Can a ported sub sound as good as a sealed one?
Low q drivers designed only for ported vs drivers designed only for sealed.Both for sq no spl
The big difference is the efficiency
But what about sq? I know that sealed is better but people didn't compare dubs designed for ported and sealed jut one that can do both with a qts around 0.45
Low q drivers designed only for ported vs drivers designed only for sealed.Both for sq no spl
The big difference is the efficiency
But what about sq? I know that sealed is better but people didn't compare dubs designed for ported and sealed jut one that can do both with a qts around 0.45
Perhaps you'll be using a multi-sub arrangement with an overall response target. Closed might be a little easier to use this way.
Group delay is a negative contributor to the "sq" of a ported enclosure. With a low-q woofer that has a strong enough motor and a high enough EBP, you can combine a smaller than normal airspace with a very low vent tuning to keep group delay low enough in the common passband of a subwoofer so that it won't stand out too far over the lower group delay of a sealed sub. I've done some testing on this with some very powerful 15's and have had great success with airspace as low as 1.5cf and vent tuning in the mid 20's. The vent area and length will obviously be large, but if you use a slot port with a sufficient cross sectional area, you can build it into the enclosure so the net size of the box (airspace + port volume) won't be alarmingly big.
Define "sound as good". Ported subwoofers have a few advantages, as well as a few disadvantages when compared to sealed subwoofers. I know that's very general, but in keeping with the general nature of the discourse. Keep in mind that there are confounding variables that serve to taint any subjective impressions that the listener may typically form regarding a sealed or vented system's sound quality (SQ).
The low-Q drivers are often well suited to QB3 (quasi-Butterworth) alignments. These have a more gradual roll-off around the cut-off frequency than do the higher-Q B4 alignments. These give you some of the benefits of a vented enclosure, with a transient response that is somewhat like that of a sealed-enclosure 2nd-order Qtc=0.71 system. It's a quite good trade-off, assuming that vent resonances are well handled, and also that port chuffing is kept to a minimum at high power levels.
In that situation, the low-frequency cut-off points of the two systems will be quite different. It's nigh on impossible to compare their "sound quality" (SQ) as such, as the listener's perception is clouded by the quite different frequency responses of the two systems. In the traditional audiophile folklore, as passed down through the ages, the sealed system has "better" sound quality due to its "tighter" bass response. However, once the differences in output levels are accounted for, the sealed system sounds surprisingly like the vented system, and vice versa. Of course, the sealed system has a cleaner response, by virtue of the fact that it has no port chuffing nor any port resonances.
Last edited:
Personally I think If the resonance is 30Hz or below, you are not going to hear the port ring, so it doesn't matter at all. It probably doesn't matter if the port is tuned to 40Hz even.
If you look at some stereophile ported measurements, it is fairly clear from the port response shape that they are using woofers with very high Q's and these reviewers are digging them.
If you look at some stereophile ported measurements, it is fairly clear from the port response shape that they are using woofers with very high Q's and these reviewers are digging them.
Two woofers in a sealed box are about the equivalent output of one woofer in a ported box for low frequency SPL capability. That's why they are so popular. To get really low bass a sealed box will need some EQ unless it is huge and has very carefully selected woofers. With EQ available, as with plate amps or system DSP, I go with sealed every time. The required box size is smaller than vented. The sealed box has output declines less rapidly with low frequency than vented and has output down to zero Hz. So at some low frequency the vented design output drops below that of the sealed box. It would seem that the group delay of the vented designs is always audible. The best bass is likely going to come from a servo controlled sealed woofer. Of course the port output always has lag as the resonance builds. So that's not great. It also doesn't stop immediately when the note stops. So the initial transient output from a ported speaker is only what the woofer contributes. The port takes a few cycles to produce it's full output. Of course with plucked notes that have a fast attack and exponential decay, this mutes the attack, the first few cycles of the wave. That beautiful flat frequency response of the ported system is the long term response. To see the initial first cycle transient response, subtract the port output.
Last edited:
Many people cant tell the difference real life.
With a Quick A/B test, most prefer ported because the apparent bass of ported seems like " more bass"
Depends what your comparing.
Generalized are we comparing a dinky turd speaker 65 Hz tuned or 45 Hz tuned.
Transient sucks, the end. Even then A/B test, ported sounds like more bass with tradeoffs.
Most dont hear the trade offs anyways.
Ron E nailed it. Once you have a driver down to 30 Hz tune or even 25 Hz
Most music wont reveal any transient garbage from ported. Who cares.
Unless your music Genre finds it.
Big difference what people " imagine" being 30 Hz is more like 45 Hz or 55 Hz
Thinking oh little 6.5" or 8" does that = nope
So if your tuned low 30 to 25 Hz . You dont have problems
45 to 55 Hz is actually accurate.
Otherwise hitting real 30 to 35 Hz content in music.
You might get picky and go to sealed.
Or watching movies with annoying sound effects wont flutter the speakers
with big bass dynamics.
Being non musical hits, Who cares a ported system with multiple drivers
cuts down flutter or distortion for movie sound effects.
I get picky for music, and only music that actually hits low.
Or in a car system with gracious bass from the cabin.
Sealed all the way. No basic transfer function in a model tells you what is happening.
And people nit pick nonsense trying to make magical models.
With a Quick A/B test, most prefer ported because the apparent bass of ported seems like " more bass"
Depends what your comparing.
Generalized are we comparing a dinky turd speaker 65 Hz tuned or 45 Hz tuned.
Transient sucks, the end. Even then A/B test, ported sounds like more bass with tradeoffs.
Most dont hear the trade offs anyways.
Ron E nailed it. Once you have a driver down to 30 Hz tune or even 25 Hz
Most music wont reveal any transient garbage from ported. Who cares.
Unless your music Genre finds it.
Big difference what people " imagine" being 30 Hz is more like 45 Hz or 55 Hz
Thinking oh little 6.5" or 8" does that = nope
So if your tuned low 30 to 25 Hz . You dont have problems
45 to 55 Hz is actually accurate.
Otherwise hitting real 30 to 35 Hz content in music.
You might get picky and go to sealed.
Or watching movies with annoying sound effects wont flutter the speakers
with big bass dynamics.
Being non musical hits, Who cares a ported system with multiple drivers
cuts down flutter or distortion for movie sound effects.
I get picky for music, and only music that actually hits low.
Or in a car system with gracious bass from the cabin.
Sealed all the way. No basic transfer function in a model tells you what is happening.
And people nit pick nonsense trying to make magical models.
I'm not entirely convinced that that is the case.
Is that because a servo control will help to reduce distortion at higher excursions within the nominal "linear" operating range of the woofer?
Although a vented system relies on the port resonance to control its dynamic properties, reproducing short-duration, half-cycle, pulse-like waveforms does not allow time for the "resonance" to build.
But doesn't it stop relatively quickly? After all, when the electrical input signal to the driver's voice-coil is altered, the driver tends to "do as it's told". Then, the driver imparts a somewhat controlling influence on the port, bringing its vibratory response under control.
The following plot, taken from Small (1973), shows the step response of many different vented-box low-frequency alignments. For the popular QB3 and B4 alignments, the sound pressure response of the system dies away very quickly, even though it "doesn't stop immediately".
If the woofer is not moving much at the box tuning frequency, Fb, then how can the woofer contribute much at all to the sound pressure output?
Referring to the previous step response plots, how is it possible for the system response to rise and then decay away within about 2 cycles if it is taking the port "a few cycles to produce its full output"?
If we have an input signal whose frequency content is centred on the port tuning frequency, Fb, then how is it possible to get much sound pressure output from the woofer when its motion is at a minimum? Wouldn't it instead be the case that when there is no woofer motion, then there is no sound pressure output from the woofer? Hence, subtracting the port output from the total response would leave only a very low sound pressure output from the woofer, wouldn't it?
Benson (1969), in his journal paper Theory and Design of Loudspeaker Enclosures, has studied this and has produced the following plot of the individual contributions of the port and the driver to the total response.
Last edited:
I'm sorry I don't have time to address all of your questions right now, but I will go for a few.
Figure 5-26 Graphs what I think is a threshold for audibility of group delay. I haven't tested if I can hear it at this level. Using WinISD I was able to produce a vented system that had group delay less than several times that threshold value. See the graph below.
Figure 2. Blue: Group delay of a 10" woofer in a ported system. Black: Blauert and Law's threshold for audibility.
I extrapolated the Blauert and Law's curve below the 300 Hz, so it's likely totally wrong. Other discussion at Audio Science Review I found discuss the audibility of a group delay equal to the time for one cycle. At 50 Hz, that would be 20 ms. I don't know what it should be. It is apparent to me that I can easily distinguish ported from sealed systems, likely due to the total absence of bass below the port frequency and likely the crushed transients in percussive or plucked rapidly decaying notes produced by the port.
Regarding the plots from the TS paper. You can see "a few cycles" of ringing in many of those. Some five cycle tone burst response plots or the response to a exponentially decaying plucked bass note would better illustrate my points. I will have to look at finding or generating those from an LTSpice model of a ported system I have someplace.
The frequency response graphs represent the output at steady state long after all the startup transients have subsided. So the system will produce that magnitude response after many cycles. The response of the first cycle doesn't look anything like that. My first description may not have been precise, saying "subtract the port output" . For the first cycle of a tone burst going to the woofer, at a frequency in the port resonance range, the woofer will essentially behave as if it were in a sealed box. The output of the port for that first cycle will be near zero. That's because the air mass in the port begins at a stand still and is coupled to the woofer by a spring. So that first cycle output will be what that woofer would produce nearly the same output as it would in a sealed box of that same size. The mass of air in the port, that has to move to produce port output is much like the mass of a passive radiator. It is coupled to the woofer by the spring formed by the air in the box between the port and the woofer. So think about two masses coupled by a spring. Move one mass slowly and the other lags and follows along in the same direction. Move the driven mass at the resonant frequency of the passive mass and spring system and after a cycle or two the masses will move in opposite directions, stretching and compressing the spring. That's what is happening in a ported or PR system. It is only after the port is resonating that it loads the woofer and impedes it's motion.
Figure 5-26 Graphs what I think is a threshold for audibility of group delay. I haven't tested if I can hear it at this level. Using WinISD I was able to produce a vented system that had group delay less than several times that threshold value. See the graph below.
Figure 2. Blue: Group delay of a 10" woofer in a ported system. Black: Blauert and Law's threshold for audibility.
I extrapolated the Blauert and Law's curve below the 300 Hz, so it's likely totally wrong. Other discussion at Audio Science Review I found discuss the audibility of a group delay equal to the time for one cycle. At 50 Hz, that would be 20 ms. I don't know what it should be. It is apparent to me that I can easily distinguish ported from sealed systems, likely due to the total absence of bass below the port frequency and likely the crushed transients in percussive or plucked rapidly decaying notes produced by the port.
Regarding the plots from the TS paper. You can see "a few cycles" of ringing in many of those. Some five cycle tone burst response plots or the response to a exponentially decaying plucked bass note would better illustrate my points. I will have to look at finding or generating those from an LTSpice model of a ported system I have someplace.
The frequency response graphs represent the output at steady state long after all the startup transients have subsided. So the system will produce that magnitude response after many cycles. The response of the first cycle doesn't look anything like that. My first description may not have been precise, saying "subtract the port output" . For the first cycle of a tone burst going to the woofer, at a frequency in the port resonance range, the woofer will essentially behave as if it were in a sealed box. The output of the port for that first cycle will be near zero. That's because the air mass in the port begins at a stand still and is coupled to the woofer by a spring. So that first cycle output will be what that woofer would produce nearly the same output as it would in a sealed box of that same size. The mass of air in the port, that has to move to produce port output is much like the mass of a passive radiator. It is coupled to the woofer by the spring formed by the air in the box between the port and the woofer. So think about two masses coupled by a spring. Move one mass slowly and the other lags and follows along in the same direction. Move the driven mass at the resonant frequency of the passive mass and spring system and after a cycle or two the masses will move in opposite directions, stretching and compressing the spring. That's what is happening in a ported or PR system. It is only after the port is resonating that it loads the woofer and impedes it's motion.
Last edited:
I have yet to find the actual LTSpice model, but I found notes from when I ran the model several years ago. I don't know what the the system alignment was.
LTSpice Software: M8N simulated acoustic response to 40 Hz tone. Fourth cycle is at full amplitude.
Vented Speaker Transient Response Simulation.
What does a ported (vented, bass reflex, tuned port) speaker response to a sine wave look like. I found a spice model for a vented speaker and entered the M8N woofer and tuning from WinISD above and computed the response to a 40 Hz sine wave. Notice that it takes several cycles before the acoustic output (driver + port) from the speaker reaches it’s final level at about 80 milliseconds.
LTSpice Software: M8N simulated acoustic response to 40 Hz tone. Fourth cycle is at full amplitude.
Nice!
I did something similar, but with an actual loudspeaker and got very similar results (scroll down to the "hard cut sine burst"):
Please find attached a .zip file containing several LTSpice files for simulating different ported and sealed systems. There is also a .html file from the author that has descriptions of the different files. Looking through the notes on the different schematics you will find Spice directive text that you edit to change the Thiel Small parameters for the drivers, box volumes, port dimensions etc.
I added a file that models the very unique SpeakerLab 30 dual woofer design from parameters and crossover schematics I found for this speaker. It has a dual woofer, active hybrid radiator configuration, similar to a passive radiator, that was invented by Mila (Mioljub) Nestorovic. The speaker has an 8" woofer and a 10" woofer that through an novel crossover is made to operate similar to a passive radiator with customer adjustable parameters. Mila left behind no instructions, that I am aware of, for designing new speakers with his configuration, but with this model you can pretty quickly figure out what woofer parameters and crossover values are required to make it work.
I found these on the web eight years ago, posted October 12, 2005 by Ahmet Feyz Pirimoglu, pirimogl@hotmail.com I am uploading them here, as I find links found on these pages often don't work after a few years. You may be able to find the original site by searching for the authors name, that date and LTSpice yourself.
I have found that running speaker models using the freeware LTSpice is an excellent way to observe the behavior of different speaker designs. You can perform time domain simulations (Transient analysis) and frequency domain (AC analysis) with a few clicks of the mouse on your PC. It is easy to graph any value in the model: voltage, current, cone velocity, acoustic output etc.
I added a file that models the very unique SpeakerLab 30 dual woofer design from parameters and crossover schematics I found for this speaker. It has a dual woofer, active hybrid radiator configuration, similar to a passive radiator, that was invented by Mila (Mioljub) Nestorovic. The speaker has an 8" woofer and a 10" woofer that through an novel crossover is made to operate similar to a passive radiator with customer adjustable parameters. Mila left behind no instructions, that I am aware of, for designing new speakers with his configuration, but with this model you can pretty quickly figure out what woofer parameters and crossover values are required to make it work.
I found these on the web eight years ago, posted October 12, 2005 by Ahmet Feyz Pirimoglu, pirimogl@hotmail.com I am uploading them here, as I find links found on these pages often don't work after a few years. You may be able to find the original site by searching for the authors name, that date and LTSpice yourself.
I have found that running speaker models using the freeware LTSpice is an excellent way to observe the behavior of different speaker designs. You can perform time domain simulations (Transient analysis) and frequency domain (AC analysis) with a few clicks of the mouse on your PC. It is easy to graph any value in the model: voltage, current, cone velocity, acoustic output etc.
Last edited:
Here's a salient provocatively titled but thorough exposition: Why Bassreflex is not Suitable for Low-Frequency Musical Sound Reproduction
I wouldn't call that thorough, I would call that biased. The eigenmode stuff is intentionally made to look terrible and written to confuse.
Extrapolating Blauert's curve is not something you can do.
Regarding the ringing - as I stated earlier, the ringing is at the port frequency and you will not hear it due to fletcher munson. I don't have to extrapolate that 😉
The coupling between the woofer and port essentially happens at a fraction of the speed of sound. I posted a bunch of plots of port, diaphragm and box (sum) step response curves for a bunch of different sealed and ported woofer alignments in another thread.
Here are some of them, just ported box ones, you can see the port isn't taking several cycles: