port resonances of mains in a 2.1 setup

[dasdoing]

this graph here made me wonder:

if we high-pass a speaker like this at let's say 100Hz (because it's very small) or even higher.....will or wont the port resonances still mess up the sound.
I always thought that by high-passing a ported speaker you would avoid the port problems....but this made me wonder if in a 2.1 steup like this sealing the port would make things better?!

 

[ernestcarl]

Here's a similar graph from the Sceptres:

if we high-pass a speaker like this at let's say 100Hz (because it's very small) or even higher.....will or wont the port resonances still mess up the sound?

With my own monitors measured in-room, the answer is no yes.

I'm also not as optimistic now to believe that the distortion in the mids can be cleared up by high-passing the sub/lower bass (perhaps reduce some of the IMD, but by how much, I don't know). Probably it would clear muddiness by removing the lower part... but the upper-bass and low-mids will not be improved significantly in terms of their own distortion.

Similar can be seen with the KH120 here -- not the port noise, I didn't measure it, yet.

 

[ernestcarl]

sealing the port would make things better?!

I have sealed the ports on all of my studio monitors... not much improvement, actually. I think the distortion finds a way out through the cabinet or elsewhere...

 

[dasdoing]

I have sealed the ports on all of my studio monitors... not much improvement, actually.

It would be intresting to meassure a ported speaker with sealed ports, and then compare to a high-passed one with comparable slope.

when you say "not much improvement" were you actualy observing those resonances, or was this a listening test?

 

[ernestcarl]

were you actualy observing those resonances, or was this a listening test?

The measurements and what I’m hearing in the upper range of the bass and low mids — if my memory is any good. BUT, “chuffing” or vibration type of noise in the lower part at very high SPL disappears with the high pass.

I’ll try to post measurements maybe this weekend or next.

 

[Frank Dernie]

this graph here made me wonder:

if we high-pass a speaker like this at let's say 100Hz (because it's very small) or even higher.....will or wont the port resonances still mess up the sound.
I always thought that by high-passing a ported speaker you would avoid the port problems....but this made me wonder if in a 2.1 steup like this sealing the port would make things better?!

I think the arrows labeled "port resonances" are actually the 3 fundamental internal resonances of the cabinet being "heard" through the port, not actual resonances in the port, so they will be there even with the port blocked and "heard" through the cone of the bass-mid driver which is fairly acoustically transparent.
The actual port resonance, that is the mass of air in the port resonating on the compliance of the air in the cabinet is the main lf peak.
On this basis getting rid of these peaks may involve damping inside the cabinet rather than blanking the port.

 

[dasdoing]

I think the arrows labeled "port resonances" are actually the 3 fundamental internal resonances of the cabinet being "heard" through the port, not actual resonances in the port, so they will be there even with the port blocked and "heard" through the cone of the bass-mid driver which is fairly acoustically transparent.
The actual port resonance, that is the mass of air in the port resonating on the compliance of the air in the cabinet is the main lf peak.
On this basis getting rid of these peaks may involve damping inside the cabinet rather than blanking the port.

that actualy makes sense. I was kind of shocked seeing that graph. you explanation makes more sense

 

[ernestcarl]

they will be there even with the port blocked and "heard" through the cone of the bass-mid driver which is fairly acoustically transparent.

Indeed! The cone is literally paper-thin so I would expect it not to put much of a resistance, if at all.

 

[pjug]

I think the arrows labeled "port resonances" are actually the 3 fundamental internal resonances of the cabinet being "heard" through the port, not actual resonances in the port, so they will be there even with the port blocked and "heard" through the cone of the bass-mid driver which is fairly acoustically transparent.
The actual port resonance, that is the mass of air in the port resonating on the compliance of the air in the cabinet is the main lf peak.
On this basis getting rid of these peaks may involve damping inside the cabinet rather than blanking the port.

So then does this kind of measurement as done by Amir on the Salk WOW1, where uneven parts of the bottom curves still seem to show resonances similar to upper open port curve, support what you are saying? To me it appears that this supports what you are saying.

 

[dasdoing]

So then does this kind of measurement as done by Amir on the Salk WOW1, where uneven parts of the bottom curves still seem to show resonances similar to upper open port curve, support what you are saying? To me it appears that this supports what you are saying.
View attachment 103559

nice find,
I think it does support since he says the port-closed meassurement above is directly at the port. the dip is therefore maintained by the woofer (since it is still there) https://www.audiosciencereview.com/...shelf-speaker-review.14842/page-4#post-462815

 

[Juhazi]

Resonance escaping through the port are mostly cabinet's internal standing waves. How much of those escape through the port depends on port's internal opening's location. Then, having the port opening on the front panel maximizes interferences with on-axis sound. Differences between mediocore and good speakers is huge in this respect.

This is one of the reasons why I hate 2-way loudspeakers... and ported loudspeakers! Fortunately these problems are not easy to hear!

I've done some tests myself

Here port and woofer nearfield with more resolution. I can't say that after blocking the port same energy escapes through the cone... Tweeter was playing simultaneously - difference in level and above 1,5kHz of "W near (with port open)" is artefact, a bit different mic location!

 

[Wombat]

Standing wave calculator:

https://vikash.info/audio/standing_wave_calc/

 

[KSTR]

I once had to investigate port resonances in the typical 500Hz..2kHz range in an active three-way, with the woofer being heavily lowpassed at 300Hz, residing in an enclosure generously filled with damping material. So internal standing waves didn't seem to apply. It turned out the midrange driver was exciting the port's pipe resonances as it was really close, only a few cm away from the ports. Resonance pattern in the FR stayed the effecively same even when the woofer was disconnected and shorted. The ports being old-school round shape and with no flare exaggerated the problem, it seems.

Now in a 2-way, on top of that we really might see internal reflections and standing waves both leaking through the cone and the port. A way to tell the contributors apart would be to excite the port's pipe resonances with a second speaker in closest possible proximity, and with the tested speakers input (or woofer) shorted and un-driven. Internal standing waves are still being excited but at much lower level as the path now is the leakages of cone and port, so any pipe resonances should stand out.

 

[dasdoing]

View attachment 103636

how come the LF roll-off doesn't differ much between the 2 lower graphs?

 

[KSTR]

but this made me wonder if in a 2.1 setup like this sealing the port would make things better?!

"It depends" would be my not very satisfying answer. At any rate, blocking the port or not, the first and most important thing to remember is that subwoofer integration is a speaker design task off-loaded to the user. A set of proper resulting acoustic crossover functions is required for the sub to blend in correctly, typically a 4th-order Linkwitz-Riley at ~80Hz is the target. And the phase contribution of the sub's lower roll-off must be factored in here, too. This calls for very specific corrections to the natural low frequency response of the mains, blocking the ports just requires a different response for the correction so that the same final frequency response target (for magnitude and phase) is reached.

OK, let's assume we have managed that step. Next we assume the port resonance was significantly lower than the XO freq, finally we assume the closed box response doesn't dive down at and above planned XO. Then closing the port is likely to be beneficial (from reduced leakage and no pipe resonances) and the cone will not suffer from a significant excursion increase when playing the lowest notes, and also there is no general efficieny penalty. With the latter two conditions not met, this would add more distortion at higher playback levels and that might spoil the whole plan.

 

[Juhazi]

how come the LF roll-off doesn't differ much between the 2 lower graphs?

"Plug" is the foam plug that came with the speaker. It effectively blocks reflex pipe resonances, but lets some higher freq cabinet resonances and sound through.
"Block" is a clean tennis sock stuffed in the port, and it practically totally blocks all air movement and sound transfer

 

[ernestcarl]

It's been a long while since my port blocking experiments and I can't remember where the mdat files are -- I have too many saved. So I redid the experiment with the S8s with improvements by adding some extra damping material.


Old stuff I have around the house

Inside the cabinet

There is no extra space on the back plate to add damping material.

So I just attached the pipe wrap to the cotton damping material separating the amplifier and driver spaces.

No, I didn't use the adhesive -- yet.

Some additional photos from an old archived review (spark digital):

The monitor temps in my own room after substantial use is much lower than the above.

Yellow trace is with just the foam "support" and foam block.
Blue trace is with additional socks and and pipe wrap.

Turns out my memory is shit. Cancellation dips seem to be reduced with this monitor. Although, I'm almost certain there was very little benefit with the KH120 as they are already flat enough there.

While there is some loss in bass SPL, it's probably acceptable. Excess phase seems to be reduced as well with the use of the monitor's internal DSP high-pass switches -- external DSP results via miniDSP or JRiver is a bit worse as these do not apply any additional phase correction by default.


Group Delay

Distortion

Distortion is lowest with the extra internal damping -- luckily the cabinet is pretty large and could accomodate the entire pipe wrap insulation.

% Distortion

*Notice that the 1kHz distortion peak from a resonance is still there and only a tiny bit attenuated -- at least the dip in the magnitude response disappears -- although there is also some loss, too, but not much.

**Actually, the 1kHz peak is part of (but not entirely) the tweeter's fundamental -- I think -- just very low in level.

Along with my speaker EQ above 1kHz (optimized for off-axis listening), I was able to lower distortion levels substantially.

The Sceptre S8's rated max continuous playback level is 102dB (current website). Previously I saw 105dB from their old promotional brochure, so I just measured them all. I personally prefer not to exceed 100dB -- which is more like 95dB per monitor at 2m at my MLP -- which, of course, is still a rarity.


Needless to say, I purchased another pipe wrap insulation via amazon to take advantage of their one day delivery.

Yeah, previous assumptions were wrong... and now I see the light! LOL

*I probably would not recommend this IF:

1. The ambient temperature of the room gets high -- my room is always 21 degrees celsius or lower.
2. If the speakers are not class D or highly efficient, and esp. if there are no external heat sinks.
3. You love to play music at max volumes for hours and hours.

 

[Juhazi]

Thank you ernestcal!

I revisited my Amphion port test measurements and seems like plugging effect in midrange is minimal (good thing), major difference is in BR range spl and distortion. (I have done box stuffing and bracing tests with other speakers too)

 

[dasdoing]

It's been a long while since my port blocking experiments and I can't remember where the mdat files are -- I have too many saved. So I redid the experiment with the S8s with improvements by adding some extra damping material.


Old stuff I have around the house

View attachment 104037


Inside the cabinet

View attachment 104038


There is no extra space on the back plate to add damping material.

View attachment 104039


So I just attached the pipe wrap to the cotton damping material separating the amplifier and driver spaces.

View attachment 104040
No, I didn't use the adhesive -- yet.

Some additional photos from an old archived review (spark digital):
View attachment 104041

View attachment 104055

The monitor temps in my own room after substantial use is much lower than the above.

View attachment 104043
Yellow trace is with just the foam "support" and foam block.
Blue trace is with additional socks and and pipe wrap.

Turns out my memory is shit. Cancellation dips seem to be reduced with this monitor. Although, I'm almost certain there was very little benefit with the KH120 as they are already flat enough there.

View attachment 104044

While there is some loss in bass SPL, it's probably acceptable. Excess phase seems to be reduced as well with the use of the monitor's internal DSP high-pass switches -- external DSP results via miniDSP or JRiver is a bit worse as these do not apply any additional phase correction by default.


Group Delay

View attachment 104077


Distortion
View attachment 104047

Distortion is lowest with the extra internal damping -- luckily the cabinet is pretty large and could accomodate the entire pipe wrap insulation.

% Distortion
View attachment 104048
*Notice that the 1kHz distortion peak from a resonance is still there and only a tiny bit attenuated -- at least the dip in the magnitude response disappears -- although there is also some loss, too, but not much.

**Actually, the 1kHz peak is part of (but not entirely) the tweeter's fundamental -- I think -- just very low in level.

View attachment 104100


Along with my speaker EQ above 1kHz (optimized for off-axis listening), I was able to lower distortion levels substantially.

View attachment 104051
The Sceptre S8's rated max continuous playback level is 102dB (current website). Previously I saw 105dB from their old promotional brochure, so I just measured them all. I personally prefer not to exceed 100dB -- which is more like 95dB per monitor at 2m at my MLP -- which, of course, is still a rarity.


Needless to say, I purchased another pipe wrap insulation via amazon to take advantage of their one day delivery.

View attachment 104053

Yeah, previous assumptions were wrong... and now I see the light! LOL

*I probably would not recommend this IF:

1. The ambient temperature of the room gets high -- my room is always 21 degrees celsius or lower.
2. If the speakers are not class D or highly efficient, and esp. if there are no external heat sinks.
3. You love to play music at max volumes for hours and hours.

pretty impresive,

unfortunatly my room temps get high, my amps are not class D, I like loud music at weekends lol

but my huge slot port (23cmx23cm) is at the bottom and there shouldn't be much heat escaping there anyways.

those are some nice speakers btw. when I buy new speakers I will defently go coax, too

 

[andreasmaaan]

I always thought that by high-passing a ported speaker you would avoid the port problems....but this made me wonder if in a 2.1 steup like this sealing the port would make things better?!

In general, high-passing a speaker will do little to remedy port issues (whether they are caused by pipe resonances, standing waves, or something else) at frequencies above the HPF frequency, because the HPF will do nothing to reduce the amplitude of excitation frequencies.

In general, the only effective solution to such problems is acoustical, i.e. stuffing, plugging, rebuilding, etc. etc.