Jump to content

Noob's adventure into accoustics


Recommended Posts

17 hours ago, almikel said:

As with most things, there's no free lunch - truly broadband absorption and thin????

Try them  - I'd love it if they worked.

 

IME if you have a lightly constructed room, absorption works fine down to 100Hz or so and EQ works fine below there.

If you have rigid walls your challenges are greater.

The broadband absorption seems to come form the unique resonance pattern of a steel plate, which has to do with the speed of sound of steel materials.

I would love to see them work, but as stated before there is hardly any supporting evidence in DIY projects that have great impact. The other things is the same problematic issue is size. To be effective in the low region (wavelength) they need a minimum dimension of 100cm. Include the minimal recommended aspect ratio of 1:1.5 and the minimum panel size is 100x150cm.

 

Left wall is solid double brick without air gap. The front and right wall are a double brick wall with air gap and a window. Back wall is a single 70mm timber stud wall with insulation and 16mm plywood and 13mm plasterboard on both sides. Ceiling is 10mm timber slats with 100mm insulation and a sheet roof.

Link to comment
Share on other sites



Guest Peter the Greek
3 hours ago, Primare Knob said:

Which trap design do they use then? By the looks of it, these guys don't do your typical small room theater either.

Keith Yates: http://keithyates.com/

Dennis Erskine: http://erskine-group.com/home-theater-acoustic-design/

 


 

 

 

I can't speak for Keith... he's a freak of nature. But you dont see them in his rooms.

 

I've been in several of Dennis', he uses the riser, ceiling (that one I linked is his) or other areas like that little room I showed you.

 

I had no peaks in my old room (same as yours more or less) and any dips would be covered by sub placement (2 subs) and eq. I only had one sub before we moved so I couldn't test that. But no peaks.

 

Dennis is often quoted in sayibg all you need is a riser and two subs.

 

Ps. All rooms are "small" :D

 

 

 

Link to comment
Share on other sites

11 hours ago, Peter the Greek said:

Ps. All rooms are "small" :D

That is true. The next room is always going to be bigger.

 

Just for clarification. Could a "riser" stage being used at the front? I don't want to pass on, on an opportunity, but a riser in the back isn't going to be a user friendly option.

Link to comment
Share on other sites

Reading a bit through the riser document poster earlier by @Peter the Greek I have come up with the following.

 

Like the stage, the room should be completely sheetrocked before your build, with any gaps between the floor and the sheet rock completely filled with a 50 year flexible chalk.

I assume this has to do with soundproofing of the room and not so much with the construction of the riser

 

 

Unlike the stage the platform can touch the walls. I assume this has to do with the tactile transducer stage described in the document.

Height seems to be based upon seating usability rather than insulation thickness. It looks like a clever and opposite version vs  absorbing ceiling treatment.

 

The platform doesn’t seem to be based upon a helmholz resonator and doesn’t seems needing an airtight space. The function of all the layers of roofing felt seems to be reducing the risk of vibrations and resonations, the different thickness of ply seems to imply this as well to stop it from turning into a resonator, and the size for the  gaps are based upon the size of available vents.

I don’t really follow why these vents/holes are there in the first place, as LF bass and room nodes would travel straight through the construction, but perhaps this has to do with the nodes in the upper bass region.

I did find the following information in relation to this.


http://www.acousticfrontiers.com/adding-a-home-theater-riser-to-an-enthusiasts-room/

“The high volume to open surface area ratio naturally acts as a low pass filter, providing absorption in the lowest frequencies only. The percentage open area must be calculated, as must the internal volume, and the number and placement of vents engineered to create the desired low frequency bass trapping effect. Just cutting a few vents in here and there will not turn your riser into a bass trap, there is some definite engineering that needs to take place here!”

 

The bigger question is, what is the minimum size for effectiveness of this type trap, and I still can’t follow the reason why this wouldn’t work at the front stage, as I can’t imagine that speakers would pass on that much vibrations into the floor construction that the unit becomes a resonator, that is with my plan to put my subwoofers halfway the room.

Link to comment
Share on other sites

A riser would have to be build as something like this, on top of the existing floor. It won't be the most practical option, wen it comes to room entry or for the couch unless I extend it to pass in front of the window or make a step in front of the couch,  but still an option.

Riser_03.jpg

Riser_01.jpg

Riser_02.jpg

Link to comment
Share on other sites



Guest Peter the Greek

I reckon that looks ok...nothing to put feet on and makes getting in/out of seat hard though....

 

Whats the exact dimensions of the room again?

 

We use green glue now, not felt.

 

You could put speakers on cast in place concrete and build a riser trap around it. Key is having contining "air" flow through the volume.

Edited by Peter the Greek
Link to comment
Share on other sites

On ‎23‎/‎01‎/‎2018 at 7:50 PM, Peter the Greek said:

Ps. All rooms are "small" :D

 

13 hours ago, Primare Knob said:

That is true. The next room is always going to be bigger.

All of our domestic rooms are "small" from an "acoustic" perspective, and they have very different issues to large auditoriums:

  • our rooms aren't large enough to ever generate a "diffuse" sound field (hence you see discussions on the interweb that reverberation times don't apply to domestic rooms - IMO they do, but in the right context - eg comparing the RT times for the same room before/after treatment
  • modal issues typically don't apply for large rooms (auditoriums) - the room is sufficiently large that the modes are very dense within the audible range (like domestic rooms above 500Hz)
  • large rooms with lots of people absorb lots of treble - just the large distance from the speakers to listeners absorbs lots of treble

The "pioneers" of Acoustical Engineering (eg Wallace Sabine) did their studies on large auditoriums - some of this doesn't apply to our small rooms.

 

11 hours ago, Primare Knob said:

I did find the following information in relation to this.


http://www.acousticfrontiers.com/adding-a-home-theater-riser-to-an-enthusiasts-room/

I have a lot of respect for the Acoustic Frontiers guys.

I regard their paper on Measurement Standards for Listening Rooms as the reference guide - you've probably read it, but linked here just in case:

http://www.acousticfrontiers.com/wp-content/uploads/2011/10/acoustic_measurement_standards.pdf

 

 

Mike

  • Like 1
Link to comment
Share on other sites

Thanks for the document, I have read it now.

I have a some questions about the first pages.

To test the RC30 value of the room I should be using the RTA in REW. Do I need any specific settings?

The other thing is reading and producing the reflected sound graphs.

First of all, I don't know what to look at as the graphs all look the same to me, and can't pick up the difference.

Second, How did they measure the different fixed frequencies, and how to smooth the impulse ETC graph?

Link to comment
Share on other sites

2 hours ago, Primare Knob said:



To test the RC30 value of the room I should be using the RTA in REW. Do I need any specific settings?
 

I've never tried to quantify my background noise, other than using a Sound Level Meter (SLM). On a quiet night my room gets down to 25-30 dB SPL on C weighting. I'm not exactly sure how to correlate RC30 compared to the single number an SLM provides.

If you've calibrated REW to an SLM then you could use the RTA or the SPL meter.

You can't change the room's background noise without construction work focusing on isolation/sound proofing, so I haven't paid it much attention.

@Peter the Greek has done much more work in this area than I have.

 

2 hours ago, Primare Knob said:



First of all, I don't know what to look at as the graphs all look the same to me, and can't pick up the difference.

 

first graph B1 has left/right speakers decaying similarly - I added the black line

5a69955608b1a_goodetc.thumb.PNG.c1aae2e0d03fb230dfda89000bb3bb74.PNG

 

2nd graph B2 has reflections around 15ms still as high as first reflections - the graph doesn't decay much over time - reflections up at 35-40ms nearly as intense as first reflections

 

B3 again has large energy reflections out to 20ms and a climbing response 35-40ms

 

3 hours ago, Primare Knob said:


Second, How did they measure the different fixed frequencies, and how to smooth the impulse ETC graph?

I don't have REW on this PC...so can't check settings easily...

They're not fixed frequencies - they're comparing different octave bands - there'll be a setting in REW to band limit the ETC and apply smoothing - when I get on a PC with REW I'll have a look.

 

This criteria is mostly about the speakers' response being consistent on and off axis, but room diffusion helps also.

 

cheers

Mike 

 

 

  • Like 1
Link to comment
Share on other sites



I have moved my subs to the halfway point of the room, and made adjustment to the LP of only 10cm up and or backward.

 

The SPL graph, looks the smoothest with the mic placed 110cm from the back wall, but the decay graph shows an increasing peak over time at 70Hz.

 

I think this has to do with the 2nd room length node which is close to 70Hz instead of the 1st room height node which is also close to 70Hz

 

The waterfall graph shows a decay after 300ms of 18.5dB at 70Hz with the mic placed 120cm from the back wall, and only 10.0dB with the mic placed 110cm from the back wall.

70Hz resonant peak SPL.jpg

70Hz resonant peak 110cm.jpg

70Hz resonant peak 120cm.jpg

Link to comment
Share on other sites

I believe the RC30 rating is an A weighted rating with different values for different frequencies.

 

https://web.iit.edu/sites/web/files/departments/academic-affairs/Academic Resource Center/pdfs/Workshop_-_Acoustic.pdf

 

I did the RTA measurement and saved the file, but I can't get REW to display it when I reopen it.

At night the room is well below RC30, but that can change easily as the room isn't sound proofed.

 

But it did show very sharp singular peaks at 1, 2, 3, 4kHz. I wonder what those could be.

Link to comment
Share on other sites

I did watch HT geeks on Youtube, the Acoustics 101 and 102 episode which helped me to get a better understanding of standing waves.

 

They show a diagram with room positions with the least impact of the room nodes.

0.2, 0.32, 0.45, 0.55, 0.68, 0.8 of the room dimensions.

 

As always, reality is different as I found out, which could have to do with SBIR.

Link to comment
Share on other sites

Guest Peter the Greek

Have you seen this? Probably the finest home theater in existence right now anywhere in the world. That threat is the home theater equivalent of what a yellow pages sized porno would be to a 15 year old. There is some real gold in Keith's explanations of how they calibrate the room too. Worth reading from start to end. Not necessarily relevant to the above, but might give you some ideas......it has me.....I want to build something a similar size as a stand alone building on our place :D

Link to comment
Share on other sites

It would probably be cheaper to buy a commercial cinema, and start living there. They come included with pop corn machines.

 

If I follow it correctly, the use the Trinnov as a glorified switcher, probably because it is the only one that offers 32 channels. It looks like the Trinnov is sending 32 channels of analogue to the Lake LM44 which converts from A to D, does the DSP per speaker, or speaker group, converts D to A, before sending of to the amplifiers and the 45 speakers.

 

I don't understand why there is a Trinnov in the first place, as it seems a big waste of money, for simply Atmos decoding.

 

Edit:

It look like the Trinnov can output 16 digital channels. Not sure if these are stereo channels.

 

Nope, 16 channels max of digital output. So they are probably not using this, as the basic version can do the same.

Link to comment
Share on other sites



On ‎23‎/‎01‎/‎2018 at 4:18 PM, Primare Knob said:

 

Left wall is solid double brick without air gap. The front and right wall are a double brick wall with air gap and a window. Back wall is a single 70mm timber stud wall with insulation and 16mm plywood and 13mm plasterboard on both sides. Ceiling is 10mm timber slats with 100mm insulation and a sheet roof.

this is a tough room to treat due to the 3 rigid walls, and likely other rooms creating modes.

 

 

Link to comment
Share on other sites

On ‎17‎/‎01‎/‎2018 at 10:15 PM, almikel said:

 

Heading down a slat + absorption path on the ceiling will provide good absorption down to 150Hz or so but keep higher frequency energy in the room - "in room" measurements would likely see benefits below 150Hz due to the size if you covered the ceiling - the ceiling absorption would be exposed to absorb "grazing" energy from all lateral modes.

 

Looking at the models - I would recommend using "fluffy" instead of anything denser above the slats - it's cheaper and models better.

Below is 100mm wide slats with 100mm spacing, with above 200mm fluffy (5000 thingies) vs 200mm mineral wool (27000 thingies) - fluffy is blue

5a5f398d70500_slats200mmfluffyvs200mmmineralwool.JPG.51ed2d06a88b073b9ad664158a01f352.JPG

That's a good bass trap (blue line)  - you may only need some EQ below 100Hz (which is where EQ works very well).

 

cheers

Mike

In the absence of professional advice - I still think the above would help significantly

Link to comment
Share on other sites

After many more measurements I thought I had found the 70Hz disturbance. Opening the back door in the room and the 70Hz peak in GD disappeared when measuring the subs. Unfortunately the opposite happens when I open the door and measure the L+R speakers.  The decay however is more in line with expectations with the door open, so I do believe the door is resonating at a certain frequency.

 

None the less, I think I am done with moving speakers an LP, and by know have found the best possible position considering room constraints.

Link to comment
Share on other sites



On 21/01/2018 at 5:59 PM, almikel said:

A key thing I learned from @davewantsmoore is that all wiggles in a frequency response create phase changes.

 

Frequency response and phase response are just two different views of the same thing.

:thumb:

 

 

Unless you're using an EQ which can distort time (eg. FIR), then if you have a wiggle in frequency (or phase), then there will be a wiggle in the other.

 

<wiggle>

Link to comment
Share on other sites

17 hours ago, davewantsmoore said:

 

Frequency response and phase response are just two different views of the same thing.

:thumb:

 

 

Unless you're using an EQ which can distort time (eg. FIR), then if you have a wiggle in frequency (or phase), then there will be a wiggle in the other.

 

<wiggle>

If we are talking about phase in an SPL graph we are talking about soundwaves and how the phase changes with frequency? But what does a phase setting do on a subwoofer?

And is phase measured in the time domain?

Link to comment
Share on other sites

On 2/6/2018 at 2:18 PM, Primare Knob said:

If we are talking about phase in an SPL graph we are talking about soundwaves and how the phase changes with frequency?

yes we are

On 2/6/2018 at 2:18 PM, Primare Knob said:

And is phase measured in the time domain?

No - phase is typically measured in the frequency domain - as per your SPL graphs showing your frequency response (more accurately described as the Amplitude Response) and phase response, both are in the frequency domain - the "X" axis on your graphs (horizontal axis) is in Hz (frequency)

 

Impulse response is in the time domain and has seconds on the X/horizontal axis.

 

On 2/6/2018 at 2:18 PM, Primare Knob said:

 But what does a phase setting do on a subwoofer?

For "old school" analog phase dials on a sub, they implement an "all pass" filter response which adjusts phase (delay), but not amplitude.

I realise this is contra to what @davewantsmoore and I have said above (ie wiggles in the frequency/amplitude response always reflect in the phase response) - in this case deliberate changes in the phase response (via the phase dial on a sub to create delay) shouldn't impact the frequency/amplitude response of the sub.

In other words, the phase dial adjusts phase/delay of the sub without changing the frequency/amplitude response of the sub.

All analog EQ is Infinite Impulse Response (IIR), as opposed to DSP, which is Finite Impulse Response (FIR).

On 2/5/2018 at 8:35 PM, davewantsmoore said:

Unless you're using an EQ which can distort time (eg. FIR), then if you have a wiggle in frequency (or phase), then there will be a wiggle in the other.

My understanding is that old school IIR "all pass" analog filters found on subs don't distort time - hopefully @davewantsmoore will chime in with a good explanation of how "IIR All Pass" analog filters change phase but not amplitude.

 

Modern DSP FIR filters have the ability to muck with phase and amplitude independently - hence Dave's comment on distorting time.

 

Coming back to analog phase controls on subs - keep in mind that all filters have delay (passive, active, digital whatever).

The filter delay is inversely proportional to frequency (delay gets higher as frequency drops) and proportional to order of the filter (delay gets higher as steepness increases).

This means that if your sub is next to your mains, to achieve time alignment, you usually need to delay your mains, rather than your sub.

The analog phase control on the sub can only delay the sub further, and is only useful if the sub is much closer to the listening position than the mains.

A compromised result can be achieved by delaying the sub some cycles to align later with the mains (to use a @Paul Spencer analogy, the sub missed the bus, let's delay it further to catch the next bus).

 

Aligning mains with sub/s is where IMO, DSP becomes essential.

You place your sub/s in the room based on the best bass response and based on their position you time align them with your mains - often requiring delay on your mains and likely individual subs in a multi sub setup.

Achieving this with DSP is hard - without DSP it's virtually impossible unless your subs are way closer to the listening position than your mains.

 

To provide an example, and I'm quoting numbers from memory - my sub is at least 1m closer to my listening position than my mains and I'm still running around 10ms of delay on the mains to time align with the sub  based on mic measurements at the listening position.

 

cheers

Mike

  • Like 1
Link to comment
Share on other sites

Great post above. Just to clarify changing the phase of the sub won't change its frequency response alone that it outputs but the frequency response at the listening position(s) will change in the sub-woofers frequency range when combined with the output of speakers.

Edited by Satanica
  • Like 1
Link to comment
Share on other sites

Guest Peter the Greek

@almikel - regarding the FIRs, I was of the understanding (probably incorrect from the sound of it) that there wasn't really any DSP's out there that you could use FIRS on down into the bass region (or even mid bass for that matter) as they were limited in their number of taps/processing power? So they're really only suited for crossover work for a tweeter?

 

Some of the more commercial dsps have a fixed delay for all filters applied. So if you have 1 or 20 filters it doesn't matter, they're all time aligned (I know QSC does this and am pretty sure Xilica do it too...and Lake....)

 

P.S. my "understanding" of this is really basic, hopefully I've said all that correctly

Link to comment
Share on other sites



  • Recently Browsing   0 members

    • No registered users viewing this page.




×
×
  • Create New...
To Top