Ron E. You made several claims with no applicable evidence. I actually showed evidence, the tone burst plot, and made clear what was conjecture regarding the Blauert curve. There is a curve for the audibility of group delay below 300 Hz, and it's likely the time constant of one wave at each frequency as I said before. I haven't verified that.
The graph of the ported system response with the sinewave burst input that grows over four cycles. You saw that right? You just figured that was all garbage and fake.
This is DIYAudio, so rather than just "talk talk talk" go do the work for yourself. Run the LTSpice model and look at sine bursts or model a plucked string as a decaying exponential sinewave. The step response isn't what I'm talking about. Or if you are not a physicist or an electrical engineer, and have zero faith in the LTSpice model, go get two microphones and measure the port and woofer output for a five cycle sinewave burst and graph it for yourself.
The speed of sound? You seem to imply the woofer and the air mass in the port move together linked at the speed of sound with no effective air spring between them. You realize it is a resonator? It stores energy in that air spring between the cone and the mass of the air in the port tube. You will notice that around the resonance the woofer and port or PR are both moving out together and in together. Stretching or compressing the air spring between them. According to you one should go out while the other goes in, because they are coupled "at the speed of sound". This guy explains the spring coupling of the port mass to the woofer.
You throw out Fletcher Munson with no rational for what you're trying to say. The lowest note on a six string bass guitar is 31 Hz, so it's the reason people tune ports that low. It would appear to me that according to that curve the ear would be more sensitive to a drop in bass level than it is in the midrange. According to Fletcher Munson, see the graph below, a 30 Hz tone at 90 dB only has to drop 25 dB down to 65 dB to become inaudible. So the modulation of the note amplitude by a ported speaker should be obvious. Thanks for finding evidence to support my position. Bass guitar notes are percussive plucked note with an instant attack and an exponential decay. The tone burst like attack, the first most impactful cycles, get crushed by a ported speaker that takes several cycles to come up to level.
It's actually possible to sort of make a sealed speaker as bad as a ported speaker. The sealed speaker has a 2nd order roll off. The ported speaker has a 4th order roll off. So it's possible to cascade a 2nd order high pass filter with a sealed speaker to get rid of that extra low bass the ported speaker people don't want. It would also be possible to add a resonator to delay the attack and decay of a sealed speaker to get the whole effect.
The graph of the ported system response with the sinewave burst input that grows over four cycles. You saw that right? You just figured that was all garbage and fake.
This is DIYAudio, so rather than just "talk talk talk" go do the work for yourself. Run the LTSpice model and look at sine bursts or model a plucked string as a decaying exponential sinewave. The step response isn't what I'm talking about. Or if you are not a physicist or an electrical engineer, and have zero faith in the LTSpice model, go get two microphones and measure the port and woofer output for a five cycle sinewave burst and graph it for yourself.
The speed of sound? You seem to imply the woofer and the air mass in the port move together linked at the speed of sound with no effective air spring between them. You realize it is a resonator? It stores energy in that air spring between the cone and the mass of the air in the port tube. You will notice that around the resonance the woofer and port or PR are both moving out together and in together. Stretching or compressing the air spring between them. According to you one should go out while the other goes in, because they are coupled "at the speed of sound". This guy explains the spring coupling of the port mass to the woofer.
You throw out Fletcher Munson with no rational for what you're trying to say. The lowest note on a six string bass guitar is 31 Hz, so it's the reason people tune ports that low. It would appear to me that according to that curve the ear would be more sensitive to a drop in bass level than it is in the midrange. According to Fletcher Munson, see the graph below, a 30 Hz tone at 90 dB only has to drop 25 dB down to 65 dB to become inaudible. So the modulation of the note amplitude by a ported speaker should be obvious. Thanks for finding evidence to support my position. Bass guitar notes are percussive plucked note with an instant attack and an exponential decay. The tone burst like attack, the first most impactful cycles, get crushed by a ported speaker that takes several cycles to come up to level.
It's actually possible to sort of make a sealed speaker as bad as a ported speaker. The sealed speaker has a 2nd order roll off. The ported speaker has a 4th order roll off. So it's possible to cascade a 2nd order high pass filter with a sealed speaker to get rid of that extra low bass the ported speaker people don't want. It would also be possible to add a resonator to delay the attack and decay of a sealed speaker to get the whole effect.
Last edited:
There is a difference. I understand this model because I wrote it and didn't just download it off the internet, so you can forget about all your silly implications I don't understand how this works. I can convolve a sine wave into it and show a sine wave going through the transfer function but I haven't done that yet because it doesn't seem that important. Have you looked at a sealed box system in your "found" model yet? Which ported box systems have you looked at?
30Hz ringing is not important because you are not going to hear it because the ear isn't that sensitive down that low, and also because your room is ringing at multiple frequencies for much longer than the speaker is. Many rooms riing for a long time at low frequencies, depending on construction. These I've shown die out rather quickly,
From just your last paragraph, everything you say about ported loudspeakers is suspect anyway. You have chosen a hill to die on. I don't see life in black and white.
30Hz ringing is not important because you are not going to hear it because the ear isn't that sensitive down that low, and also because your room is ringing at multiple frequencies for much longer than the speaker is. Many rooms riing for a long time at low frequencies, depending on construction. These I've shown die out rather quickly,
From just your last paragraph, everything you say about ported loudspeakers is suspect anyway. You have chosen a hill to die on. I don't see life in black and white.
Ron E. Again with unsubstantiated conjecture. I don't find that at all compelling as I have run models and made measurements that agree with the models. Go do the work and present it here if the results are different. The dynamic equations for ported systems are readily available. If you look at the notes in the files I uploaded the author has references. I guess if you took graduate level dynamics classes, you have the training to easily build your own models. Or look at the model I presented and use your knowledge to correct what you think is wrong. It's pretty common to model spring, mass, damper systems using electronically analogous parts with SPICE. The notes in the files I uploaded describe how to do it.
I'm not sure what model you have or if a convolution of a 5 cycle burst with the impulse response would help you, but go for it and present it. If your impulse response was derived from the infinite time frequency response rather than a transient simulation or measurement, my first thought is that it will not accurately represent the transient response of this type of system. Judging from your history as recorded in your Reaction Score you won't do it. It would be interesting to capture a step response or impulse response directly with something like Audacity and do the convolution. While your at it convolve the system impulse response with a plucked bass note. The old Stereophile test CD has three tracks with plucked bass notes. Watch how the first few cycles get reduced and flattened.
I'm not sure why you would question there being an audible difference between a two piece spring coupled system that forms a fourth or fifth order high pass response that relies on high Q resonances to fill in the lower octave and a system with well damped second order high pass response. From a dynamic systems perspective these are obviously huge differences. So it's not a stretch to understand they would be audibly different. I find it hilarious that on the same forum there are people claiming to hear huge differences in DACs, cables, fuses, capacitors and line cords that are identical in response with in a fraction of a dB and here we have people claiming there's no audible difference, what so ever, in two systems that have wildly different response characteristics. The sealed woofer that can pretty faithfully reproduce a 5 cycle sine burst waveform while a typical ported system that makes a rising and falling amplitude resonant burp in response to the same sinewave burst. It invents a waveform that looks nothing like the input. It's bell that rings when you hit it. That is the fundamental concept behind the resonant port or PR.
Not really a hill to die on, just presenting some observations made from simulations, measurements and listening. The Fletcher Munson curve was a great addition.
As further confirmation, my small LS3/5a inspired speaker with a sealed woofer out scored all the other 2-way speakers in the local speaker building contest Oct 5th. Most of the other contestants had ports or passive radiators.
I'm not sure what model you have or if a convolution of a 5 cycle burst with the impulse response would help you, but go for it and present it. If your impulse response was derived from the infinite time frequency response rather than a transient simulation or measurement, my first thought is that it will not accurately represent the transient response of this type of system. Judging from your history as recorded in your Reaction Score you won't do it. It would be interesting to capture a step response or impulse response directly with something like Audacity and do the convolution. While your at it convolve the system impulse response with a plucked bass note. The old Stereophile test CD has three tracks with plucked bass notes. Watch how the first few cycles get reduced and flattened.
I'm not sure why you would question there being an audible difference between a two piece spring coupled system that forms a fourth or fifth order high pass response that relies on high Q resonances to fill in the lower octave and a system with well damped second order high pass response. From a dynamic systems perspective these are obviously huge differences. So it's not a stretch to understand they would be audibly different. I find it hilarious that on the same forum there are people claiming to hear huge differences in DACs, cables, fuses, capacitors and line cords that are identical in response with in a fraction of a dB and here we have people claiming there's no audible difference, what so ever, in two systems that have wildly different response characteristics. The sealed woofer that can pretty faithfully reproduce a 5 cycle sine burst waveform while a typical ported system that makes a rising and falling amplitude resonant burp in response to the same sinewave burst. It invents a waveform that looks nothing like the input. It's bell that rings when you hit it. That is the fundamental concept behind the resonant port or PR.
Not really a hill to die on, just presenting some observations made from simulations, measurements and listening. The Fletcher Munson curve was a great addition.
As further confirmation, my small LS3/5a inspired speaker with a sealed woofer out scored all the other 2-way speakers in the local speaker building contest Oct 5th. Most of the other contestants had ports or passive radiators.
I didn't stay long enough to know the results, but there were lots of diy builds and all the ones I listened to sounded great.
Thanks to all the folks who gave their time and energy to put the event on and share their amazing builds.
Thanks to all the folks who gave their time and energy to put the event on and share their amazing builds.
Last edited:
Equal loudness contours show how people of average hearing respond to continuous sine waves, they do not show how long it takes to perceive them.
They do show that at ~30Hz, a difference of only 5dB sounds twice as loud, the port response can easily increase level by that or more around Fb.
Bass guitar note envelope depends on the instrument, strings, and playing style.
Upright acoustic bass note envelopes may have a leading transient and a build up in level as the instrument's Helmholtz resonance increases level, then begins to decay.
"Crushed" is a subjective term.
While the port is coming up to level, the driver's excursion is largely undamped, it's level contributes more than the port, then is reduced around Fb as the port's output increases.
No question there is a difference in transient response between a ported and sealed system as STV's tests clearly show, but obviously there is a large percentage of the population that don't find it audibly "crushes" the transients which lead the fundamental frequencies.
Telling people what they can or can't hear does not change what they can or can't detect, or what they find most important.
I still long for the years when I could hear 4000 Hz louder than 40 Hz..
Art
Thanks for posting a link to this paper. It is a much better explanation of everything I said in my posts.
It is very clear and complete, like a college level physics or dynamic systems book, and shows undamped and optimally damped configurations for ported speakers. The undamped configurations are useful just to see what happens in the limit. The build up of the port resonance with a sinewave input looks like what I have measured and calculated myself. It doesn't get any better than this.
Grimm Audio builds technically advanced and highly regarded studio monitors. As I can't afford a pair of the LS1 speakers, I recently mocked them up using a baffle with similar dimensions loaded with the commercially available versions of the SEAS drivers they use. While my prototype is an ugly hack jobs based on the left overs from a previous build, they sound amazing. The woofer is in a sealed box.