Audibility of port resonances

Vineethkumar01

Well-Known Member
Joined
May 19, 2015
Messages
473
Points
93
Location
Bangalore
Hi All,
I have attached nearfield measurements that I took of a 2-way bass reflex speaker .
The blue line is the nearfield response of the midwoofer.
The red line shows the nearfield response of the port that is at the back of the box.
As can be seen from the pic, a frequency of around 38.2 Hz is where the woofer output is minimum and port output is maximum.In my understanding, this corresponds to the tuning frequency of the bass reflex box. I alsoa reasonably high Q peak around 720 Hz and other peaks at its integer multiples. I am suspecting that is a port resonance and something which we dont want. Please correct me if there is any inference that I have made wrongly.
In this context my question is as follows. Have anybody had any experience hearing and identifying port resonances.
1) How does it sound like??
2) What kind of music when we play is it most audible?
 

Attachments

  • port_woofer.JPG
    port_woofer.JPG
    158.8 KB · Views: 33
Nice post, @Vineethkumar01. Appreciate your passion to get to the fundamentals, experiential and through experimentation - like a true researcher. I hope that your study posts get more responses and we all benefit from some collective insight.

I am a layman when it comes to this technology. But reading your post I wondered if the (undesired) port resonances might change in amplitude as the distance from the front wall (behind the port) is changed? Worth trying that out?
 
Nice post, @Vineethkumar01. Appreciate your passion to get to the fundamentals, experiential and through experimentation - like a true researcher. I hope that your study posts get more responses and we all benefit from some collective insight.

I am a layman when it comes to this technology. But reading your post I wondered if the (undesired) port resonances might change in amplitude as the distance from the front wall (behind the port) is changed? Worth trying that out?
Thanks a lot for your reply.

Port resonances as I understand are caused by the dimesions of the port itself. So ideally, it should be independent of the location at which measurement was done (as long as it is far away from immediate reflecting surfaces). Also, nearfield measurement technique is usually classified as a high SNR (signal-to-noise ratio) measurement. When done properly, at low enough frequencies. the measurement does not even see the surface where the port is mounted on. Therefore it will also not see the walls and the surrounding environment. Usually, such port-related artifacts are got rid of, at least to some extent, by stuffing the speaker box properly; I think.

Apart from solving this problem, my primary objective in asking this question here is to detect the problem itself using our ears, without the calibration mic. I mean can we do it with our ears and if so, how?

Thanks again
 
Last edited by a moderator:
Hi All,
I have attached nearfield measurements that I took of a 2-way bass reflex speaker .
The blue line is the nearfield response of the midwoofer.
The red line shows the nearfield response of the port that is at the back of the box.
As can be seen from the pic, a frequency of around 38.2 Hz is where the woofer output is minimum and port output is maximum.In my understanding, this corresponds to the tuning frequency of the bass reflex box. I alsoa reasonably high Q peak around 720 Hz and other peaks at its integer multiples. I am suspecting that is a port resonance and something which we dont want. Please correct me if there is any inference that I have made wrongly.
In this context my question is as follows. Have anybody had any experience hearing and identifying port resonances.
1) How does it sound like??
2) What kind of music when we play is it most audible?
These are Interesting questions @Vineethkumar01,
I have read about port chuffing in reviews of ported subwoofers but have not heard it myself.
My understanding is that whenever air is forced out of a hole there is some noise caused by the turbulence and current (of air)…think nose.
larger the amount of air and faster it is forced out (larger diaphragms, louder) would cause more noise.
I recall some speaker designers came up with engineering solutions - golf ball type dimples on port outlets (B&W), tube channels (PMC, EPOS?)
Good initiative in measuring this!
 
Found these


and

 
Found these


and

Hi.
Thanks for the links.
In these links, they are discussing about port chuffing.
As per my understanding, chuffing usually occurs when the port is trying to output frequencies at which it is supposed to give its peak output. In our case, this region is around the "system resonance" frequency which is about 38.2 Hz as shown in the plot that I attached. The port is the "star" in this range of frequencies. I think chuffing happens because the flow of air inside the port turns from a laminar flow to a turbulent flow, which means instead of air being able to move smoothly through the port, it is finding it hard to move so smoothly. This could be due to design of the port itself due to various factors.

One of the techniques that is used to mitigate chuffing to some extent is to ensure that when we model the low frequency bass reflex alignment for woofer using the T/S parameters in softwares like WinISD etc, we keep the dimensions of the port and the resultant port tuning such that the port velocity is under 17-20 meters/sec (to be conservative) even at sufficiently high power levels that we want the system to play.

The issue that we are discussing here has more to do with the port output "leaking" into the midrange. These are the frequencies where we see narrow peaks in the port output (red line in the figure that i attached in previous post) around 700 Hz. I think at these frequencies, the port is supposed to "play dead", like its not there. The entire sound radiation should be taken care of by the midwoofer to whom we have entrusted this duty to.. However, in this particular case, the port is sticking out like a sore thumb and calling attention to itself.

Thanks
Vineeth
 
Saw one video where the outer surface of port inside box is damped, also in Cornwall, tweeter section is heavily damped...

I don't have a clue about what I'm talking about :) just threw that in
 
I think what you're referring to as port resonance is pipe resonance of the port, right? That resonance is mainly academic for the most part. The farther from the port you measure, the lesser in magnitude it gets.
Saw one video where the outer surface of port inside box is damped
Yes, that is one way to lower the magnitude of the resonance. Another way is to use a flexible pipe of suitable material for the port. KEF does that for some of its speakers.

also in Cornwall, tweeter section is heavily damped...
That is to damp the outside of the tweeter "enclosure" so that the in-box resonances do not in turn cause the tweeter to resonate thus causing a "problem" with tweeter output.

Harman has a whole paper/s on this. I'll see if I can locate it on my machine and post it. Too much junk to wade through before I find it. ;)

In MY opinion, like I said before, for the most part, these are academic but if one can implement these mitigation easily, why not? If one were anal like me, I would. I'd sleep better at night.
 
I think what you're referring to as port resonance is pipe resonance of the port, right? That resonance is mainly academic for the most part. The farther from the port you measure, the lesser in magnitude it gets.

Yes, that is one way to lower the magnitude of the resonance. Another way is to use a flexible pipe of suitable material for the port. KEF does that for some of its speakers.


That is to damp the outside of the tweeter "enclosure" so that the in-box resonances do not in turn cause the tweeter to resonate thus causing a "problem" with tweeter output.

Harman has a whole paper/s on this. I'll see if I can locate it on my machine and post it. Too much junk to wade through before I find it. ;)

In MY opinion, like I said before, for the most part, these are academic but if one can implement these mitigation easily, why not? If one were anal like me, I would. I'd sleep better at night.
Thanks a lot @keith_correa . Yes I think the correct term is pipe resonance.
The port resonance is at the port tuning frequency I think. Over at htguide.com & diyaudio.com, I have seen FM augerpro doing detailed studies on this. So I thought, it would be very audible. Hence my questions about the audibility of it.
Thanks in advance for the paper also if you are able to find it... :)
 

If you haven't see this before, you might find this interesting. The wonders of science! :cool:

Edit: I found this in my "notes". This is for only ONE pipe resonance:

One pipe resonance occurs at F=c/(2*L+0.8*d)
where:
c = speed of sound (m/s)
L = length of vent (m)
d = diameter of vent (m)

How other resonances are calculated, I have no clue. @Vineethkumar01, with your background, maybe you can help and educate.

Over at htguide.com & diyaudio.com, I have seen FM augerpro doing detailed studies on this. So I thought, it would be very audible. Hence my questions about the audibility of it.
If you have seen augerpro's posts about this on DIYA then I'm quite sure he has linked to the Harman paper that I was referring to. You should also look in his posts for the paper from N.B. Roozen formerly of Philips that I had pointed him to. Augerpro was also doing extensive tests on ideal enclosure material/s to build speaker enclosures of. THAT, is WAY more audible than pipe resonances et al.
 
Last edited:
Hi.
Thanks for the links.
In these links, they are discussing about port chuffing.
As per my understanding, chuffing usually occurs when the port is trying to output frequencies at which it is supposed to give its peak output. In our case, this region is around the "system resonance" frequency which is about 38.2 Hz as shown in the plot that I attached. The port is the "star" in this range of frequencies. I think chuffing happens because the flow of air inside the port turns from a laminar flow to a turbulent flow, which means instead of air being able to move smoothly through the port, it is finding it hard to move so smoothly. This could be due to design of the port itself due to various factors.

One of the techniques that is used to mitigate chuffing to some extent is to ensure that when we model the low frequency bass reflex alignment for woofer using the T/S parameters in softwares like WinISD etc, we keep the dimensions of the port and the resultant port tuning such that the port velocity is under 17-20 meters/sec (to be conservative) even at sufficiently high power levels that we want the system to play.

The issue that we are discussing here has more to do with the port output "leaking" into the midrange. These are the frequencies where we see narrow peaks in the port output (red line in the figure that i attached in previous post) around 700 Hz. I think at these frequencies, the port is supposed to "play dead", like its not there. The entire sound radiation should be taken care of by the midwoofer to whom we have entrusted this duty to.. However, in this particular case, the port is sticking out like a sore thumb and calling attention to itself.

Thanks
Vineeth
I agree. I was referring to unwanted noise from the ports (chuff?). If I don’t hear this I assume it’s all fine. I was probably unclear in my understanding of the questions.

If you haven't see this before, you might find this interesting. The wonders of science! :cool:

Edit: I found this in my "notes". This is for only ONE pipe resonance:

One pipe resonance occurs at F=c/(2*L+0.8*d)
where:
c = speed of sound (m/s)
L = length of vent (m)
d = diameter of vent (m)

How other resonances are calculated, I have no clue. @Vineethkumar01, with your background, maybe you can help and educate.

If you have seen augerpro's posts about this on DIYA then I'm quite sure he has linked to the Harman paper that I was referring to. You should also look in his posts for the paper from N.B. Roozen formerly of Philips that I had pointed him to. Augerpro was also doing extensive tests on ideal enclosure material/s to build speaker enclosures of. THAT, is WAY more audible than pipe resonances et al.
Sure. Thanks for the information and equation.. :) I am already going through the KEF paper. I will go through the Philips paper and augerpro's posts once more, during my breaks from work.. :) Once I gather more info, I will try and summarize the relevant port-related (good practices) material here.

From my preliminary understanding, I think the equation you gave is for the finding out the fundamental resonant frequency of the column of air enclosed by the port. I remember seeing a similar equation long back over at the techtalk parts express forum. Such resonant behaviour (identified by narrow peaks in frequency response) will also seen at the harmonics of this fundamental frequency (F). ie at 2*F, 3*F, 4*F.... for the system under consideration. In the plot that I had posted also, it can be seen upto the 1st harmonic clearly, I think. (Ideally we dont want to hear any of those coming out from the port).

Edit:
I went through the KEF LS 50 white paper, Roozen's paper and relevant parts of FM augerpro's related thread here:

The Roozen paper was very interesting to read. It calls the kind of resonant behaviour I saw in plots posted above above as acoustic port resonances. In general, one way in which noises from the port occur is due to "chuffing" (vortex formation), which is broadband in its behaviour, meaning it is spread out over a wide range of frequencies. The other kind of noise is what we saw as sharp peaks (acoustic port resonances) in my plot, which occur approximately at integer multiples of the frequency, whose wavelength is equal to half the effective length of the port.

In the case that I presented above, This occurs at F= (n*c/2*L), where n is a positive integer (takes values 1, 2,3,... corresponding to the fundamental frequency (n=1), and the harmonics (other values of n =2,3,....) ), c = speed of sound, L = effective length of the port. In my particular example, I had a rectangular slot port with effective length (L)= 23 cm in length, c= 350 m/s, etc. This gives the fundamental frequency of acoustic port resonance at around 730 Hz, 1st harmonic occurring around 2*730 = 1460 Hz,.. and so on... This behaviour is clearly visible in the above graph.

In order to solve these kind of port resonances, even acadamically, we might need to adopt advanced port designs. Harman has a special flexible port. Roozen paper says about a specific design with holes poked at the center and 3/4the distance from one end. Augerpro's experiments also confirm their effectivety, and he has produced his own variations of the above ports. These specialized ports apart from the KEF one achieve high port resonance suppression, but seems to cause some amount of harmonic distortion. Again all these occur at high drive levels (when we are try to play music at high volumes).

So, even if the port resonances themselves are audible or not in real world uses cases (specifically with music) if we take care of these issues in port design, we might be able to sleep better at night.... :) This answers my questions about audibility of these and I am happy with the strategies I learnt for their potential mitigation..

Thanks everyone for your contributions.

Thanks
Vineeth
 
Posting this thread here in case someone finds it useful in future:
This deals with a simple, yet very effective port resonance suppression technique employed in the Technics SB-C700 speaker. Measurements say that it has been even more effective than KEF design in terms of suppresion of resonant peaks. Nothing is mentioned about the overall harmonic distortion though.
 
The Marantz PM7000N offers big, spacious and insightful sound, class-leading clarity and a solid streaming platform in a award winning package.
Back
Top