What exactly happens if you seal a speaker which was specifically designed to be ported? KEF Reference 3 Meta

[Goldhamster916]

Hi,

following scenario: the KEF Reference 3 Metas have two ports with a tuning frequency of ~40 Hz and two 6.5'' high excursion woofers each. I would listen at 1.5m at low volume (~60DB), speakers placed within a few inches to front wall. As you can see I really don't need the SPL the Reference 3 provides. The idea is to seal the ports to have the air inside behave like a spring and decrease woofer distortion which, in the ported configuration, increases a lot under the ports tuning frequency. Apart from that i would maybe try to apply a few DB bass boost to get, combined with the more gradual role off from the sealed design, a flat FR till ~30hz. I would then try to find a way to limit SPL below ~100HZ. Amp has 250W (hypex based), power handling of the speaker is 300W.

1. Are there any flaws in my reasoning?

2. would i get the desired results?

3. would there be group delay or other non FR related issues with this approach? if yes would it be audible? in case of group delay: would linear phase filters be a solution?

4. From what i have read about KEF i would assume the Reference 3 and all components within are modeled specifically with this exact port in mind. would the effects of a sealed cabinet negatively interact with/contradict design choices on other parts of the loudspeaker and therefore lead to negative effects on sound quality? Dampening and motor structure for example are parts i can think of which might be moduled to this specific design.

KEF Reference 3 Meta CEA2034:

I think I basically try to emulate what many active speakers do with a passive speaker, like here on the LS60 but at even lower SPL. I would be happy to have a headroom to ~70DB at 30hz. the question is if there is anything in the Reference 3 design which makes it unsuitable fo this approach.

LS60 low frequency compression:

I appreciate anyone who takes his time to answer my questions!

 

[pablolie]

Don't they provide different options to control the ports with foam stuff or tubes with different depths? [1] You can stuff a sock in it, just means you have to optimize bass output in some other way (if needed). Ideally with a room correction tool, of course. I assume if you feel compelled to do that, it means you're not happy with the response in your room.
That said, a truly flat response to 30Hz seems a questionable goal given the speakers. Just checked and they are specced at - 6dB for "typical room response" at 28Hz. So any natural linear response at 30Hz is questionable .

[1] I just checked and the Reference 3 come with 2 different length port tubes to optimize bass response for your room, just like the Reference 1 (which I have) do. Have you tried to see if that makes the response more to your liking?

 

[DVDdoug]

I suspect you'll notice a loss of bass. I also assume it's a well designed and optimized speaker and there's probably nothing "simple" you can do to improve it.

And yes, you should be able to stuff the ports with something. It won't be "sealed" but it will behave more like a sealed than ported speaker. Stuffed or completely sealed will probably sound the same.

You can make a good speaker sealed or vented and whatever is best depends on the driver characteristics and cabinet size, etc., and I assume KEF made all of the right trade-offs. (It's easier to badly foul-up a ported design.)

I saved a link to this post which shows the generalized difference between ported and sealed designs. At some point the curves cross and the sealed box is putting out more bass, so with EQ, enough power, and big enough woofer (that can also handle the power) you can get more bass than from the ported box.

Often ported designs are tuned with a slight bump that's hopefully not big enough to be a problem, and maybe even be desirable, and that can extend the cutoff frequency a bit more.

At 60dB you probably aren't hearing any deep bass anyway... Our hearing sensitivity is poor at low frequencies.

 

[Duke]

In general woofers that perform well in a ported enclosure have parameters (Qes in particular) that do not result in very good low-end extension in a sealed-box configuration. The -3 dB frequency will move up, possibly by an octave or more. The roll-off will be quite gentle. You might need 6 dB or so of EQ in order to reach the same -3 dB frequency as before, resulting in more heating of the voice coil.

I don't know enough about your speakers to make specific comments about what you have in mind. My guess is there are power levels (after EQ) at which it's perfectly safe, and power levels at which it may be problematic.

You might try sealing up ONE of the two ports. This will lower the tuning frequency a little and result in a small reduction of bass energy and a little bit more gentle roll-off, but not nearly as much change as sealing both ports. The potential downside is, you might get turbulence ("chuffing") at high SPLs since you would have halved your port cross-sectional area.

 

[jEDGEc0m]

I experimented with sealing the port of a subwoofer not too long ago.
The excursion of the woofer went up substantially. It also actually performed audibly worse, with sometimes a flappy noise - i.e. worse than any artefact the open port did.

ChatGPT suggested this, which correlates

• One of the big benefits of a ported design: lower excursion near tuning, reducing distortion and power demands.
• If you seal the port, you remove this mechanism. The woofer now moves more (excursion increases) at those same low frequencies, especially below the original port tuning

At the low volumes you're mentioning, the excursion might not be an issue, even with added eq to boost bass. But I imagine it probably overall is going to be a less favourable outcome with the ports sealed.

(Your distance to the wall shouldn't be a problem for the open ports either. and being close to the wall has benefits)

Certainly experiment with eq to get the bass boost/extension that pleases your ear - but I suspect leaving the ports open might end up better.

And certainly if you don't already, would recommend some kind of mic to measure the low freq in-room response at your listening position, and eq down main errant peaks, including some that might occur from speakers being close to the wall.

For low frequencies, I understand one doesn't need/get much benefit from a specific measurement mic - so an app on your phone might be good enough to be useful, if you don't have other means of measuring.

(ps: I always like seeing measurements of before and after)

I recently got a pair of AsciLab F6B.
Here are some initial impressions and measurements. Speakers are close to the wall also. You can see how i used 2 eq filters to drop down some peaks, which was an improvement. I have since also tuned with eq to raise up the bass down ~45 hz, similar to what you're trying to achieve. It works well and have had no issues with the playback volumes I need.

That said, I have also tuned a setup with subwoofers, which is still quite favourable with more bass warmth and punch, whether at lower volumes or higher. This might be in part the extra extension, but also having the subs in different locations helps fill up the dip from the room at ~60hz.

AsciLab speakers are about to launch

F6B. Received today. Once dispatched, delivery was very quick and sooner than estimate. Departed AsciLab late Wednesday, arrived Friday morning, Korea to Australia. Early impressions, very positive. Used without subs so far, 3.5x3.5x2.4m room (some treatment of makeshift insulation batts)...

www.audiosciencereview.com

TL;DR. Get eq, measure, applyi eq, measure, and enjoy!

 

[bmc0]

In general woofers that perform well in a ported enclosure have parameters (Qes in particular) that do not result in very good low-end extension in a sealed-box configuration.

Yes, the total Q will be a bit low, but this can be corrected with equalization (Linkwitz transform or similar). As long as you have enough power, the usable low-frequency extension depends largely on the linear volume displacement and the desired SPL.

 

[sigbergaudio]

Well, the whole point of the port is to get more low end extension with less distortion and excursion, so if your goal is to extend the bottom end / get more bass, this is likely not a great idea. The roll off will be slower in a sealed configuration, but will start much sooner.

You will reduce distortion / increase headroom by plugging the port, as the speaker will play less of the deepest bass that requires the most excursion from the driver. If you then EQ the low bass up again, you will remove the benefit that you created by plugging the port, and force the driver to work even harder than it did before plugging the port. If the port is tuned to 40hz, the driver is unlikely to be designed to do a great job at 30hz.

A better solution is probably to invest in a subwoofer.

All that being said, you are of course free to experiment (that is part of the fun), and as long as you don't play too loud / watch the excursion of the driver after EQing, you should be fine. Just remember that each 3dB of boost requires twice the power, so it's much more than it sounds, and you will likely need way more than 3dB boost. So you will relatively quickly reach a level the driver is not designed to handle.

 

[ppataki]

would there be group delay or other non FR related issues with this approach? if yes would it be audible? in case of group delay: would linear phase filters be a solution?

In a sealed cabinet, regarding the low frequencies, group delay will be way lower and there will be much less phase rotation

But then you apply a low shelf filter or a Linkwitz Transform or equivalent to compensate for the low end and the group delay will increase....
However, as you rightly pointed out, if you apply a linear phase low shelf filter then the group delay will not increase

 

[RickS]

Here is an example for something not as well designed as the KEF…

What really happens when you plug a speaker port/vent?

As an extension of my comparison test of speaker damping materials, thought I would see how plugging the vent affects its output. In this case, for the Dayton C-Note speaker. As you will see, the result depends a lot on what material you use... So, the red trace is the baseline speaker as...

www.audiosciencereview.com

Actual results depend on the specific design. As I mention elsewhere, stuffing a port is often a sub-optimization. If you have eq or dsp, seems a lot more straightforward to just reduce the bass (HPF?). Without driver specs, a sim is a lot of effort. Did you try contacting KEF? We have have members that work for KEF but less likely to get their attention unless your thread title mentions KEF.

 

[DanielT]

So the pros and cons of, well constructed, sealed vs. ported constructions is the question, you could say.

The problem seems to be that there are so many poorly constructed ported speakers. It seems that every other one that Amir tests shows clear, ( probably audible?), port resonances.

A curved slit port to avoid resonances seems to work well though. For example, as on these bass boxes. Measured after construction and no problems with resonances:

faktiskt.io • Visa tråd - Slitsport i basstöd

Edit:
It's probably best if the slot is curved at both ends. 7 degrees of curve should be a good goal if I remember correctly.

From the thread linked above:
The best slot gate has this shape:


Symmetrical expansion and a properly baffled mouth (also on the inside) is optimal.

 

[obsessive9119]

Don't they provide different options to control the ports with foam stuff or tubes with different depths?

My LS50s came with two different sized foam inserts to partially or fully block the ports, and guidance on when to use them (based on distance to the wall). I’d have to check but I believe I have mine fully blocked, but have a sub covering the lower frequencies.

 

[Hear Here]

I experimented with sealing the port of a subwoofer not too long ago.
The excursion of the woofer went up substantially. It also actually performed audibly worse, with sometimes a flappy noise - i.e. worse than any artefact the open port did.

Surely precisely the opposite. If you plug the port, the air inside the enclosure will present a greater resistance to cone movement than allowing the air to escape. A ported enclosure will offer less resistance to cone movement; an open baffle will offer no resistance to movement and this the reason for this design being generally considered as unsatisfactory.

The cabinet dimensions (internal volume) will largely dictate whether a ported or fully-enclosed cabinet will suit the driver best.

 

[JohnnyNG]

I experimented with sealing the port of a subwoofer not too long ago.
The excursion of the woofer went up substantially. It also actually performed audibly worse, with sometimes a flappy noise - i.e. worse than any artefact the open port did.

Indeed. People assume there's "more pressure" in a sealed cab but near tuning in a reflex cab it's a very different story.

How a Hole-in-the-Box Works - HomeTheaterHifi.com

Even though loudspeakers utilizing bass reflex techniques (i.e., ported or passive radiator designs) are quite common, or maybe even the norm if you...

hometheaterhifi.com

 

[jEDGEc0m]

Surely precisely the opposite. If you plug the port, the air inside the enclosure will present a greater resistance to cone movement than allowing the air to escape. A ported enclosure will offer less resistance to cone movement; an open baffle will offer no resistance to movement and this the reason for this design being generally considered as unsatisfactory.

The cabinet dimensions (internal volume) will largely dictate whether a ported or fully-enclosed cabinet will suit the driver best.

I read the article @JohnnyNG linked (Thank you), and if I understood correctly, it does seem like a well implemented, ported speaker is likely to be sub-optimal if you seal the port and still want the low frequency range/extension the port was intended for.

And if you ask those frequencies of that system once sealed, the driver excursion will increase (which I guess explains what I witnessed when I did it - the difference in excursion was not subtle - very visible, and I could quickly put the plug in or out and immediately see the difference).

So perhaps plugging a well-designed ported enclosure is only sensible if you're going to high-pass above the port's tuned frequency range. (i.e. with ported speakers high-passed , and crossing over to a subwoofer which will then take the burden of reproducing those delightful frequencies)

Once plugged, and not causing extra excursion, there may be some benefit of reduced port resonances - which depending on the design, may or may not have been relevant/audible anyway.

OP's Kef speakers seem surely like an excellent speaker, and unless crossing to a sub, seems one would be better off letting the ports do their thing as designed.

 

[Goldhamster916]

I appreciate all the replies!

KEF provides port plugs for the step down R series. For the reference series they provide two different port tubes with different tuning frequency but no plugs. The ,,long'' port tube has a 40hz tuning frequency the ,,short'' port tube is tuned higher. A big part of the 2014 reference series was a extensive rework on the ports and LF driver. You can look it up here in their white paper which really is a scientific paper and not some superficial marketing fluff:

chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/https://assets.kef.com/pdf_doc/REF/REF-White-Paper-201219-LR.pdf

Port noises or chuffing of any kind are a non issue with these speakers as outlined in the white paper and also subjectively.

For Context i currently have the older 8 woofer Reference 5 which are, after moving into a much smaller listening space, complete overkill for my room aesthetically. Thats why i think of getting the newer Reference 3 metas but i fear missing out the below 40hz range the reference 5 provides. Interestingly the bookshelf reference 1 have more output below 40hz (although i assume high distortion), maybe that would be a better option. I contacted KEF support but they say my questions exceed their possibilites. If i already owned the Reference 3 metas i would just test it but so i can just work out the theory:


the reference 3 metas have a sensitivity of 86DB at 1 Watt. At my listening volume of 60DB i use a small fraction of a Watt. Even with a bass boost to 70db to make the bass more audible at such a low volume the power requirement could be several hundreds of times higher and still shouldnt cause any issues with a 250W amplifier and 300W power handling of the speaker.

the idea is then to seal the box and basically apply a linkwitz transform where i trade in sensitivity to get more bass extension. The reference 3 is a 120x22x47cm floorstander so the volume (third parameter in the linkwitz formula) should be there. the questions is of course if the LF drivers are mechanically capable of doing this. I know there are only two 6.5'' woofers (on each speaker) but they are very well built and although i cant find a specific xmax number KEF reworked them for the 2014 reference line and i believe them when they claim in their white paper that they achieved very large excursion capability.

the air is basically working as a ,,air cushion'' or spring in a sealed box and thus preventing excess driver movement which would usually occur under the ports tuning frequency in the ported design. With all these factors combined i thought it might be possible to get some usable output below the 40hz tuning frequency of the ported design at low volume. The question is if there are other design aspects of the speaker which are specifically modeled to work within a ported design and would affect sound quality negatively in a sealed design. I don't know if the speaker might then be overdamped for example.

would of course be awesome to get a answer of @jackocleebrown .


thanks again for all the replies, i read them alll carefully but can't answer all of them unfortunately!

 

[bmc0]

If you seal the port, you remove this mechanism. The woofer now moves more (excursion increases) at those same low frequencies, especially below the original port tuning

I missed this before. This is another example of an LLM giving the wrong answer. Sealing a ported enclosure will in fact reduce excursion at frequencies significantly below the port tuning.

Surely precisely the opposite. If you plug the port, the air inside the enclosure will present a greater resistance to cone movement than allowing the air to escape. A ported enclosure will offer less resistance to cone movement

Not the case near the tuning frequency—the resonance opposes the cone motion, meaning the cone "sees" a very stiff air spring.

i thought it might be possible to get some usable output below the 40hz tuning frequency of the ported design at low volume

It seems unlikely to me that you'll get a meaningful benefit. You can't hear very low frequencies at low SPL anyway (threshold of hearing at 30Hz is circa 60dB SPL), and based on the measurements here you're better off with the ported configuration above 30Hz or so (edit: oops, wrong speaker...). By sealing the port, you're likely to get slightly better (measured) performance at frequencies you can't really hear at the expense of inferior performance at frequencies you certainly can hear.

 

[Goldhamster916]

I missed this before. This is another example of an LLM giving the wrong answer. Sealing a ported enclosure will in fact reduce excursion at frequencies significantly below the port tuning.


Not the case near the tuning frequency—the resonance opposes the cone motion, meaning the cone "sees" a very stiff air spring.


It seems unlikely to me that you'll get a meaningful benefit. You can't hear very low frequencies at low SPL anyway (threshold of hearing at 30Hz is circa 60dB SPL), and based on the measurements here you're better off with the ported configuration above 30Hz or so. By sealing the port, you're likely to get slightly better (measured) performance at frequencies you can't really hear at the expense of inferior performance at frequencies you certainly can hear.

thanks for the responses!

i would try to get to 70DB via EQ for the bass frequencies at 60DB average listening level in order to make them a bit more audible.

thea measurements you posted are from the KEF R3 which is their step down series bookshelf. the speaker i meant here is the KEF Reference 3 Meta which is a higher end floorstander. KEF really made unfortunate naming choices there, i have seen many people confuse the two. this is the Reference 3 Meta:

 

[bmc0]

thea measurements you posted are from the KEF R3 which is their step down series bookshelf. the speaker i meant here is the KEF Reference 3 Meta which is a higher end floorstander.

Oops, sorry about that. I think that what I stated still more-or-less applies because the bass tuning is broadly similar. Since the Reference 3 Meta has two ports, plugging just one (as @Duke mentioned) may be a workable option. This should lower the tuning to approx. 30Hz (half an octave or so).

 

[Goldhamster916]

Oops, sorry about that. I think that what I stated still more-or-less applies because the bass tuning is broadly similar. Since the Reference 3 Meta has two ports, plugging just one (as @Duke mentioned) may be a workable option. This should lower the tuning to approx. 30Hz (half an octave or so).

that would be nice. Care to explain how plugging one port would lower the tubing frequency of the other port?

 

[ssashton]

Sealing the port will increase low end distortion for the same SPL, because the woofer has to work harder. If your main goal is to get more extension, at the cost of some sensitivity, you should tune the port lower, then EQ to get the bass response you desire.

If its tuned to 40hz now, you could try extending the port to get a 30hz tuning.

If you seal the cabinet you'll need a lot of EQ to get back what you lost and that will stress the amp and woofer significantly. Here is a quick example of 2x 6" woofer in sealed vs ported box. As you can see, even the 40Hz tuning is not overtaken by the sealed box until 26Hz so you don't gain a thing until below that frequency. It will be a similar story for the KEF even if the box volume and drivers are different.

Sealed = Blue
40Hz Ported = Green
30Hz Ported = Red

Bare in mind, a difference of 3dB is double or half the amp power required.