do I need to address long low freq decay?

[ahmedmo1]

Hi there,

Hope all is well! Am trying to improve the in-room performance of the sounds systems in my room (bedroom). Would really appreciate the feedback of the more technically savvy (and not dogmatic) folks in this group.

I have two similar 2.1 systems in the same room, one is a desktop setup and the other is a TV setup:

Each contains a:

· A pair of Eve Audio SC208 powered studio monitors (AMT tweeter with 8” woofer)- freq response 35Hz -3dB

· Martin Logan 1000W 12” sealed sub on an Auralex subdude

· Topping DX3 Pro DAC used for both setups



The only differences are as follows:

The desktop setup has the speakers connected to isoacoustic stands on an hardwood desk. Appropriate ear height and equilateral triangle. I use Equalizer APO for EQ. The sub is behind the listening position but nearfield (not much farther than either speaker). The connection to the SUB is a via a wireless receiver. It introduces negligible latency vs. a wired connection.

• The TV setup has the speakers on adjustable-height floor stands and are connected to the sub via interconnect. It uses a Topping DX3 connected to a miniDSP 2x4HD for DSP/EQ (and DAC function). Theyre a few meters from the listening position so it’s not an equilateral triangle.

I have taken several measurements using a UMIK-1 calibrated mic and REW



I have 2 problems:

1. High reverb/decay at low frequency (~1 second decay times below 100Hz on desktop setup and ~0.7 seconds with the TV setup). My reverb/decay looks good above 100Hz.
2. Two 20-30dB nulls @ ~45Hz and ~85Hz in the TV setup. The frequency setup of the desktop setup actually looks great with some minor EQ.

I think the solution to #2 is to just add a second sub and smooth out the in-room frequency response. Thoughts?



My questions:

1. How much does reducing reverb/decay matter at low-frequencies? From my understanding, it’s hard to identify differences in decay times/reverb at low frequencies (100Hz and lower). Research papers suggest long decay times (well over 1 second) can’t be distinguished from shorter times at low frequencies. If that’s the case, what’s the use of using thick (very thick) bass traps to reduce low-freq reverb?

Research paper:

https://assets.ctfassets.net/4zjnzn...04_karjalainen_antsalo_makivirta_valimaki.pdf



2. Is the problem re. decay times that the difference in decay times across the frequency range? My decay time above 100Hz is short (under 0.4 seconds) but does the low-end decay times being longer mean it makes the overall sound less coherent?

3. Can changing the subwoofer location help with decay times/reverb for the desktop setup?

4. Does using dual subs help with low-end decay times?

 

[MRC01]

The decay time in the bass is always longer than in the mids & treble. You can see this in just about any CSD plot. In the bass the most important thing is how uniform it is across frequencies. If you see longer decay times at specific frequencies, this is what to tackle first.

For example at one point my room CSD looked like this:

The room had a mode at 74 Hz that I had to kill. Other than that you can see the decay times increasing in the bass more or less evenly is normal. Most rooms do benefit from broadband porous bass absorbers to make bass decay faster and more even across frequencies. But if you add too much, you can over-deaden the room.

 

[ahmedmo1]

Thanks! Good to know. Learning so much. I'm learning more as I read up on treatment/ room acoustics and it's actually quite fun.

My concern is as follows; doing the math on the size of traps needed to improve decay at low frequencies, seems like the porous bass absorbers that would work would also be prohibitively for a hobbyist, no? At under 100Hz, I'm looking at 1 foot or more of thickness in corner traps, no?

Finally, I was reading the below link that indicated that not only does multiple subs help with evening out the frequency response across multiple listening positions, it also helps even out the decay times across the frequency response. Is this accurate or incorrect in that it's missing key details?

https://www.audioholics.com/home-theater-calibration/bass-optimization-for-home-theater

 

[MRC01]

... My concern is as follows; doing the math on the size of traps needed to improve decay at low frequencies, seems like the porous bass absorbers that would work would also be prohibitively for a hobbyist, no? At under 100Hz, I'm looking at 1 foot or more of thickness in corner traps, no? ...

That's right. Broadband room treatments have to be sized according to the wavelength of sounds they affect. Alternately, you can have more narrow band treatments like resonators. Different tools for different jobs. You can also use EQ to squash peaks but EQ doesn't work as well for raising nulls.
As a hobbyist, you can build your own room treatments. For example the tube traps I built are 2' in diameter and 7' tall. Buying traps of this size would cost thousands, but I built mine for about $400 in parts and some of that cost was finish to make them look professional.

 

[Pio2001]

Hello,
First, remember that in low frequencies, 95 % of the audible problems come from equalization, and 5 % from long decay time (Numbers made up as I'm writing ).
Second, you need to look at a decay or waterfall or CSD graph, not to the RT60 curve. It is not accurate enough.
Third, if the long decay times occur at the Bottom of dips in the frequency response, then it doesn't matter.

This said, and if you still have a long decay time at a given frequency, that is not at the Bottom of a dip in the frequency response (unsmoothed), look for the group delay, minimal group delay and excess group delay curves, unsmoothed.
If the excess group delay is below 20 ms at all frequencies below 100 Hz, (except at the Bottom of dips), then everything is OK. It means that the variations in group delay are minimal. That's the best behaviour.

If you have more than 50 ms of excess group delay at a frequency below 100 Hz, that is not the Bottom of a dip, you might have an audible problem, and you can start investigate.

 

[ahmedmo1]

That's right. Broadband room treatments have to be sized according to the wavelength of sounds they affect. Alternately, you can have more narrow band treatments like resonators. Different tools for different jobs. You can also use EQ to squash peaks but EQ doesn't work as well for raising nulls.
As a hobbyist, you can build your own room treatments. For example the tube traps I built are 2' in diameter and 7' tall. Buying traps of this size would cost thousands, but I built mine for about $400 in parts and some of that cost was finish to make them look professional.

Thanks- took a look at your measurements and it's re. smoothing the frequency response. Which is great but looking more for time domain measurements (decay times). I can just pick up another sub to smooth out frequency response no?

 

[MRC01]

The frequency & time domains are related. Room treatments that improve frequency response by squashing modes or weakening (lifting) nulls, also improve time domain performance.
Intuitively, the long decay time is the room resonating at those frequencies when the distance between reflective surfaces (walls) is related to the wavelengh of certain frequencies. Room treatments weaken those reflections which changes the room's response, weakening those modes.

 

[q3cpma]

Not really on topic, but any impressions about the Eves? Also, any reason for having chosen the SC208 instead of the 207 or 205 when using them with a sub?

 

[Pio2001]

Room treatments weaken those reflections which changes the room's response, weakening those modes.

It depends on what you call room treatment.
Building a new special wall in front of the original regular wall might work.
If you're talking about adding carpets and curtains, forget it. Low frequency don't care about them.

The first thing in order to improve low frequency resonances is to move the speakers where the resonances don't exist.

Electronic EQ might also have a positive effect on long resonances. Below is a test signal at 69 Hz, where a strong resonance exist in my room.
Top picture : the original test signal.
Second picture : how it looks after being amplified by a strong resonance (vertical scale zoomed out in order to compare the decays).

And here are recordings made from my listening position.
On top, the signal without equalization. It sounds horrible. But it doesn't sound horrible because the decay is too long. It sounds horrible because it plays way too loud compared to other frequencies.
Bottom : same thing after the level has been equalized with a minimal phase equalizer (vertical scale zoomed in in order to compare the decays). Reducing the level of this frequency also improved the decay time, although I did nothing about it. I just corrected the frequency response.

 

[MRC01]

It depends on what you call room treatment.
Building a new special wall in front of the original regular wall might work.
If you're talking about adding carpets and curtains, forget it. Low frequency don't care about them.
The first thing in order to improve low frequency resonances is to move the speakers where the resonances don't exist. ...

I've seen 6-9 dB improvements from room treatments like big tube traps, bass traps and tuned membranes.
Rearranging the speakers & listener position can also make a big difference, since SBIR and LBIR can contribute to modes.
I use EQ as a last resort, after fixing as much as I can with room treatments.

 

[mitchco]

One can also reduce low frequency decay times using DSP: https://audiophilestyle.com/ca/reviews/home-audio-fidelity-room-shaper-review-r855/

Room Shaper is not room eq. Room eq can smooth out the low end by a combination of minimum phase response eq and excess phase correction below the rooms transition frequency. "Room Shaper" DSP software is designed specifically to reduce long low frequency decay times. It is measurable and audible (to my ears anyway).

It reminds me of when I worked in really well acoustically designed studio control rooms where the bass is really tight sounding. Note these controls rooms were designed with proper room ratios and geometric shapes designed to a) distribute the room modes as randomly as possible and b) attenuate the long decay times by passing the low frequency wavelengths through the inner shell of the control room into a secondary outer shell room that "trapped" the wave. This prevented low frequency room build up. Cool designs, but really expensive! This is because traditional bass traps and membranes are not too efficient below 100 Hz and you would need a lot of them to control anything meaningful below 50 Hz. Been there done that. One ends up stuffing room and at the detriment of higher frequencies. Not saying they don't help, but there is a point of diminishing returns on investment.

Note it is a VST plugin, so your playback software needs to be able to handle VST plugins like JRiver for example.

Good luck!

 

[MRC01]

... "Room Shaper" DSP software is designed specifically to reduce long low frequency decay times. It is measurable and audible (to my ears anyway). ... traditional bass traps and membranes are not too efficient below 100 Hz and you would need a lot of them to control anything meaningful below 50 Hz. Been there done that. One ends up stuffing room and at the detriment of higher frequencies. Not saying they don't help, but there is a point of diminishing returns on investment.
...

In my case, room treatments were quite effective even at low frequencies. A pair of tube traps 2' diameter x 7' tall in the room corners lifted bass response by 3 dB @ 40 Hz and 8 dB @ 80 Hz, flattened the overall response, and improved the CSD. This did over-damp the room, so I rearranged using half the amount of tube traps (same diameter), with reflective material around half the tube trap to reduce its absorption of high frequencies and the curve improves dispersion, and added more narrow band tuned bass membranes (GIK Scopus T70s) and bass traps (RealTraps MegaTraps). This achieved the same benefits without over-damping the room. Then I also apply EQ, but after these treatments fixing most of the problem, the EQ is quite gentle.

This had 2 key benefits over using EQ alone. FIrst, It improved a null that no amount of EQ could ever fix. Second, it improved the CSD more than EQ alone did. When you EQ to attenuate a peak, you're reducing energy at the room's mode which also improves time decay. But EQ doesn't change the fact that the room still resonates at that frequency. When you use room treatment to attenuate a peak, you are reducing the room's resonance at that frequency.

 

[mitchco]

@MRC01 no disagreement, I use bass traps myself and have spec'd many a room for them and measured the before and after results.

Just to be clear, "Room Shaper" is not "EQ" nor is it a FIR filter or any kind of "static" equalization filter. The Room Shaper DSP works in the time domain only to reduce room modes that have excessively long decay times. It is a DSP trick in the time domain. Room Shaper can be used in conjunction with both passive acoustic treatments and room eq. It is not my product, nor do I derive any financial benefit from it, other than it can really reduce excessive long low frequency decays times.

 

[dshreter]

The first thing in order to improve low frequency resonances is to move the speakers where the resonances don't exist.

For home listening, I would put most of your energy into subwoofer placement as well as how you've configured your crossover. Your monitors have quite extended bass response, and you can high pass them down to the lowest frequency where they still have flat frequency response and avoiding any of their major room modes. If you're lucky, you can then figure out subwoofer placement that doesn't excite any room modes that are below your low pass frequency to the sub.

This can take some experimentation depending on how much flexibility you have with both the monitor and subwoofer placement. But if you look at the in room response from the monitors and subwoofer independently, that can give you some good ideas of the most advantageous crossover point.

I think this article has a pretty good summary of techniques.
https://www.gikacoustics.com/room-setup-speaker-placement-201-part-two-subwoofers/

 

[Pio2001]

Just to be clear, "Room Shaper" is not "EQ" nor is it a FIR filter or any kind of "static" equalization filter. The Room Shaper DSP works in the time domain only to reduce room modes that have excessively long decay times. It is a DSP trick in the time domain.

Time and frequency domains are mathematical abstractions. The decay of a room mode is completely defined by its frequency and phase response (that's the frequency domain), which is the same thing as its impulse response (that's the time domain representation of the same thing).

For home listening, I would put most of your energy into subwoofer placement as well as how you've configured your crossover. Your monitors have quite extended bass response, and you can high pass them down to the lowest frequency where they still have flat frequency response and avoiding any of their major room modes. If you're lucky, you can then figure out subwoofer placement that doesn't excite any room modes that are below your low pass frequency to the sub.

Thank you for the advice, but I don't plan to upgrade my system. I am happy with it, and it would be a very long work to find a result as balanced as the one I have now in such a room. The low frequency room modes are dealt with with a MiniDSP 2x4, with a 17 bands parametric EQ.

 

[mitchco]

@Pio2001 speakers are typically minimum phase devices. However, when speakers are placed in rooms, we get both minimum phase response and non minimum phase response due to room resonances. https://www.roomeqwizard.com/help/help_en-GB/html/minimumphase.html

Most room correction software will take the acoustic measurement and split it into a minimum phase response and excess phase response components. When you correct the magnitude for the minimum phase response, you are correcting both the amplitude and phase response of the loudspeaker. But we are still left over with the excess phase response of the room. The room correction software applies an independent timing response (i.e. excess phase correction) that is independent of the minimum phase response. This is why you will see most room correction software using FIR filters as it allows us to independently apply both a frequency and timing response correction. This is why most PEQ's don't work in correcting room responses, as they are not dealing with the excess phase response of the room.

In this portion of the article, I demonstrate this clearly as both frequency responses are exactly the same, but switch to the time domain view, we can see that they are indeed very different, with one having no excess phase correction and the other with excess phase correction (aside from the time alignment of the drivers). https://audiophilestyle.com/ca/ca-academy/ integrating-subwoofers-with-stereo-mains-using-audiolense-r712/#timedomain Scroll up a bit to see the frequency and excess phase correction and compare.

Not giving away Thierry's Room Shaper algorithm, but it is dealing with the time domain only and not the minimum phase response of the loudspeaker. If you reference the article I wrote on the subject, you can see before and after spectograms that show clearly the frequency response has not changed, but the amplitude of the room resonance has been attenuated significantly, like in the order of 20 dB below 100 Hz and between 250ms and past 500ms of decay. No amount of subwoofering or placements or acoustic treatments is going to get that kind of surgical reduction in room resonances. This is what makes Room Shaper unique.

 

[dshreter]

Thank you for the advice, but I don't plan to upgrade my system. I am happy with it, and it would be a very long work to find a result as balanced as the one I have now in such a room. The low frequency room modes are dealt with with a MiniDSP 2x4, with a 17 bands parametric EQ.

I wasn't suggesting you upgrade your system, I was agreeing with your comment that moving speakers is the most important thing for the OP to do

 

[tjcinnamon]

@Pio2001 speakers are typically minimum phase devices. However, when speakers are placed in rooms, we get both minimum phase response and non minimum phase response due to room resonances. https://www.roomeqwizard.com/help/help_en-GB/html/minimumphase.html

Most room correction software will take the acoustic measurement and split it into a minimum phase response and excess phase response components. When you correct the magnitude for the minimum phase response, you are correcting both the amplitude and phase response of the loudspeaker. But we are still left over with the excess phase response of the room. The room correction software applies an independent timing response (i.e. excess phase correction) that is independent of the minimum phase response. This is why you will see most room correction software using FIR filters as it allows us to independently apply both a frequency and timing response correction. This is why most PEQ's don't work in correcting room responses, as they are not dealing with the excess phase response of the room.

In this portion of the article, I demonstrate this clearly as both frequency responses are exactly the same, but switch to the time domain view, we can see that they are indeed very different, with one having no excess phase correction and the other with excess phase correction (aside from the time alignment of the drivers). https://audiophilestyle.com/ca/ca-academy/ integrating-subwoofers-with-stereo-mains-using-audiolense-r712/#timedomain Scroll up a bit to see the frequency and excess phase correction and compare.

Not giving away Thierry's Room Shaper algorithm, but it is dealing with the time domain only and not the minimum phase response of the loudspeaker. If you reference the article I wrote on the subject, you can see before and after spectograms that show clearly the frequency response has not changed, but the amplitude of the room resonance has been attenuated significantly, like in the order of 20 dB below 100 Hz and between 250ms and past 500ms of decay. No amount of subwoofering or placements or acoustic treatments is going to get that kind of surgical reduction in room resonances. This is what makes Room Shaper unique.

Have you tried their SoundStage Shaper plugin?