How do you like your multisub setup? Did you make any additional absorption on any walls?
What do you have above the grid/drop ceiling? With the curves and abundance of asymmetry, isn't that still a large surface parallel to the floor? If it's built anything like others I've seen, you could hide an awful lot of bass trap up there.
Cheers
edit: Almost forgot -- for pics, try using the Attachment facility (the little paper clip icon) in the Advanced editor. Dunno if others have the same problem, but I have to fire up a different browser (Firefox--eats another ~400 MBytes) and Copy/Paste the URL from your link into it to see 'em. Skipping around to several different ones arbitrarily is simply impractical.
Cheers
edit: Almost forgot -- for pics, try using the Attachment facility (the little paper clip icon) in the Advanced editor. Dunno if others have the same problem, but I have to fire up a different browser (Firefox--eats another ~400 MBytes) and Copy/Paste the URL from your link into it to see 'em. Skipping around to several different ones arbitrarily is simply impractical.
Last edited:
Hello Rick,
above the the drop ceiling is concrete plate. Yes, all walls are parallel to each other, also the ceiling is parallel to the floor surface. Unfotunatelly the surfaces on the photo are curved because of the panorama mode on my phone when I took the photo.
Only the 1st reflection points are treated on the ceiling with rock wool insulation. Still need to work on that. The room is rented, I´m not the owner of that building so some large encroachment in the room is not possible. I´m planning to replace the drop ceiling plates with some additional diffusion elements but that will not solve room modes of the low end region.
I would like ask you guys if there is any chance to place absorbers on the floor insted on ceiling to treat that modal region?
PS: Sorry, I´ve never attached the photo here before so thak you for the advice, I´ll try to do that.
above the the drop ceiling is concrete plate. Yes, all walls are parallel to each other, also the ceiling is parallel to the floor surface. Unfotunatelly the surfaces on the photo are curved because of the panorama mode on my phone when I took the photo.
Only the 1st reflection points are treated on the ceiling with rock wool insulation. Still need to work on that. The room is rented, I´m not the owner of that building so some large encroachment in the room is not possible. I´m planning to replace the drop ceiling plates with some additional diffusion elements but that will not solve room modes of the low end region.
I would like ask you guys if there is any chance to place absorbers on the floor insted on ceiling to treat that modal region?
PS: Sorry, I´ve never attached the photo here before so thak you for the advice, I´ll try to do that.
Can't problems with the lowest modes be dealt with delayed + phase inverted subs at the back wall? Especially in rooms with a high length vs width ratio?
What are the cons, besides lower total efficiency? Altering the wall structure usually isn't practical - if not forbidden
- in a domestic environment.If you mean two independent sub inputs from my receiver yes I do. I’m also using these same subs in my room. This is a studio. 1000 square feet. The room measurements are pretty similar to this too.
I’m in process of adding a 12” with a lot more power so not sure what’s going to happen to the others yet. I wanted one directly in front of me under the ottoman I built.
Thanks Dr. Geddes. I have always wanted to know if an airspace was required behind the CLD wall. It seems it is recommended. My situation is a finished wall to which I would like to add low frequency absorption. I should probably add a 2 X 4 or 2 X 6 framed airspace on the front of the finished wall and then the resilient channel/CLD wall. Correct?
That´s awesome dr. Geddes, thank you.
Well my place is rented so some kind of big and heavy absorbers is the only way...
Basically that is correct.
The air space acts as a spring, as long as the edges are sealed, which they need to be. If this space is too small then the wall is too stiff.
The means of absorption is two-fold. First the energy is dissipated through the CLD, but the panel must actually move for this to occur. But then the sound radiatiates again on the back side due to the panels motion. This sound needs to be dissipated as well, or it will just reflect back through the panels. So the air space should be heavily damped as well.
Hi, I would like to optimize my 7.4 system (4 = 2 bass reflex boxes + 2 Linkwitz dipoles) in a room 5 x 7 x 3.75 m3. T_60 in the middle frequency range is 0.3 ... 0.4 s. Large damping elements for the bass range should not be installed. There are no long corridors, ventilation channels etc. with exotic reverb characteristics (=> waterfall diagram not so important?).
Both film material with LEF signals and stereo music should be listened to. The goal is less a studio-like linear frequency response than
a) A clear dynamic (punchy) bass e.g. for heavy metal. At the moment the bass is not very accentuated.
b) not to be " pressed down " by organ music with longer low tones. At the moment there are resonances where you can really listen to how the resonance builds up and covers the "breathing" of the sound in the concert hall or church.
For reference I have a stereo system in which I can use FIR filters. FR and phase are optimized there for the listening position, in the bass they are not boosted. But in practice 90% prefer the uncorrected bass. In the long run the corrected version sounds more exhausting in the bass range: too unfree, not breathing and open, slightly over-regulated, unbelievable. Maybe this is related to the psychoacoustic effect that the ear wants to perceive the visible listening space, maybe the increase suggested by Earl Geddes is missing (3..6 db rise per decade from 200 Hz downwards). My conclusion is that < 200 Hz is no longer a studio-like linear FR.
The current goal is to find as systematically as possible what is possible with the current material. There will probably not be a procedure that leads to an optimal result.
SETUP
Receiver: Denon, the subwoofer channel gets a LFE + main signal. XO 80 Hz, filter details unknown, probably 12 dB/oct.
Besides receiver and signal sources the following components are involved:
1) Front L+R Magnat Chromnium Style (with BR system up to about 30 Hz). Set as "Large" in the receiver, so they also get the full frequency response - unlike the subs R/L separated. So XO in the receiver (Denon) is basically an LP for the subs.
2) Bass Reflex Boxes (in future "BR") in the edge of the room under the shelf - they should stay there. (self made 60 l, chassis Seas Prestige CA26RFX (H1305), Free Air Res. 29 Hz, Q_t 0,34 Installation and cabinet resonance
tuned so that they do not hit the 33 Hz room mode).
3) 2 Linkwitz dipoles (future "LW") with 12" chassis (95 dB/W). 1 near the wall, 1 in the middle of the room, later under the coffee table.
4) Behringer DCX 2496 (3 input and 6 output channels phase, delay, EQ... adjustable), 4-channel Amp Img Stageline STA 2000D. Currently 1 In / 4 Out are in operation, you can define e.g. an EQ-filter for In, which will be applied to all 4 Out etc.
In all measurements, Front L and Front R are included for the temporal reference. The double measurements (Sub+L, Sub+R) are also good for a consistency check. For all measurements I set the bass channels 10 dB louder than Front L+R to get (with sufficient accuracy) isolated data only for the Sub's. Previous sample measurements (results not usable for MSO due to lack of time reference) only with the subs's show that the results are usable: The differences in FR between Sub alone, Sub + R and Sub + L are much smaller than the differences between different listening positions or different Sub's.
A very nice feature of MSO is that you can long-click the +/- buttons of the delay filters and the calculated sum FR will change in real time:
animated gif upload was only possible using zip:
View attachment mso_delay_screencam_4.zip
Sum of 2 BR Speakers: Measured vs. individually played and calculated with variable delay for one of them. "1"... "3" are the listening positions, "R", "L" the front speakers for the temporal reference "G+S" (garden, street) are the names of the BR speakers. "0.00m" means that the temporal reference for the phase starts at the maximum of the IR.
So you can watch how the calculated FR of BR1[with Delay]+BR2 approaches and moves away from the measured FR BR1+BR2 and get an intuition that deep narrow dips are noticeable but rather artifacts of calculation, while sums behave comprehensible. Finally the sound changes rather little if you change the delay a few ms.
In the animated version of the picture you can see how the dips become comb filters that shift to lower frequencies with increasing delay. Random calculations with wavelengths and distances show that the values are plausible.
As a physicist in the field of computational physics I have the experience that the optimization of such systems (with a very heterogeneous parameter landscape and very well understandable behaviour during parameter changes) with reasonably given start values and constraints gives the best results. The "machine to let everything do" is imho. mostly only at first glance, before knowing the artifacts, "purer" (discussing this in detail would fill many pages). So I would suggest the following:
BR Speakers
If possible, the position BR at the edge of the room should not be changed (matching the shelf, see picture). My ambition is to get the system with the speakers in the edge of the room acoustically under control - do you think this is possible? They are mainly responsible for the "heavy work" in the low bass range, since the other speakers cannot generate any sound pressure there. They should be operated up to 40 or 50 Hz (12 dB) to cut off localizable frequencies safely.
Proposal for MSO specifications:
1) The resonance @33 Hz should be lowered by EQ, constraint: -5...-10 dB @33 Hz
2) How many filters should be used (many narrowband or few broadband filters?)
Is there a danger of audible artifacts (distortion) with narrow-band filters? I would suggest a maximum of 4.
3) Both BR speakers should have the same sound level.
4) Delay: Has already been calculated from the average group delay, depending on the concept:
EITHER EITHER in phase with the Quadral (delay time difference to listening positions adjusted to 10..20 ms to Front R and L, maybe some fine tuning for smooth XO)
OR counter-phase (as DBA) to Front R and L (i.e. minus and approx. 5 m transit time difference).
I would only allow +/- 10 ms variation by MSO
Linkwitz Dipoles
The LW speakers are about 10 dB weaker (despite significantly more watts). I would fix this for MSO and adjust the amplifiers afterwards if necessary. First listening tests show that they contribute less to the bass, but the sound seems to float more in the middle of the room and the bass gets lightness, which sounds good. Outdoor measurements show that not much is to be expected below 35 Hz. I would suggest something like the following:
1) LW polarity and phase vs. level in the XO range: delay +/- 15 ms free, phase also free optimize
2) FR 2 or 3 filters with fixed frequency and bandwidth at different peaks, maybe 10% freedom in frequency if peaks move a little bit due to overlapping.
What do you think about this?
TARGET:
Target frequency range: up to approx. 80 Hz separately for the seats, then a further optimization for the sum of all listening positions: 80 Hz is 1/4 of the wavelength approx. 1m the sofa is over 3m wide, the 3 seats may vary.
No target curve (depending on source material LFE, stereo) the volume has to be readjusted anyway.
Imho he main task is "mastering BR speakers in room edge". I think, in the end you have to take energy from the resonances via EQ and counteract it with the LW dipoes.
Do you have experience with a DBA setup? Would the use of higher quality subwoofers near the front R and L improve the characteristics so much that these attempts are not worthwhile?
Both film material with LEF signals and stereo music should be listened to. The goal is less a studio-like linear frequency response than
a) A clear dynamic (punchy) bass e.g. for heavy metal. At the moment the bass is not very accentuated.
b) not to be " pressed down " by organ music with longer low tones. At the moment there are resonances where you can really listen to how the resonance builds up and covers the "breathing" of the sound in the concert hall or church.
For reference I have a stereo system in which I can use FIR filters. FR and phase are optimized there for the listening position, in the bass they are not boosted. But in practice 90% prefer the uncorrected bass. In the long run the corrected version sounds more exhausting in the bass range: too unfree, not breathing and open, slightly over-regulated, unbelievable. Maybe this is related to the psychoacoustic effect that the ear wants to perceive the visible listening space, maybe the increase suggested by Earl Geddes is missing (3..6 db rise per decade from 200 Hz downwards). My conclusion is that < 200 Hz is no longer a studio-like linear FR.
The current goal is to find as systematically as possible what is possible with the current material. There will probably not be a procedure that leads to an optimal result.
SETUP
Receiver: Denon, the subwoofer channel gets a LFE + main signal. XO 80 Hz, filter details unknown, probably 12 dB/oct.
Besides receiver and signal sources the following components are involved:
1) Front L+R Magnat Chromnium Style (with BR system up to about 30 Hz). Set as "Large" in the receiver, so they also get the full frequency response - unlike the subs R/L separated. So XO in the receiver (Denon) is basically an LP for the subs.
2) Bass Reflex Boxes (in future "BR") in the edge of the room under the shelf - they should stay there. (self made 60 l, chassis Seas Prestige CA26RFX (H1305), Free Air Res. 29 Hz, Q_t 0,34 Installation and cabinet resonance
tuned so that they do not hit the 33 Hz room mode).
3) 2 Linkwitz dipoles (future "LW") with 12" chassis (95 dB/W). 1 near the wall, 1 in the middle of the room, later under the coffee table.
4) Behringer DCX 2496 (3 input and 6 output channels phase, delay, EQ... adjustable), 4-channel Amp Img Stageline STA 2000D. Currently 1 In / 4 Out are in operation, you can define e.g. an EQ-filter for In, which will be applied to all 4 Out etc.
In all measurements, Front L and Front R are included for the temporal reference. The double measurements (Sub+L, Sub+R) are also good for a consistency check. For all measurements I set the bass channels 10 dB louder than Front L+R to get (with sufficient accuracy) isolated data only for the Sub's. Previous sample measurements (results not usable for MSO due to lack of time reference) only with the subs's show that the results are usable: The differences in FR between Sub alone, Sub + R and Sub + L are much smaller than the differences between different listening positions or different Sub's.
A very nice feature of MSO is that you can long-click the +/- buttons of the delay filters and the calculated sum FR will change in real time:
animated gif upload was only possible using zip:
View attachment mso_delay_screencam_4.zip
Sum of 2 BR Speakers: Measured vs. individually played and calculated with variable delay for one of them. "1"... "3" are the listening positions, "R", "L" the front speakers for the temporal reference "G+S" (garden, street) are the names of the BR speakers. "0.00m" means that the temporal reference for the phase starts at the maximum of the IR.
So you can watch how the calculated FR of BR1[with Delay]+BR2 approaches and moves away from the measured FR BR1+BR2 and get an intuition that deep narrow dips are noticeable but rather artifacts of calculation, while sums behave comprehensible. Finally the sound changes rather little if you change the delay a few ms.
In the animated version of the picture you can see how the dips become comb filters that shift to lower frequencies with increasing delay. Random calculations with wavelengths and distances show that the values are plausible.
As a physicist in the field of computational physics I have the experience that the optimization of such systems (with a very heterogeneous parameter landscape and very well understandable behaviour during parameter changes) with reasonably given start values and constraints gives the best results. The "machine to let everything do" is imho. mostly only at first glance, before knowing the artifacts, "purer" (discussing this in detail would fill many pages). So I would suggest the following:
BR Speakers
If possible, the position BR at the edge of the room should not be changed (matching the shelf, see picture). My ambition is to get the system with the speakers in the edge of the room acoustically under control - do you think this is possible? They are mainly responsible for the "heavy work" in the low bass range, since the other speakers cannot generate any sound pressure there. They should be operated up to 40 or 50 Hz (12 dB) to cut off localizable frequencies safely.
Proposal for MSO specifications:
1) The resonance @33 Hz should be lowered by EQ, constraint: -5...-10 dB @33 Hz
2) How many filters should be used (many narrowband or few broadband filters?)
Is there a danger of audible artifacts (distortion) with narrow-band filters? I would suggest a maximum of 4.
3) Both BR speakers should have the same sound level.
4) Delay: Has already been calculated from the average group delay, depending on the concept:
EITHER EITHER in phase with the Quadral (delay time difference to listening positions adjusted to 10..20 ms to Front R and L, maybe some fine tuning for smooth XO)
OR counter-phase (as DBA) to Front R and L (i.e. minus and approx. 5 m transit time difference).
I would only allow +/- 10 ms variation by MSO
Linkwitz Dipoles
The LW speakers are about 10 dB weaker (despite significantly more watts). I would fix this for MSO and adjust the amplifiers afterwards if necessary. First listening tests show that they contribute less to the bass, but the sound seems to float more in the middle of the room and the bass gets lightness, which sounds good. Outdoor measurements show that not much is to be expected below 35 Hz. I would suggest something like the following:
1) LW polarity and phase vs. level in the XO range: delay +/- 15 ms free, phase also free optimize
2) FR 2 or 3 filters with fixed frequency and bandwidth at different peaks, maybe 10% freedom in frequency if peaks move a little bit due to overlapping.
What do you think about this?
TARGET:
Target frequency range: up to approx. 80 Hz separately for the seats, then a further optimization for the sum of all listening positions: 80 Hz is 1/4 of the wavelength approx. 1m the sofa is over 3m wide, the 3 seats may vary.
No target curve (depending on source material LFE, stereo) the volume has to be readjusted anyway.
Imho he main task is "mastering BR speakers in room edge". I think, in the end you have to take energy from the resonances via EQ and counteract it with the LW dipoes.
Do you have experience with a DBA setup? Would the use of higher quality subwoofers near the front R and L improve the characteristics so much that these attempts are not worthwhile?
May i ask what kind of room build is this? im in the process of building a small listening room, and gathering as much as data as i can to make this happen with minimum issues....
"May i ask what kind of room build is this? im in the process of building a small listening room, and gathering as much as data as i can to make this happen with minimum issues.... "
Deserves a thread in the acoustics forum. But the moderators have hidden that someplace far away when it ought to be here. Pity.
B.
Deserves a thread in the acoustics forum. But the moderators have hidden that someplace far away when it ought to be here. Pity.
B.
Yes, far far away. Maybe as much as 650 pixels away.
Attachments
Hi Earl,
I'm looking at building another sub or 2 with 15" prosound drivers like the B&C 15TBX100 which I believe you use or used at one point in your subs?
Thiele Small calculators would predict a driver like the 15TBX100 to have an -F3 of around 88 Hz and be down 25 dB at 20 Hz. Obviously this has a lot less meaning to a sub used in-room and I'd expect once a lpf is applied and room effects are measured any boost required at 20 Hz would be much less. For your subs do you have to apply much, or any eq to increase the low end extension of your subs?
thanks,
Dan
I'm looking at building another sub or 2 with 15" prosound drivers like the B&C 15TBX100 which I believe you use or used at one point in your subs?
Thiele Small calculators would predict a driver like the 15TBX100 to have an -F3 of around 88 Hz and be down 25 dB at 20 Hz. Obviously this has a lot less meaning to a sub used in-room and I'd expect once a lpf is applied and room effects are measured any boost required at 20 Hz would be much less. For your subs do you have to apply much, or any eq to increase the low end extension of your subs?
thanks,
Dan
I do like the 15TBX100, or better the neo version. This driver has so much headroom in a closed box that it seems impossible to overload it. At 20 Hz, in my room, I actually need to cut the response as that is where the lowest mode is. Above that not much EQ is required - in terms of gain - certainly not enough to cause a problem.