Have you determined from where the most of offending sound travel? Are the ducts the most offending culprit? If this is i've seen some kind of helmotz resonator used to mitigate the effect in some older studios. It is placed right at the output of the AC. In there is some kind of labyrinth path with lead foam lining the interior.
I know it is forbidden in EU now ( because risk of bacteriological hazard) but if you make the thing accessible and cleanable maybe that could be part of the answer too.
Contact me via pm i may have something for you.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.