Electronic bandwidth limiting for bassreflex speakers?

[Heinrich]

Bass reflex loudspeakers have the characteristic that the diaphragm movement increases drastically below the tuning frequency. This not only produces harmonic distortion, but also intermodulation. In extreme cases, a loudspeaker can destroy itself. Also the ineffective pumping of air back and forth through the port also produces clearly perceptible wind noise.

In passing, I have seen that modern content quite often has components of relevance below 40Hz. These are mostly exceptions, such as thumping noises, but now also regularly recurring 'musical' parts.

Speakers are getting smaller these days, are mostly vented but then show tuning frequencies well above 40Hz. The consequential distortions create a certain “drama” that replaces the boom from the lowest frequencies to a certain extent, but it would be cleaner to switch off the distortion machine.

What are the simplest means of limiting the signal bandwidth downwards?

Tuning at 50Hz, output @-10dB in that register, though, still 5%HD (86db median):

PMC result6 Monitor Review

This is a review, listening tests, EQ and detailed measurements of the PMC result6, active professional monitor (speaker). It is kindly sent to me by a member and costs US $3000 for a pair. I have to confess: this is one ugly looking monitor! Let's hope all the accents are functional. The...

www.audiosciencereview.com

Tuning at 70Hz (!), on the verge of unusable at least for EDM:

Cambridge Audio Minx XL Speaker Review

This is a review, listening tests, EQ and detailed measurements of the Cambridge Audio Minx XL bookshelf speaker. It is on kind loan from a member and costs US$299. I should note however that I can't find new stock anywhere so maybe it is discontinued. As you can see, speaker is rather...

www.audiosciencereview.com

Tuning at 60Hz, output @-15dB (!) in that register, 100%HD (!) at 40Hz (86db median):

Micca MB42X G2 Budget Speaker Review

This is a review, listening tests and detailed measurements of the Micca MB42X G2 low cost 2-way bookshelf speaker. It is on kind loan from a member and costs US $100 for a pair, shipping included. I expected thin, cheap plywood construction but instead, found a robust composite enclosure that...

www.audiosciencereview.com

 

[ppataki]

What are the simplest means of limiting the signal bandwidth downwards?

Simply use a high-pass filter, preferably a steep one (24dB/oct. for example). It will protect the driver and will tame distortion too

 

[Heinrich]

Simply use a high-pass filter, preferably a steep one (24dB/oct. for example). It will protect the driver and will tame distortion too

Sure, but how?

 

[ppataki]

Sure, but how?

I use Jriver's DSP Studio + various VST plugins to apply filters like this

You can also use hardware, like miniDSP and various other DSP solutions

You can also apply it with a passive circuit but I will let others comment on that, I am not familiar with that domain

 

[Heinrich]

I use ...

Thanks! Can you share any personal impressions of how much it helps or not?

 

[ppataki]

Thanks! Can you share any personal impressions of how much it helps or not?

I use a linear phase 24dB/oct. high-pass filter on my main front speakers (in case of all these projects) at 80Hz where I to cross them to my subs (currently I use Fabfilter Pro-Q 4)
And I also use the same filter but at 40Hz with my rear speakers to simply take the burden off of them
It definitely helps and it makes a big difference, especially if you turn up the volume - cone excursion will be limited by the filter so you get more control and less distortion

 

[Oddball]

Ideally you wound also want to understand in-room response of the speakers by measuring them (REW+Umic 1). That would give you a more precise idea where to apply the filter and what roll-off to use. Since you will be cutting the signal (and not crossing over to subs) probably best to cut only what is necessary.

 

[ZolaIII]

Sure, but how?

Right under the port tuning and higher order if need be.

 

[JS2]

Bass reflex loudspeakers have the characteristic that the diaphragm movement increases drastically below the tuning frequency. This not only produces harmonic distortion, but also intermodulation. In extreme cases, a loudspeaker can destroy itself. Also the ineffective pumping of air back and forth through the port also produces clearly perceptible wind noise.

In passing, I have seen that modern content quite often has components of relevance below 40Hz. These are mostly exceptions, such as thumping noises, but now also regularly recurring 'musical' parts.

Speakers are getting smaller these days, are mostly vented but then show tuning frequencies well above 40Hz. The consequential distortions create a certain “drama” that replaces the boom from the lowest frequencies to a certain extent, but it would be cleaner to switch off the distortion machine.

What are the simplest means of limiting the signal bandwidth downwards?

Tuning at 50Hz, output @-10dB in that register, though, still 5%HD (86db median):

PMC result6 Monitor Review

This is a review, listening tests, EQ and detailed measurements of the PMC result6, active professional monitor (speaker). It is kindly sent to me by a member and costs US $3000 for a pair. I have to confess: this is one ugly looking monitor! Let's hope all the accents are functional. The...

www.audiosciencereview.com

Tuning at 70Hz (!), on the verge of unusable at least for EDM:

Cambridge Audio Minx XL Speaker Review

This is a review, listening tests, EQ and detailed measurements of the Cambridge Audio Minx XL bookshelf speaker. It is on kind loan from a member and costs US$299. I should note however that I can't find new stock anywhere so maybe it is discontinued. As you can see, speaker is rather...

www.audiosciencereview.com

Tuning at 60Hz, output @-15dB (!) in that register, 100%HD (!) at 40Hz (86db median):

Micca MB42X G2 Budget Speaker Review

This is a review, listening tests and detailed measurements of the Micca MB42X G2 low cost 2-way bookshelf speaker. It is on kind loan from a member and costs US $100 for a pair, shipping included. I expected thin, cheap plywood construction but instead, found a robust composite enclosure that...

www.audiosciencereview.com

I had big passive speakers n the 80s with a horn, woofer, port and 15 inch bass reflex. Eventually the woofer dried out which may actually be a result of the bass reflex pushing air back and forth.

 

[voodooless]

Simply use a high-pass filter, preferably a steep one (24dB/oct. for example). It will protect the driver and will tame distortion too

You really don't need such a steep filter. 12 dB is also fine. If you model the excursion, you'll see that a 12 dB filter is usually enough to keep the excursion low enough. But it will need some tweaking of the high-pass frequency to get it optimal. WinISD (or comparable) is your friend

Adding more slope would also increase group delay. Ideally, you would like to have this as low as possible. The best way to do that is to tune as low as the driver/enclosure combo allows, even if that does not yield a linear response. You can always EQ that flat. But you will push down the group delay bump in frequency, meaning it's less of an issue with music content.

 

[YSDR]

You really don't need such a steep filter. 12 dB is also fine. If you model the excursion, you'll see that a 12 dB filter is usually enough to keep the excursion low enough.

You don't even need a high-pass filter, because a good parametered low-shelf filter is typically enough and this introduces lesser phase shift/group delay. Of course a high-pass reduces the excursion more at very low frequencies.

 

[gberchin]

Section 4.1 here. PM me for a copy, if interested.

 

[markstein]

I use a linear phase 24dB/oct. high-pass filter on my main front speakers (in case of all these projects) at 80Hz where I to cross them to my subs (currently I use Fabfilter Pro-Q 4)
And I also use the same filter but at 40Hz with my rear speakers to simply take the burden off of them
It definitely helps and it makes a big difference, especially if you turn up the volume - cone excursion will be limited by the filter so you get more control and less distortion

Don't you have issues with pre-ringung when using a FIR linear phase filter for this purpose? Wouldn't a IIR minimum phase filters the correct choice?

 

[voodooless]

Don't you have issues with pre-ringung when using a FIR linear phase filter for this purpose? Wouldn't a IIR minimum phase filters the correct choice?

If perfectly designed, the low- and high-pass ringing would cancel out.

 

[Heinrich]

... will push down the group delay bump in frequency, meaning it's less of an issue with music content.

There are other criteria, that according to urban legend are more reliable predictors for a pleasent audio experience. Group delay is in itself a quite complex matter--the definition depends explicitly on calculus, no single tone can have a group delay and such. In short, it is often pretty well misunderstood.

What 'we' might know is that contemporary recordings carry much of low frequengy content, down below to 30Hz with not so few recordings. From my experience it is needed as a relevant contribution, if present. Hence, the tuning of any bassreflex should not be lower than that, or a bit above, say 33Hz.

Two points:
1) with small speaker boxes the available sound output at such low frequencies is too low as to be heard, see Fletcher/Munson curves aka equal loudness contour. Hence it makes no sense to long for that w/ the tuning: tune higer, but cut-off.
2) if the tuning is well at 33Hz or so, an elecronic cut-off isn't needed no more, but the speaker might, after room integration, need some boosting for linearization.

On point (2) one might ask, what is more or less the distractor, 'group delay', whatever that means, or severe distortion (HD, IM, Doppler) reaching up to the mids?

On WinISD, that's a very good and easy to learn tool. But the alignment suggestions from the books are basically useless today, once we have digital EQ at hand. To start with a new design needs to know how to weigh parameters against the others, a quite different approach than it used to be in the past. E/g, most reflex housings are severely limited by the speed of air in the port, available power might dictate a minimum enclosure volume, and what not. Now, as bandwidth isn't the target anymore, as we simply EQ that, and Xmax is good enough to nearly always hit the envelope due to Doppler distortion a new trick is to be performed.

 

[ppataki]

Don't you have issues with pre-ringung when using a FIR linear phase filter for this purpose? Wouldn't a IIR minimum phase filters the correct choice?

No, not even when using a standalone linear phase HPF using 24dB/octave slope
It is the question of how the filter was implemented - the best implementation from a pre-ringing perspective (at least that I am aware of - and believe me I have tried many) is thEQorange

 

[voodooless]

There are other criteria, that according to urban legend are more reliable predictors for a pleasent audio experience.

In-room response is probably the single most important metric. There is no point in designing a super nice, smooth, and flat extended response if the room wreaks havoc with it.

 

[gnarly]

Don't you have issues with pre-ringung when using a FIR linear phase filter for this purpose? Wouldn't a IIR minimum phase filters the correct choice?

Yes, a FIR linear-phase sub/system high-pass does have pre-ring potential. There is no way to avoid it.

Crossovers can have little if any, pre-ring potential, as they have two sides to cancel each other's pre-ring potential.
But a system high-pass has nothing below it to cancel with.

Whether or not the pre-ring potential is audible, is debatable. It's not debatable that it can be measured via impulse/step.
For me, it's not an issue.

Switching gears.....Personally, any and all reflex designs need an electrical high-pass filter. Letting a driver unload, and flop around aimlessly below its port tuning is bad juju at best. Subs need a high-pass, however implemented. Reflex mains need one too, or a crossover which effectively provides high pass.
Only sealed boxes are OK without a high-pass.,...... well I should say maybe OK. Perhaps not, with all the ridiculous boost we often see put on them.

 

[Heinrich]

In-room response is probably the single most important metric. There is no point in designing a super nice, smooth, and flat extended response if the room wreaks havoc with it.

Exactly--in-room the group delay depends solely (by an order of magnitude more severely) on room interaction. The other parameters I was referring to are harmonic distortion, intermodulation and Doppler induced amplitude modulation, that reach up to the mids w/ the (in?)famous two-way.

Only as to motivate the concern. The original question was, if bandwidth limiting is worth it. According to my notion (1) yes for small speakers, that won't deliver deep bass that can be heard at the output given as a max (depends on cone area mostly). According to (2), no if the bass is big enough, which should be true for all real subs.

I missed to mention the observation, that most EDM today is strictly limited to 40Hz, a so called brick-wall filter applied. But more artsy recordings may sport the 30Hz as 'bumps' from plucking a bass guitar and other spurious sounds. You may not want to miss out ;-)

ps: see also this review, giving an example for very deep bass https://www.audiosciencereview.com/...lac-df63-floor-standing-speaker-review.65613/ Don't quote me when saying it was worth the effort, but there are other recordings that integrate such deep effects effortlessly and in a real musical context ...

 

[Putter]

Excuse my ignorance, but doesn't your average home theater receiver with subwoofer out(s) do the same thing. A little research shows Denon and I believe other brands use a 24 db/octave high pass filter which typically can be set at different frequencies between 60 hz and 200 hz although some are fixed at 80 hz.