advertisement


Measuring the effect of acoustic treatments in the room

Is there a delay/timing or whatever option in REW where you can largely dial-out the effect of the room across the whole bass area, say 20Hz through to say 600Hz? I’m curious whether part of what you are unhappy with is some kind of cancellation artefact of the 66’s ABR design.

I certainly like 66s based on memory, but there are always imperfections with this kind of loading as at certain frequencies the air-spring of the cabinet volume along with the mass of the ABR means the main driver is moving ‘out’ and the ABR ‘in’, therefore cancelling rather than doubling or augmenting bass. Celestion knew what they were doing, and they are clearly good speakers, but the passage of 45 years or whatever may have impacted the surrounds, spider etc and I guess there is a possibility this artefact has shifted in frequency or consistency from its original specification.

To be honest I have issues with anything aside from sealed boxes, bass horns or dipoles. I do not like ports, I’m not a huge fan of transmission lines, and ABRs would again not be my conceptual choice, though they can offer some benefits over ports. I certainly view them as being far closer to a port than a sealed box in the way they behave and interact with a room. Despite saying all this you’ll find a large pair of Lockwood cabs in my living room, they are technically ‘aperiodic’ loading, a very damped low-q port that behaves more like an open back guitar cab to my mind, i.e. it doesn’t really create or reinforce bass. Again not ideal, but they work in my room as whatever that big hole is doing it is nowhere near my primary 41Hz room node!

I’d love to see measurements for a good hefty sealed box speaker in your room, e.g. NS1000s, Gale 401s, even a pair of 44s, something like that, and also an anechoic plot for the 66 to see what its ABR bass arrangement is actually producing, i.e. where exactly it is cancelling, doubling or impacting phase, as I suspect it will be somewhere.
 
Is there a delay/timing or whatever option in REW where you can largely dial-out the effect of the room across the whole bass area, say 20Hz through to say 600Hz? I’m curious whether part of what you are unhappy with is some kind of cancellation artefact of the 66’s ABR design.

I'm not sure. I can gate out early reflections in the measurements but doing this reduces the resolution of the FR and makes the measurement accurate only down to around 200Hz-300Hz-ish.

I certainly like 66s based on memory, but there are always imperfections with this kind of loading as at certain frequencies the air-spring of the cabinet volume along with the mass of the ABR means the main driver is moving ‘out’ and the ABR ‘in’, therefore cancelling rather than doubling or augmenting bass. Celestion knew what they were doing, and they are clearly good speakers, but the passage of 45 years or whatever may have impacted the surrounds, spider etc and I guess there is a possibility this artefact has shifted in frequency or consistency from its original specification.

To be honest I have issues with anything aside from sealed boxes, bass horns or dipoles. I do not like ports, I’m not a huge fan of transmission lines, and ABRs would again not be my conceptual choice, though they can offer some benefits over ports. I certainly view them as being far closer to a port than a sealed box in the way they behave and interact with a room. Despite saying all this you’ll find a large pair of Lockwood cabs in my living room, they are technically ‘aperiodic’ loading, a very damped low-q port that behaves more like an open back guitar cab to my mind, i.e. it doesn’t really create or reinforce bass. Again not ideal, but they work in my room as whatever that big hole is doing it is nowhere near my primary 41Hz room node!

I’d love to see measurements for a good hefty sealed box speaker in your room, e.g. NS1000s, Gale 401s, even a pair of 44s, something like that, and also an anechoic plot for the 66 to see what its ABR bass arrangement is actually producing, i.e. where exactly it is cancelling, doubling or impacting phase, as I suspect it will be somewhere.

I'll get round to measuring some of my other speakers in this new location at some point, though probably not my Edinburghs as it's too much of a faff for my dad to heave those back into my room! My IMF Professional Monitor mk3 are on castors and easy to manoevre but I have a feeling I'd need to double up on the number of GIK 42Hz Scopus traps in my room with those! My other IMFs have better-damped bass but are on stands and are a real bugger move. The only big sealed box I have is the Lancaster which is fitted with MG15s, I can try these at some point. I'll also try measuring the Dittons at different heights to see how that affects the 100Hz-300Hz cancellations.

BTW - Ditton 66 anechoic measurement here: https://we.tl/t-rjlORWVGW7
 
Looks like the cancellation point is around 200Hz. It’s a very nice response, and I suspect that 200Hz dip is deliberately engineered-in to aid with wall proximity placement. Staggeringly inefficient though, 90db for 4.8W?! That is insane! Worse than say LS3/5As!
 
Looks like the cancellation point is around 200Hz. It’s a very nice response, and I suspect that 200Hz dip is deliberately engineered-in to aid with wall proximity placement. Staggeringly inefficient though, 90db for 4.8W?! That is insane! Worse than say LS3/5As!
I suspect that efficiency figure is a misprint as I only need to turn the volume of my amp up by a couple of dBs to have the Dittons match the SPL of my Tannoys. Also the Ditton is a 4Ω-8Ω load so perhaps that makes the Wattage value appear lower, if they were rated by voltage sensitivity I suspect the dB/V value would be a lot higher. This is a measurement I took of my Dittons and S3/5R2 last year with the amp's volume setting held constant. The Dittons are on average 8dB louder in the mid and high frequencies for the same volume setting, but the amp will of course be delivering twice the power into the frequencies rated at 4Ω.
 
To get an idea of what adding more bass trapping to the corners will offer me, I filled in some of the space behind my inverted GIK corner traps with wedges of RS45 rockwool. I didn't have enough to match the height of the GIK traps, only enough for around a 75cm high stack in each of the four corners. I also tried putting all the RS45 in the two front corners and then moved it all to the two rear corners to see which made the biggest difference.

These measurements were taken with the speakers against the front wall and the mic in the front half of the room (240cm from the rear wall), because this location produced one of the smoothest frequency responses when I measured without any treatments in the room:

attachment.php


As I explained a few posts back, the FR between 100Hz-400Hz is poorer when the corner treatments are in place:

attachment.php


This is what happens when I backfill the front, rear, and all corners with additional bass trapping. The differences are too small to obtain reliable RT60 and waterfall data from, but we can use the reduction in FR peaks as a proxy for reduced ringing in the time domain:

attachment.php


attachment.php


attachment.php


attachment.php


The plots suggest that the 41Hz axial length modal peak is reduced more by putting the treatment on the front wall than the rear wall ('front corners' performance at 41Hz is on a par with 'all corners' performance). I'd be interested to know if this because the front wall is the wall that's closest to the speakers (the theory being that bass absorption is maximised the closer to the source you place the absorption), or if it's because the listening position is in the front half of the room (i.e. would the results be reversed if the listening position was in the rear half of the room)?

The 53Hz axial height mode appears to be most affected by adding treatment to the rear corners, which is interesting as I expected it to be reduced the most by distributing the additional absorption to all four corners of the room. Perhaps the height at which the treatment was placed has a bearing on this (I sat the rockwool on shelves that were approximately 40cm above floor level).

The 80Hz-120Hz region appears to be improved more by treating the front corners over the rear corners, although performance here is arguably the smoothest when the additional treatment is distributed equally among all four corners.

I'll need to place an order for more rockwool to take this experiment further!
 
Last edited:
I certainly like 66s based on memory, but there are always imperfections with this kind of loading as at certain frequencies the air-spring of the cabinet volume along with the mass of the ABR means the main driver is moving ‘out’ and the ABR ‘in’, therefore cancelling rather than doubling or augmenting bass. Celestion knew what they were doing, and they are clearly good speakers, but the passage of 45 years or whatever may have impacted the surrounds, spider etc and I guess there is a possibility this artefact has shifted in frequency or consistency from its original specification..

Measurements I've previously taken of other speakers against the (previous) front wall suggests that the 150Hz dip isn't unique to the Ditton, e.g. see here and here.

Below are the individual, non-smoothed responses of left and right Dittons in their current location, which is more or less the same as they are in this pic but about 20cm or so closer together. As you can see, the left speaker does not suffer as much from cancellation at 150Hz. The only difference in the front wall geometry is the backwave from the left speaker is partially radiating into a cupboard stuffed with rockwool whereas the right speaker has a solid wall, but the bay window on the left side wall could also have something to do with it.

01-L-R-speakers-45cm-from-front-wall-mic-240cm-from-rear-wall.jpg


There is a glimmer of hope in that adding additional absorption to the front corners of the room does improve the 150Hz in the right speaker a little. Hopefully this is an indication that it will continue to improve if I keep adding more treatment to this area.

02-L-R-speakers-45cm-from-front-wall-mic-240cm-from-rear-wall-rear-corners-backfilled.jpg


03-L-R-speakers-45cm-from-front-wall-mic-240cm-from-rear-wall-front-corners-backfilled.jpg


04-L-R-speakers-45cm-from-front-wall-mic-240cm-from-rear-wall-front-rear-corners-backfilled.jpg
 
Could it partly be due to floor bounce cancellation?

https://mehlau.net/audio/floorbounce/

According to that calculator the bass driver would need to be 120cm above the floor to cause a cancellation at 150Hz with a seating distance of 155cm and a seating height of 103cm. The active bass driver in the Dittons is currently only 56cm (approx) above the floor and the ABR only 25cm (approx) above the floor (measured from the floor to the centre of each driver), which should cause cancellations at 286Hz and 625Hz, respectively.

I will be experimenting with stands at some point, so that should hopefully tell us if floor bounce is a contributing factor.
 
Listening in the "front half" of the room is a novel experience, not only because I'm not used to sitting this close to these speakers and having such a large amount of open space behind me, but also because the speakers are sounding leaner than they ever have! Some familiar songs are lacking that hefty 'thud' that I'm used to hearing from the kick drum. I suspect the kick in these tracks is tuned to 45Hz, which corresponds to the most significant null in the low bass that I have at this listening position. This null is also present in the rear half of the room but not to the same extent. What's also interesting is the speakers' roll-off below the 41Hz axial length mode is steeper in the front half of the room than in the rear. (See graphs in post #18).

After EQ I'm left with a differential of just 4dB between the lowest and highest frequencies, which is significantly less than the 8dB downward-sloping response than I had when I was listening in the rear half of the room and had the speakers firing down the shorter dimension. Re-installing the grilles on the Dittons will attenuate the high frequencies by around -2dB (I removed them for measurement purposes in case they caused spurious reflections in the ETC plots and haven't refitted them yet). However, I suspect I may still end up reverting to my natural habitat of listening in the rear half of the room to increase the ratio of reflected sound to direct sound, which is a pity because I was hoping to finally have space after all these years for a small worktable behind my listening seat that is central to my speakers! :( An alternative would of course be to use EQ to artificially create the desired downward sloping response, but even I have to draw the line somewhere on the overuse of EQ! :D Reintroducing my subs is also still under consideration but I was intending to use them only to fill in below 40Hz so that I don't need to introduce additional DSP.

01-L-R-Avg-with-vs-without-EQ-sitting-in-front-half-of-room-spkrs-45cm-from-front-wall-mic-240cm-f.jpg
 
An alternative would of course be to use EQ to artificially create the desired downward sloping response, but even I have to draw the line somewhere on the overuse of EQ! :D

Also, above around 500Hz you should EQ the anechoic response, not the in-room response.
 
Four packs of Rocksilk RS45 arrived today. One has already been cut up, bagged and stuffed into 40x40x40cm boxes, the others will be done in the coming days assuming the incoming heatwave isn't too extreme.

attachment.php


My parents can't believe the four vertical corners of my room require this much rockwool to run floor-to-ceiling. Imagine their horror when I told them I'd have needed an additional two packs if I wasn't incorporating my existing TriTraps into the corner treatment! :D

The TriTraps are going at the bottom of the stack facing into each other to make a 40x40x120cm cuboid and I'm stacking five 40x40x40cm rockwool-stuffed boxes on top which will take me almost to the top of the 325cm high ceiling.

I'll take REW measurements and then replace the 20 boxes of RS45 with 20 boxes of lighter density 'Acoustic Roll' glasswool and repeat the measurements to see which product has better absorption at 40Hz-50Hz.

The challenge will then be to find a local carpenter capable of building me 50cm x 50cm x 160cm curved birch plywood-faced corner frames that I'll stuff with the better performing absorber material. The curved plywood should hopefully look a bit better than a stack of cardboard boxes!... ;)
 
Last edited:
Bolstering the amount of bass traps in the four vertical corners of my room could be summed up as “a lot of work for little reward”…

First the good points:
- Even before listening to music or taking measurements you can tell that the room has a calmer feel to it, it’s still lively but there isn’t as much reinforcement in the midrange when you speak or clap your hands.
- System sounds cleaner and more precise, the treble in particular has become crisper and there is less midrange forwardness.
- Decay times lowered, now more balanced in the mids and treble, I’m getting closer to my 400ms target in the mids without the highs being choked off.
- 53Hz axial height mode has been reduced.
- 65Hz dip has filled in.

Now the bad points:
- Upper bass and lower mid frequencies are further attenuated, exacerbating the recessed area between 150Hz-350Hz (I was expecting this to happen but hoped for a miracle anyway…).
- 78Hz dip has deepened.

And finally the inexplicable:
- No change to the amplitude of the 41Hz axial length mode.

I find this bizarre given the huge amount of treatment I’ve added to the corners (2.5 times more rockwool than before). Whenever I’ve added porous absorption to the length of the room I’ve always seen an incremental reduction in the 41Hz peak, the more absorption I add the more the peak is reduced. Why not this time?! Is it because the rockwool is in plastic bags inside cardboard boxes and not exposed to the air? Or a problem with the density? I find it strange that it’s working at 53Hz and 65Hz but not 41Hz. If it was going to be ineffective on a particular room mode I’d have expected it to be at 53Hz given that this height mode needs to penetrate through 3 metres of 45kg/m3, whereas the 41Hz length mode only needs to penetrate 40cm. However, the Porous Absorption Calculator shows that absorption at 40Hz-60Hz is very similar between 45kg/m3 rockwool at 40cm and 300cm depths:

attachment.php


The following measurements are with the corner treatment configured as per this photo, then as per this photo, and finally as per this photo. (Note the speakers were positioned hard against the front wall for all measurements, I just forgot to take up-to-date pics of the first two configurations!).

Detailed data is shown for right speaker only. (L&R Avg frequency responses are shown at the end).

attachment.php


attachment.php


attachment.php


attachment.php


attachment.php


attachment.php


attachment.php


attachment.php


attachment.php


attachment.php


attachment.php


attachment.php



attachment.php


attachment.php
 
Last edited:
I’m pretty sure the plastic and boxes are affecting performance. When building acoustic panels they always say to pick a fabric you can blow air through. Not sure about your lungs but i have a hard time blowing through cardboard and plastic ;)
 
I’m pretty sure the plastic and boxes are affecting performance. When building acoustic panels they always say to pick a fabric you can blow air through. Not sure about your lungs but i have a hard time blowing through cardboard and plastic ;)
This is also what I learned back in 2013 when I was first researching acoustic treatment, but in the years since I've learned that low frequencies can still be absorbed through plastic provided the absorption material behind it isn't compressed. (Indeed if the plastic is thin enough then middle and high frequencies will also pass through it into the absorption behind!).

This is, however, the first time I've tried cardboard boxes, and my reason for doing so was to create a low-pass filter so that middle and high frequencies would be reflected back into the room. Apparently, if the face of an absorber is stiff enough, only sound waves whose wavelengths are shorter than the width of the face will be reflected, those with wavelengths longer will diffract around and also be absorbed. My boxes are 42cm wide, which sets the low-pass filter at a theoretical 820Hz. However, I was told on Gearspace that cardboard is partially absorptive at mid and high frequencies and that I should instead use a stiffer plywood-fronted enclosure. I wanted to test with cardboard first before going to the effort and expense of employing a carpenter to build plywood-faced corner traps. I'm having second thoughts now!
 
Last edited:
This is also what I learned back in 2013 when I was first researching acoustic treatment, but in the years since I've learned that low frequencies can still be absorbed through plastic provided the absorption material behind it isn't compressed. (Indeed if the plastic is thin enough then middle and high frequencies will also pass through it into the absorption behind!).

This is, however, the first time I've tried cardboard boxes, and my reason for doing so was to create a low-pass filter so that middle and high frequencies would be reflected back into the room. Apparently, if the face of an absorber is stiff enough, only sound waves whose wavelengths are shorter than the width of the face will be reflected, those with wavelengths longer will diffract around and also be absorbed. My boxes are 42cm wide, which sets the low-pass filter at a theoretical 820Hz. However, I was told on Gearspace that cardboard is partially absorptive at mid and high frequencies and that I should instead use a stiffer plywood-fronted enclosure. I wanted to test with cardboard first before going to the effort and expense of employing a carpenter to build plywood-faced corner traps. I'm having second thoughts now!
I never played with those kinds of panels, i sit quite close and on axis so not really needed for my usecase. All my panels are fabric and rockwool :)
Maybe the cardboard and plastic combines into some kind of membrane absorber effect?
 
I’m pretty sure the plastic and boxes are affecting performance. When building acoustic panels they always say to pick a fabric you can blow air through. Not sure about your lungs but i have a hard time blowing through cardboard and plastic ;)

It would be interesting to see a comparison without the boxes and plastic.
 
I'll do a measurement without the boxes, but not without the plastic, - too much risk of filling my room with airborne fibres when stacking/unstacking.

I doubt that the plastic will make any difference below 1kHz. The flexing carboard may, but it is still a relatively small area.
Something else is at play here, probably a narrow-band absorption issue.

Maybe you could keep the stacked-up boxes and remove the stuff from the cupboard?
 
Porous absorbers need high air particle speed. The slowest speeds will typically be seen next to a boundary at its associated bass mode frequency. (In contrast that's the ideal place for a membrane absorber, which is a pressure absorber).

A porous absorber is better at absorbing non-mode frequencies at boundary (which you observe) and bass mode frequency away from its boundary.

(As you know ToTo Man, membrane is recommended for bass, this is another reason why.)
 
An old studio trick was always to roll up old carpets and stand them up in the corners of the room. I tried it when I was having the bad room boom issue with the MEGs ages ago and it did attenuate it noticeably (though not enough not to irritate). I took them out again when I went back to the Tannoys as they don’t excite the room in the same way at all.
 


advertisement


Back
Top