Helmholtz - constuction advice

[boggy]

Originally Posted by bedope

........

My questions are:

1. Where to place the tube. Narrow side? Wide side? Corner? Center? Please find a sketch attached.

Tube opening needs to be at place where sound pressure level at target frequency is maximal.

Originally Posted by bedope

2. Secondly, according to spreadsheet provided in the above thread, for this size cabinet and frequency 36Hz my tube should be ~60mm (2.3 inch) diameter and 120mm (4.7 inch) long. Does it sound right?

It's not need to be... I always tune my absorbers before usage. Formula is only approximative. For measuring working frequency you can use guide I wrote at Pro Sound Web R/E/P forum: Measuring Helmholtz absorber resonant frequency in non-anechoic space.

Originally Posted by bedope

3. If I increase tube diameter my resonant frequency is getting higher. Is it really correct?

Yes, if tube length is same.

Originally Posted by bedope

3a. What is better, wider and longer tube or shorter and narrower? What a ballpark for tube size/diameter in regard of resonator volume?

Best result is maximum possible absorption surface (tube diameter) of Helmholtz absorber, with appropriate AND reasonable tube length. Wider and longer tube is better, in short.

Originally Posted by bedope

4. In there anything else anyone would advice in my case regarding the resonator?

Perhaps you can try this: Tim's Limp Mass Bass Absorbers when you get tired of HH absorbers.

Originally Posted by bedope

Thanks for all advices,

You're welcome.

[bedope]

Boggy,

thanks a lot for advice. Few more questions that strike me after reading thread provide.

1. It looks like the conclusion is: more holes is better then one, which probably determines the wide surface is more suitable. Also it make sense for surface pressure point of view. Ok, that fine. Now, how do I refer it to calculation provide in the spreadsheet? Does it mean, if my spreadsheet gives me a result for 60mm diameter tube and I would prefer to have lets say 10 tube, each tube should be have area = (60mm tube area)/10?

2. Also is my box size (7x11x27 inches) too small and potentially having 0 effect at 36Hz? The spreadsheet give me Q around 12 with damping factor 10 to the dimensions. Also is Q=12 good enough to shave of few dBs and cut ringing at 36Hz by lets say extra 0.2ms for 0.6-0.7s?

3. Thanks for measurement advice, that will be definitely useful when the box is build, but first I would like to have some design close enough to the final dimensions.

4.I tried vpr panels, did not work so well for me. Probably cause more trouble then it solved. That is why I try to focus on the lowest modes: 36Hz and 50Hz and targeted dumping. The rest (above 100Hz or so) is ok, around +/-7db (24 octave smoothing) and and in 0.3-0.4s T30.

Thanks, bedope!

[boggy]

Originally Posted by bedope

Boggy,

thanks a lot for advice. Few more questions that strike me after reading thread provide.

1. It looks like the conclusion is: more holes is better then one, which probably determines the wide surface is more suitable. Also it make sense for surface pressure point of view. Ok, that fine. Now, how do I refer it to calculation provide in the spreadsheet? Does it mean, if my spreadsheet gives me a result for 60mm diameter tube and I would prefer to have lets say 10 tube, each tube should be have area = (60mm tube area)/10?

Yes.

Originally Posted by bedope

2. Also is my box size (7x11x27 inches) too small and potentially having 0 effect at 36Hz? The spreadsheet give me Q around 12 with damping factor 10 to the dimensions. Also is Q=12 good enough to shave of few dBs and cut ringing at 36Hz by lets say extra 0.2ms for 0.6-0.7s?

Possibly.

Originally Posted by bedope

3. Thanks for measurement advice, that will be definitely useful when the box is build, but first I would like to have some design close enough to the final dimensions.

No problem...

Originally Posted by bedope

4.I tried vpr panels, did not work so well for me. Probably cause more trouble then it solved. That is why I try to focus on the lowest modes: 36Hz and 50Hz and targeted dumping. The rest (above 100Hz or so) is ok, around +/-7db (24 octave smoothing) and and in 0.3-0.4s T30.

Ok.

[bedope]

Testing... testing.. testing

3 specimens for fighting against 36Hz mode:

B. test helmholtz tune to 36.5Hz, 2x 28cm x 12cm x 34cm, opening 29mm diam, 11cm long (not so big and undumped)

C. vpr panel 2mm x 1200mm x 600mm metal glued to 100mm high density poly panel ( I know, not ideal)

D. poly panel, high density, 100mm thick (reasonable sized panel, thick and dense)

Please find some comparison graphs attached.

My conclusion are:

1. that small Helmholtz is actually quite effective. Tuning according Excel spreadsheet provided by shawn19 (thanks a lot!!!) seems to be quite accurate. SPL graph clearly shows 0.8dB reduction at 36Hz. Waterfall shows nice cut off right in the middle of 36Hz ringing as well and drop in SPL by about 3-4dB at 600ms. A lot of effect from a small box. This has been consistent with number of measurements I have done with this box in different location around my room.

Other solutions seems to have 0 (zero, null, cero, ling, shunya, rei, sifr) positive effect at 36Hz. I know, I should have more volume. I could have. I could have my whole room stuffed with insulation bats. But I don't want. I want something thin, not obstructive and easy to compose into the room. Think Helmholtz trap brings promise of that.

2. I tried some dumping the Helmholtz box. With different option from filling cavity with very fluffy polyester insulation, putting small strip of that insulation inside tube or in front of tube. All has negative effect, which mean it was too much dumping.

3. In general, I could easily build few more boxes like that, stuff in the corner like a tower and hopefully could shave of next few decibels and few hundred ms in ringing.

Now, perhaps I should try wide opening and check if there is any positive effect. Also maybe wide box, something like 6 inches rather then 4 could help.

Any comments welcome! Thanks for help everybody,

bedope

[boggy]

Originally Posted by bedope

........
2. I tried some dumping the Helmholtz box. With different option from filling cavity with very fluffy polyester insulation, putting small strip of that insulation inside tube or in front of tube. All has negative effect, which mean it was too much dumping.

They all have "negative" effect because tube isn't a place for damping material in HH absorber. If you want your HH resonators to works, put porous absorption (this fluffy thing you already use) in the box itself, without stopping air circulation from tube, even at side which is inside box. Air in tube is a "mass" in vibration system where air in the box is spring. If air in tube cannot circulate, HH absorber wouldn't work (as you already realise).

Absorption in HH absorber is needed to modify a strong acoustic resonator (which, overall, makes things worse in the room), to absorber.

Originally Posted by bedope

3. In general, I could easily build few more boxes like that, stuff in the corner like a tower and hopefully could shave of next few decibels and few hundred ms in ringing.
........

Great, build this boxes until you become fully satisfied with your room bass response...

[OpusOfTrolls]

Absorber in the box also lowers the resonant frequency, so the tube needs to be resized afterwards.

[boggy]

Originally Posted by OpusOfTrolls

Absorber in the box also lowers the resonant frequency, so the tube needs to be resized afterwards.

Yes, and it lowers Q factor too, which means that "absorption" is lower at resonant frequency, but working frequency range is wider.

They are two reasons for (possible) "negative effect" (which really isn't negative at all). Main problem is to find ratios for all parameters (box volume, quantity of absorbing material, diameter and length of tube) which best fits in some particular room (impedance matching)

[OpusOfTrolls]

Thanks for the info guys! I have been avoiding HH because I didn't know how they worked. I thought it was the cavity volume that defined the resonant frequency

[OpusOfTrolls]

If you ask me, your HH is too resonant. A Q of 12 is somewhat resonant compared to Q of 2 or 3.

[bedope]

Yes, I agree its very resonant and that's why I tried to dump it.

With dumping I am aware that over dumping will kill HH. That's why my fluffy in the tube was really, really think, very stretched, one could see through or blow through even easier and we are talking about volume of maybe 2-3cm3. Really tiny. But it looks like it's too much.

Anyway, I will try different versions of dumping with this HH before moving to larger opening test. I think I am on a right track to cure my 36Hz problem.

Btw, I tried insulation in a corner, mix of high density in front and a lot low density behind. We are talking a lot, like 1m x 1m x 2m volume. Yes it help at 36Hz, probably as much as this HH

Thanks, bedope

[kasmira]

Yes, dampening inside the actual HH panel will lower the Q and (hopefully) make it more effective. As stated, the dampener used should not be inside the tube but inside the panel.

[bedope]

Thanks a lot guys. I did some experiments with dampening (not dumping ) my HH. Please refer to graph attached.

Grey line: no HH in the corner

Blue line: HH has 3 cavity walls covered with fluffy 1 inch insulation. HH is tuned to 36.5H, where my nasty mode sits. It can be observed HH's resonance frequency has lowered, because the tip moved slightly toward higher frequencies (more efficiency below 36.5Hz). Still not bad, SPL at 36.5Hz lowered by about 0.5dB. Also judging by waterfalls (which I did not include) resonance, with additional fluffy is around 32Hz - new dip in waterfall there.

Pink line: So I adjusted opening length to tune HH at 38Hz, just buy a guess. That brought the tip again around 36.5Hz and lowered SPL by further 0.4dB to total 0.9dB. At the same time SPL at 600ms dropped by 2db. Thats the best result in this series. Good, but I hoped for more effect from dampening.

One more try to do is try 40Hz tube with pink line configuration (fluffy on walls only). Perhaps I might shoot in better and shave off SPL a bit more.

Orange line: so I filled the cavity further with fluffy insulation to about 35% volume and at the same time tuned opening to 40Hz. SPL seems to stay at the same level, but at waterfall at 600ms shows only 1db difference between grey and orange. So filling the cavity more the just walls seems to not helping that much.

On other note the pink configuration show very reasonable improvement in shard dips around 144Hz and 366Hz, which is good.

Any toughs on this particular results and HH before I move to larger opening?

thanks, bedope

[boggy]

Originally Posted by bedope

.......
Any toughs on this particular results and HH before I move to larger opening?

Be careful, larger opening needs longer tube for the same frequency, so check if you have enough space (from the back side) in the box for longer tube.

Originally Posted by bedope

thanks, bedope

You're welcome.

[bedope]

Originally Posted by boggy

Be careful, larger opening needs longer tube for the same frequency, so check if you have enough space (from the back side) in the box for longer tube.

Yes, I have noticed that. For 40mm diameter opening I need 210mm depth. Cavity depth is 280mm. Is 70mm clearance between end of the opening and cavity wall enough. My guess is: should be enough, double size of the diameter, so there should not be particular resistance for fast air entering the cavity and blocking dispersion. But that's only my guess

Additional question: does the whole tube/opening have to sit in the cavity? Cannot be half inside, half outside? Perhaps could be considering a bottle, classic example of HH has whole opening outside cavity

thanks, bedope

[boggy]

Originally Posted by bedope

Yes, I have noticed that. For 40mm diameter opening I need 210mm depth. Cavity depth is 280mm. Is 70mm clearance between end of the opening and cavity wall enough. My guess is: should be enough, double size of the diameter, so there should not be particular resistance for fast air entering the cavity and blocking dispersion. But that's only my guess

Yes, your intuition works well. About one pipe diameter is enough clearance to still have linear system.

Originally Posted by bedope

Additional question: does the whole tube/opening have to sit in the cavity? Cannot be half inside, half outside? Perhaps could be considering a bottle, classic example of HH has whole opening outside cavity

You can try this by yourself and share your experience...

90 degree elbow is useful to extend length in the box.

[bedope]

I continued testing my little box with different configuration for opening and insulation. What really mattered was better dampening, which I done with very high density and flow resistance polyester insulation.

Results are stunning for this little box. 1.8 dB reduction at 36.5Hz peak. 3.5dbB reduction at 36.5Hz at 600ms. And last, but not least 12db reduction in dip at 32Hz. From my perspective that really, really stunning.

Required a lot of testing and fine tuning. Once I am ready to build final boxes, will have to double check results and fine-fine-tuned again. I am going to put one more in the same corner in the same position, each on top of each other and perhaps two more in other corner.

Lets the graph speak for the rest.

bedope

[bedope]

Ok, here we come, after some extra work and even more tests I got a good result. I build two final boxes, it total four chambers.

Results:

- SPL at 36.5Hz dropped less then I hoped, still having 1.1dB drop. Not as good as one of my test boxes (I could not repeat that), but still ok result. Also there are improvements around 50Hz and 32Hz as a side effects, which is great.

- However... waterfall at 600ms at 36.5db dropped by 9.5dB!!! At that's what I want to see!!! Very good result for two small boxes in a corner. I recon one more box at 36Hz (which is being build) and I am done with this mode.

- Also there is improvement in harmonics, about 5db in 330-340Hz dip and 385Hz. Plus number of small improvements across the board.

Looks good. On the other hands HH are very sensitive to dampening, opening, sealing ect. Construction has to be very careful and final boxes has to be tuned.

I have not have time to asset result by ear. Maybe today evening.

Any toughs?

Thanks, bedope

[OpusOfTrolls]

Good work. Helmholtz won't hit hard into frequencys like a porous absorber will, unless it is tuned wide. The construction looks solid which is important. You can always add damping compound to the box to make it resonate less, but it might not work well at such a low frequency. I don't see how your trap will affect harmonics so much, maybe that is the result of the trap placement and it's displacement?

Anyway, I think these things have to be very large at low frequencys to have a better impact on frequency response. The absorbent inside the box should have the same effect. but only if it IS absorbent at the frequency of interest.

[bedope]

... and one more update. I been tuning the second box better and so I got slightly better result. SPL at 36.5 dropped by 1.8dB and at 600ms 9.7dB. Also I include spectrogram, which shows 100ms improvement at T30 at 36Hz mode.

OpusOfTrolls, yes I understand i need more boxes. And thats fine, I have place to next 8-9 if I need to. Those are just pilots

I though about harmonics at 330Hz and 380Hz as changes there seems to correspond to changes around 32Hz and 38Hz. However I agree, placing a box in a corner could slightly change dynamics of the system and affect those higher frequencies purely by geometric.

There is one more box coming. I will try to damp this one more evenly across the board and see how it will behave. Boxes itself are made from 16mm cheap board and extra divider in the middle should help to rigid the construction. Seems to be pretty solid for the size. I buy of shelf kitchen cabinets for that. Easy and quick. Material I use for damping seems to be pretty good at lower frequencies (0.6 at 125Hz) so it should be alright.

So far I am happy with the results. If I can get with next 2 boxes further 2dB at SPL and extra 10dB at 600ms, I think I will be done with 36Hz and will move to 40-50Hz dip.

bedope

[DanDan]

Interesting experiments bedope.
It is amazing enough to see changes caused by such small devices.

A couple of points.

Sorry to disagree, but I have seen many recommendations to place damping at, near, or in, the holes of Helmholtz panels. At a position of such high particle velocities very little is needed. Fabric across the holes is often enough.
I would have thought the same principles would apply to a single holed trap.

REW allows Waterfalls up to 1500mS. There is no need to restrict your views to 600mS.
Your WF graphs peak at around 85dB but many of the modes are still up at 50dB or so at 600mS. 35dB is not enough of a span to derive a decay curve.
Topt is your best (opt)ion in REW for simplistic decay figures. Third Octave.
Or better still, do your Waterfalls with sufficient dynamic range.
The SPL level in the Room should be say 80-90dB C Slow. Enough for you to need ear protection.
The length of the window also affects these graphs. Try different lengths and you will see.

It is critical to have the microphone and speakers in the same position for before and after tests. For the modal work you are doing, corner placements would likely be best.
Even a 50mm change of position can cause large LF response changes.

The dip at 40-50Hz is more likely a cancellation null than a modal effect IMO.
Probably Back Wall BIR. http://recording.org/studio-construc...en-matter.html

DD

[OpusOfTrolls]

The nulls at 43 and 86 hz are caused by a boundary reflection. I would assume first that it is caused by the wall/ceiling or wall/floor edge, but I could be wrong. The nulls appear at a half wavelength reflection. You need to take the distance from the listening position to the monitors, then add 13-14 feet of reflection travel distance. The offending surface is most likely be at half this distance. I don't know how large your room is, but the reflection travel distance could also be 26-28 feet.

[shawn19]

hi Bedope,
very nice results. sorry for the late answer, this may be useful for any other boxes you make.

Originally Posted by bedope

1. Where to place the tube. Narrow side? Wide side? Corner? Center? Please find a sketch attached.

the best place to put the tube is in the center. this comes from a couple of assumptions made when deriving the helmholtz formulas. the formulas assume that the pressure is the same everywhere inside the resonator. placing the tube in the center makes the opening about the same distance to the different walls. also, the tube end corrections assume that one side is in the free field, and the other in an infinite wall. again, placing the tube in the center makes the infinite wall assumption just a little better.

Originally Posted by bedope

1. It looks like the conclusion is: more holes is better then one, which probably determines the wide surface is more suitable. Also it make sense for surface pressure point of view. Ok, that fine. Now, how do I refer it to calculation provide in the spreadsheet? Does it mean, if my spreadsheet gives me a result for 60mm diameter tube and I would prefer to have lets say 10 tube, each tube should be have area = (60mm tube area)/10?

2. Also is my box size (7x11x27 inches) too small and potentially having 0 effect at 36Hz? The spreadsheet give me Q around 12 with damping factor 10 to the dimensions. Also is Q=12 good enough to shave of few dBs and cut ringing at 36Hz by lets say extra 0.2ms for 0.6-0.7s?

im sorry if it was not clear, but the fewer holes the better. as Boggy mentioned, the absorption is proportional to the square of the opening size, so the larger the better. also, one large hole is better than two of the same size. of course, with a larger opening, the tube has to be longer for the same frequency, which reduces the efficiency. it is a trade off, and the easiest thing is just use the spreadsheet to find a good solution.

as far as the damping of the resonator, both resistance in the body of the resonator and resistance across the opening widen the Q. the problem is estimating how much resistance needs to be added. if it is inside the body of the resonator, changing it later is more difficult than if it is across the opening. you mention you have two resonators, and may make more. if you find the optimal resistance for two resonators, then need to add more, the resistance for all the resonators will now be different. also, adding resistance to the body changes the resonant frequency, where adding resistance to the opening does not.

hope this helps,
Shawn

[bedope]

Originally Posted by DanDan

Interesting experiments bedope.
The dip at 40-50Hz is more likely a cancellation null than a modal effect IMO.

I agree with that and have done some test a while ago with BIR. However I also tunes my test boxes to 40hz and 50Hz and could see some good improvements. Perhaps both, mode and BIR affect this area. Something to check once i finish with 36Hz. But, thanks for pointing and reminding! And thanks to OpusOfTrolls as well. I will look closer at your suggestion guys regarding BIR next.

Originally Posted by shawn19

hi Bedope,
very nice results. sorry for the late answer, this may be useful for any other boxes you make.

the best place to put the tube is in the center. this comes from a couple of assumptions made when deriving the helmholtz formulas.....
Shawn

Thanks Shawn for detailed explanation. The next box will have opening in the middle of the chamber. I will be a good exercise to compare how two similar boxes can perform differently (better or worse) when opening moves. Good I did not cut wholes in the front panel yet

Originally Posted by shawn19

hi Bedope,

im sorry if it was not clear, but the fewer holes the better...

as far as the damping of the resonator, both resistance in the body of the resonator and resistance across the opening widen the Q. the problem is estimating how much resistance needs to be added. if it is inside the body of the resonator, changing it later is more difficult than if it is across the opening. you mention you have two resonators, and may make more. if you find the optimal resistance for two resonators, then need to add more, the resistance for all the resonators will now be different. also, adding resistance to the body changes the resonant frequency, where adding resistance to the opening does not.

hope this helps,
Shawn

I agree with what you said regarding bigger or smaller opening. I had 3 diameters tested, 29mm, 38mm and 44mm. 38mm performed better then 29mm. And 44mm would not fit at all due to length limit So I naturally settled to 38mm.

As I mentioned before I had problems with adding resistance to the opening, so I got into direction filling cavity and yes, resonance has changed. As an example, those boxes tune to fight 36Hz have tubes tuned to 39Hz and 40Hz.

I know its a dynamic system, so I might adjust it as it goes. PCV tubes are cheap and easy to cut. I already have a collection of tubes marked with different frequencies


Thanks again guys. I will keep posted once more progress in done.

bedope

[DanDan]

The frequency of the nulls caused by BIR will change with speaker position. Mode frequency never changes.

I made a tuneable Helmholtz way back. It had a thick lid. I was able to alter the tuning by sliding the tube in and out a little. Just a cm or so. This was enough to get spot on the actual modal resonance.
Acoustics Forum • View topic - Helmholtz formula, is this incorrect?

DD

[bedope]

.. forgot to publish this. I knew my response looked somehow short

Thanks a lot guys for your feedback. Let me quickly provide some responses.

First of all I would like to clarify, all graphs provided are smoothed.

Originally Posted by DanDan

Interesting experiments bedope.

Sorry to disagree, but I have seen many recommendations to place damping at, near, or in, the holes of Helmholtz panels. At a position of such high particle velocities very little is needed. Fabric across the holes is often enough.
I would have thought the same principles would apply to a single holed trap.

Yep, thats what I read about internal dampening. Thats what I heard. Thats what I was advised. Thats what I concluded by myself. Yet, it does not work.

I tested this option before and it almost rendered HH useless. Even now, just following your comment and to triple check it, I modified all openings in my current HH boxes - covered backs of each tube with very transparent fabric. And I mean very transparent fabric, really loosely woven maybe with 50% coverage. Results are more then twice worse, in all SPL and waterfall. I dont know why its like that, but by working with my previous test box I got to a conclusion that absorbent at sides of the cavity seems to work the best.

BTW, absorption shown on the photo has been modified before publishing previous results. Its half of the thickness and spread more evenly around cavity.

Originally Posted by DanDan

Interesting experiments bedope.
REW allows Waterfalls up to 1500mS. There is no need to restrict your views to 600mS.

Actually there is one good reason 600ms point is right in the middle of time when changes supposed to happen, which mean that point should show the biggest change. So any progress, even minimal is the best traced in that point. Angle of the slope is the highest there. I also generate 1200ms waterfalls and those are great, but also those are more "big picture" type of data.

Originally Posted by DanDan

Interesting experiments bedope.
The SPL level in the Room should be say 80-90dB C Slow. Enough for you to need ear protection.

Thats too high for me. I never listen at that level. 60-70db, thats all. I hear your argument about higher levels, but I aim at practical solution for practical listening levels. For me 70db is max in most cases and effectively 40dB dynamic range (ambient noise at around 30dB). I also play trumpet and never reach theoretical 110dB. but can easily play at 40-50dB. I am a quiet guy

Originally Posted by DanDan

Interesting experiments bedope.
It is critical to have the microphone and speakers in the same position for before and after tests. For the modal work you are doing, corner placements would likely be best.
Even a 50mm change of position can cause large LF response changes.

Yep, I am aware of that. All test shown so far are with mic in pretty much the same position +/- 5mm and HH boxes in the same position +/-2mm. You should see my diary I recognize importance of close enough measuring condition to eliminate as many variables as possible.

bedope

[DanDan]

I don't think a 40dB measurement range, which I would say is optimistic, is good for decay estimations. If your ambient is indeed 30dB, then you would need 90dB to estimate RT60. Or thereabouts.
The high level is for decay measurement purposes only, nothing to do with real or practical listening levels. As you are focussed on decay here it seem a wast of your great efforts to use a faulty measurement practice.
Interesting point regarding 600mS but I am afraid I can't buy it.
Decays are averaged over 60dB one way or another.
You might Zip and post one of your more successful measurements here.
Playing around with the view parameters can snatch defeat from the jaws of victory.......

DD

[bedope]

Originally Posted by DanDan

I don't think a 40dB measurement range, which I would say is optimistic, is good for decay estimations. If your ambient is indeed 30dB, then you would need 90dB to estimate RT60. Or thereabouts.
DD

Well, I followed REW tutorial with setup: Check Levels

Its says: "...your SPL meter shows a level of around 75dB". Thats 15dB short of recommended 90dB. So what does it mean, REW tutorial is faulty and recommended setup is not efficient enough to generate any RT60, T20, T30 nor waterfall by REW?



bedope

[boggy]

Originally Posted by bedope

.....
Yep, thats what I read about internal dampening. Thats what I heard. Thats what I was advised. Thats what I concluded by myself. Yet, it does not work.

I tested this option before and it almost rendered HH useless........

This is expected (at least to me)... you CAN change Q factor of acoustical HH filter if you add resistance in tube, but for what price? If you stop air circulating in tube (with resistive barrier), so filter is degraded, and not enough particle movement energy enter in the box...
All damping need to be in the box... so... no later tuning.

Originally Posted by bedope

Actually there is one good reason 600ms point is right in the middle of time when changes supposed to happen, which mean that point should show the biggest change. So any progress, even minimal is the best traced in that point. Angle of the slope is the highest there. I also generate 1200ms waterfalls and those are great, but also those are more "big picture" type of data.

Clever reason. Especially for waterfall graphs, where is very hard to see clear progress sometimes...

Originally Posted by DanDan

....If your ambient is indeed 30dB, then you would need 90dB to estimate RT60......

It is not needed to have minimum 60dB dynamic to measure RT60... if we measure slope of decay in 20-40dB dynamic range, that is enough to extrapolate where will be level in the room if we will have full 60dB.

[bedope]

Originally Posted by bedope

Well, I followed REW tutorial with setup: Check Levels

Its says: "...your SPL meter shows a level of around 75dB". Thats 15dB short of recommended 90dB. So what does it mean, REW tutorial is faulty and recommended setup is not efficient enough to generate any RT60, T20, T30 nor waterfall by REW?



bedope

Just to follow this logic and digging a bit more.

If we really need 60dB dynamic range to measure RT60 and calculate waterfalls accurately, measuring equipment should excess 60dB dynamic range to be accurate, right? Because surely linearity at the edges of the range we would need something with 70dB at least. Right?

So lets take a look at equipment recommended for REW.

1. Galaxy 140 (which I have). It has 3 measuring ranges: 32-80, 50-100, 80-100. All of them is are 50dB range, way to short. Not to mention there no range to measure our range of interest 40-100dB.

2. Behringer ECM8000, sensitivity -60dB. So even this mic is below our safe standard.

3. RadioShack - thats even worse then any above.


So, following this logic there is not semi-pro equipment that is able to measure RT60 in any confident manner. AND if that the case, literally thousand of measurement I have seen on the net go our of window, just like mine

Sounds a bit extreme for me

HOWEVER... assuming REW uses Fourier transform to calculate RT60 and so waterfall, one wonder about all those thousands graphs put on the internet IF REW uses that formula to calculate waterfalls. (?)

bedope