Extending bass response with the help of dsp

[Tangband]

One cant fool the acoustic laws in loudspeakers, there are no free lunches even when using dsp to extend the bass response in a bass reflex loudspeaker .
This is what we have learned.

But this is an interesting article :

https://audioxpress.com/files/attachment/2721

I had to test this .

In a bass tube speaker, the drivers cone stand almost still at the fb frequency - the tube makes all the sound . In my chn110 this tuning frequency is 36 Hz . At the same time, the backpressure of the tube is protecting the driver cone from doing extensive movements around that fb frequency.

Using test tones - I can see that the cone of the driver stand still at 36 Hz and the tube makes all the sound .

By using a high pass shelving filter , combined with PEQ I can extend the bass response from my Markaudio chn110 speakers by boosting slightly lower than fb , then using a high pass shelving filter -12 dB at 20 Hz . And this is without any loss of spl or xmax .

The high pass shelving filter at 20 Hz is protecting the cone below 30 Hz where it dont have any backpressure protection anymore from the tube.

In practice, this can be done with all loudspeakers.

This is easily done with a WiiM pro .

Comments on this ?

 

[Sokel]

Won't the amp suffer from this 5db boost down there?
That would need lots of power,on top of the already demanding one there.

 

[Tangband]

Won't the amp suffer from this 5db boost down there?
That would need lots of power,on top of the already demanding one there.

Yes but only at around 34 Hz , below 30 Hz the amplifier is relieved because of the high pass shelving filtering and need much less power. Below 20 Hz the response for the amplifier is down -12 dB .

 

[ppataki]

Generally speaking you can DSP closed cabinet speakers much better than ported ones
That is due to the fact that the low frequency roll-off of sealed cabinets is more gentle (around 12dB/octave) compared to ported cabinets
So in other words, there is 'material' down there that you can work with
That's the reason why I design and build sealed cabinets only
Having said that, obviously you will sacrifice SPL in exchange for low extension in any case - not to mention that (much) more amp power will be needed too
One shall take all this into account already at the design phase to avoid any unpleasant surprises

 

[Tangband]

Generally speaking you can DSP closed cabinet speakers much better than ported ones
That is due to the fact that the low frequency roll-off of sealed cabinets is more gentle (around 12dB/octave) compared to ported cabinets
So in other words, there is 'material' down there that you can work with
That's the reason why I design and build sealed cabinets only
Having said that, obviously you will sacrifice SPL in exchange for low extension in any case - not to mention that (much) more amp power will be needed too
One shall take all this into account already at the design phase to avoid any unpleasant surprises

Yes, but bass reflex has the benefit of the backpressure from the tube at fb making the speaker cone stand still ( the tube makes all the sound ) with the output from a correctly done bass reflex speaker, and because of this can play much louder at fb , about 10 dB as long as the tube dont overload.

 

[Keith_W]

Interesting technique! Thanks for posting. Do you have before-after measurements?

I echo @Sokel's concern about overtaxing your amplifier.

In practice, this can be done with all loudspeakers.

*all loudspeakers with ports

 

[Sokel]

Interesting technique! Thanks for posting. Do you have before-after measurements?

I echo @Sokel's concern about overtaxing your amplifier.



*all loudspeakers with ports

The problem is not only clipping but that in some occasions an amp running out of juice can produce the famous one note (50 or 60Hz,depending the mains) bass.
People easily blame that to the speakers/room/whatever but is something to take into account too.

 

[Tangband]

With the WiiM pro , one can also play with highpass shelf filtering and use a higher Q=4 slope.
It will give a peak at the chosen frequency and a sharp roll of below it.

This is explained in the article.

Then I can get about the same results as using two filters ( high shelving with Q=1 and a peq )

 

[Tangband]

Interesting technique! Thanks for posting. Do you have before-after measurements?

I echo @Sokel's concern about overtaxing your amplifier.



*all loudspeakers with ports

In practise, one can extend the response about 0,85 lower than the fb frequency.
With my chn110 speaker I gain about 6 Hz ( flat to 30 Hz ) without any more cone movements .

People owning a Kef Blade 2 Meta can make them go much lower in the base

To use this technique, one can use frequency test tones to see at which frequency the cone is standing still = the tuning frequency of the speaker.

Then set a PEQ 2- 5 Hz below that frequency ( + 5 dB Q=4 ) and a high pass shelving filter ( with Q= 1 ) 10 Hz below that frequency , with gain -12 dB.

 

[Purité Audio]

Buy subwoofer.
Keith

 

[Tangband]

Buy subwoofer.
Keith

Though I agree that two good stereo subwoofers might be better , you also have to deal with the eventual integration problem.

 

[sigbergaudio]

I would be a little careful EQing below the tuning frequency. The high pass filter is a good security, but if you play loud / towards the max capacity of the speaker, very small changes can be the difference between safe and exceeding xmax.

Moving a high pass filter from 18 to 20hz will as an example significantly alter the xmax at a given power level.

 

[sigbergaudio]

Here is the excursion of the same ported speaker with 18hz high pass vs 20hz high pass with the same power level in. The red line is xmax.

(y= excursion in mm, x = frequency)

 

[alex-z]

Though I agree that two good stereo subwoofers might be better , you also have to deal with the eventual integration problem.

Mono subs are usually better than stereo, managing room modes outweighs the small benefit to stereo imaging. Assuming you use a reasonable crossover, anything above 80Hz is a grey area.

Subwoofer integration is rarely difficult, the long wavelengths give you a lot of placement flexibility. The only unfortunate thing is that WiiM doesn't seem to offer an option to add subwoofer delay, only a 0 vs 180 phase option.

 

[ZolaIII]

@alex-z every sub including hundred of them together are mono down there (under 60 Hz) and no every 2x will give you additional +3 dB.
@Tangband just use room, in order for first peak to be approximately there room needs to be about 5 m long.
Edit: using sub's one per each chenel gives you possibility to set crossover higher (110~130 Hz) and really stress out mains also it enables better scaling adjustment to psyho-acoustic perceived loudness curves (equal loudness compensation).

 

[Tangband]

With this technique, its possible to extend the bass from a Kef ls50 Meta with 51 Hz ( tuning frequency ) *0.85 = 43,3 Hz . This is, in practice, a big difference.

KEF LS50 Meta loudspeaker Measurements

Sidebar 3: Measurements I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the KEF LS50's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield and spatially averaged room responses. My estimate of the LS50 Meta's voltage sensitivity was...

www.stereophile.com

 

[sigbergaudio]

With this technique, its possible to extend the bass from a Kef ls50 Meta with 51 Hz ( tuning frequency ) *0.85 = 43,3 Hz . This is, in practice, a big difference.

KEF LS50 Meta loudspeaker Measurements

Sidebar 3: Measurements I used DRA Labs' MLSSA system and a calibrated DPA 4006 microphone to measure the KEF LS50's frequency response in the farfield, and an Earthworks QTC-40 for the nearfield and spatially averaged room responses. My estimate of the LS50 Meta's voltage sensitivity was...

www.stereophile.com

You will also increase distortion significantly in a frequency range where the distortion is already high. This is fine if you play at very moderate levels, not a great idea if you want maximum capacity / dynamics.

 

[Tangband]

You will also increase distortion significantly in a frequency range where the distortion is already high. This is fine if you play at very moderate levels, not a great idea if you want maximum capacity / dynamics.

This is correct and all of you out there , be careful if trying .

However, at the same time, the high pass shelving filtering at maybe 40 Hz will prevent the speaker cone from moving at lower frequencies . Without dsp , this speaker has no protection at all below 40 Hz where it looses the backpressure advantage from the bass tube . This is as you know the dilemma with all bass tube speakers.

It would be interesting to see measurements if there is any serious concerns about max spl with music using dsp to gain a couple of Hz .

Using sheets of A4 paper , rolled as a tube and put inside the bass tube making it smaller can also lower the tuning frequency - at every ones own risk ofcourse

 

[ZolaIII]

@Tangband I really, really wouldn't do that.

https://dl.dropboxusercontent.com/s/k51inulgdej7zrs/Kef%20LS50%20Wireless%20II_Compression_full.png?dl=0

After looking at impedance/phase and frequency response down to woofer they are already pushed too much out of the box. They should be high passed sharply at 50 Hz and crossover point is high at 130 Hz if you want to stress them out. Unfortunately they did far worse with Metas in this case to non Metas when it comes to tweater crossing and impedance peek but you can't do anything regarding that other than use not load dependent amplifier with them.

 

[Tangband]

@Tangband I really, really wouldn't do that.

https://dl.dropboxusercontent.com/s/k51inulgdej7zrs/Kef%20LS50%20Wireless%20II_Compression_full.png?dl=0

After looking at impedance/phase and frequency response down to woofer they are already pushed too much out of the box. They should be high passed sharply at 50 Hz and crossover point is high at 130 Hz if you want to stress them out. Unfortunately they did far worse with Metas in this case to non Metas when it comes to tweater crossing and impedance peek but you can't do anything regarding that other than use not load dependent amplifier with them.

Ok you got a point. Maybe this Elac speaker is a better candidate for this :

Elac Debut Reference DBR-62 Speaker Review

This is a review and detailed measurements of the Elac Debut Reference DBR-62 bookshelf speaker. It was kindly purchased new and drop shipped to me for testing. The DBR-62 costs US $600 a pair from multiple sources/dealers. The DBR-62 comes in black and walnut finish or the distressed Oak and...

www.audiosciencereview.com