I don't know if this is real or an urban legend. I am seeing a lot of chatter about how low frequency response of a ported enclosure drops off at low volumes, with people saying the larger the port is, the more low frequency response you lose at low levels.
Is there any truth to this? I see this discussion in car audio forums mainly, and I'm not sure what is going on and whether or not it happens with a properly designed subwoofer in a home environment.
I am assuming a comparison between two subwoofer cabinets that have been modeled and design correctly, with the port length increasing as the port area increases, and the box volume increasing to compensate for the volume the larger port displaces.
So if we are comparing correctly designed subs with the same tuning and the same driver, can we generalize anything about the effect of port area on low frequency output at low listening levels? Ports with more area will produce less noise at high volumes, but do they hurt performance at lower levels?
The specifics of my design is 100L/ 3.5ft^3 box, tuned to 21hz, with either a single 3.875" port, or two 3.875"ports. Driver is RSS265HF-4 10" sub. The max port velocity with the single port is simulated to be 32m/s, so using 2 ports would cut that in half. But is there a down side to doubling the port area, aside from the fact that port length increases and the entire box in turn needs to be bigger?
Any thoughts?
Is there any truth to this? I see this discussion in car audio forums mainly, and I'm not sure what is going on and whether or not it happens with a properly designed subwoofer in a home environment.
I am assuming a comparison between two subwoofer cabinets that have been modeled and design correctly, with the port length increasing as the port area increases, and the box volume increasing to compensate for the volume the larger port displaces.
So if we are comparing correctly designed subs with the same tuning and the same driver, can we generalize anything about the effect of port area on low frequency output at low listening levels? Ports with more area will produce less noise at high volumes, but do they hurt performance at lower levels?
The specifics of my design is 100L/ 3.5ft^3 box, tuned to 21hz, with either a single 3.875" port, or two 3.875"ports. Driver is RSS265HF-4 10" sub. The max port velocity with the single port is simulated to be 32m/s, so using 2 ports would cut that in half. But is there a down side to doubling the port area, aside from the fact that port length increases and the entire box in turn needs to be bigger?
Any thoughts?
If anything you should be losing output at higher port velocities. 16m/s is already on the high side so port compression has to be taken into account. Always flare your ports because that makes a huge difference with port noise.
A bigger port (or more smaller ports) is almost always a plus, but make sure that the port resonances aren't in your bandwith of operation if making a subwoofer as sound deadening will do close to nothing at <200Hz.
A bigger port (or more smaller ports) is almost always a plus, but make sure that the port resonances aren't in your bandwith of operation if making a subwoofer as sound deadening will do close to nothing at <200Hz.
Hi,
Measurements reveal what's happening and what bass-deaf people like to churn into myth and legend and hearsay. Think about it. Ever measure a huge port with a microphone to see it's output curve at lower SPL values? Near field measurements? We are deaf compared to microphones at low frequency.
Your idea of two ports relates back to total surface area and you have to use appropriate end correction factor due to drag and laminar flow based on the geometry of the port type. Splitting it into two separate ports is still just one port with more surface area, so it effects the effective length of the port and thus tuning frequency while lowering total air velocity as you increase the aperture of the port throughout.
Very best,
Measurements reveal what's happening and what bass-deaf people like to churn into myth and legend and hearsay. Think about it. Ever measure a huge port with a microphone to see it's output curve at lower SPL values? Near field measurements? We are deaf compared to microphones at low frequency.
Your idea of two ports relates back to total surface area and you have to use appropriate end correction factor due to drag and laminar flow based on the geometry of the port type. Splitting it into two separate ports is still just one port with more surface area, so it effects the effective length of the port and thus tuning frequency while lowering total air velocity as you increase the aperture of the port throughout.
Very best,
Yeah, there is a lot of myth and legends, but usually there is a grain of truth somewhere. I work as a service tech, and I'm always given grossly inaccurate descriptions of problems, and it is my job to figure out what is actually going on. But in this case I'm a novice at this myself, so I'm not sure how much truth there is here.
I agree that 2 ports of the same length act as a single port with a larger area.
It makes sense to me that as the driver plays at lower and lower levels, at some point the the driver will not impart enough energy to cause the air in the port to resonate, or to control the movement of the air in the port. So it won't create the low frequencies.
The difference of air mass in the port when changing from a single 4x8.5" port to two 4x28.5" ports is a 6.7x increase in air mass. That isn't a trivial increase. That means it takes 6.7x the energy to excite the port.
This seems like basic physics to say there is a point at which the port stops working, but I don't know where this point is. If that point is approaching the inaudible level, then all is well. But if that point is 50-60dB, it'll be a problem.
No and no. You're think about this in the way an engine works.. it's needs a certain amount of RPM before it comes onto the cam. That's not how a port works and is not what is goin on here, in fact the problem is not with the speaker/sub at all, its with human hearing. Lookup the Fletcher-Munson curve, it describes what we percieve as equal loudness at frequencies across the audible spectrum. The bottom line is our hearing is most sensitive in the vocal range(surprise) and least sensitive at the very low and high end, so not only do low freq sounds require more SPL to be heard but the softer the overall level the more differential there needs to be between the vocal range and extreme low freq tones.
I think you misunderstand. I don't work on engines, and I wasn't thinking about engines.
My question is not about perception of loudness. It's about the physical process that happens within the ported speaker cabinet to produce low frequencies.
Maybe it is easier to explain using a passive radiator as an example? A passive radiator resonates in response to the driver, and the passive radiator in turn creates the low frequency extension. If the force pusing against it isn't strong enough, it won't actually move. That force needs to overcome the radiator's mass, air resistance, and friction. And it needs to move forcefully enough that it creates a pressure wave. If the driver doesn't create enough pressure to move the radiator, it won't move and there won't be any low frequency extension.
A port works in exactly the same way, so at some point the port will stop producing low frequency extension.
Vic1184 stated that having a small port will reduce low frequency output at high SPLs. I suspect the opposite is true, but I don't know when it begins to matter. The port I have in mind seems reasonable based on the simulation, but I will have to chop up a lot of wood just to test it. So I think it's worth asking first!
Hi,
It's not a 1:1 relationship. There's more going on. However, you can simply this greatly and completely become worry free when you realize 50~60db is probably lower than the actual noise floor of your home. This is why a calibrated microphone is so important to help get rid of speculation and also solidify what things actually are. It's pretty typical for a quiet home to be in the upper 50's db to lower 60's db. Listening at 75db is not loud but its not quiet either. 85~95db is loud. Over 95db is very loud and will cause damage over time to your ears with prolonged exposure. The port output on a sub at sub frequencies simply will not be even something to consider under 75db or less. Think about it. We don't hear bass the same as mids/treble, we are bass deaf. 75db at 2khz and 20khz is very audible to us, but 75db at 20hz? You'd probably no even notice it's on. You really do not need to worry about having too little of anything with a huge large surface area port, because frankly you'll never have it low enough to experience or even measure this as a problem. I think you're allowing this to become way more of an issue than it is. Frankly, it's more important to make sure your ports are not creating too high of air velocity and becoming an audible source of noise. So with a port, if you can, yes, go big.
Very best,
No it won't, it produces output proportional to the driver output all the way from 0 up to the limits of the driver... asssuming the port is even roughly appropriately sized.
For a given overall volume occupied by an enclosure, a cabinet with large ports and the same Fb will have less output at low level, simply because the net volume of the chamber must be reduced.
No, they do not hurt performance at low levels, though the longer ports have a port resonance that could be heard above the LF output if crossed over with the port resonance in the pass band.
Fb is Fb regardless of the port size, and small signal performance is identical.
If you can afford the extra volume the large ports can afford, and cross over low and steep enough to avoid the port resonance, there are no down sides at low level, and considerable advantage at high power, where the small port will be "blown out", raising Fb, increasing distortion, and reducing port output as it behaves like a hole in the box.
Art
Thanks. That seems to settle it. I've heard this idea stated repeatedly from a multitude of sources, from studio professions to car audio guys. It seems strange that there are so many who believe this is true, and so few saying that it is not actually true.
I understand your point about noise floor. And I agree that what happens below the noise floor doesn't matter.
I'm curious about your statement that the noise floor is 50-60dB? Is that legit? That seems awful loud to me. The noise floor in my house is usually ~35dB, and casual listening level for me is usually 55-60dB. (I have a decent quality SPL meter, so I believe it is pretty acurate.)
35db is an intimate whisper.
60db is low level close proximity conversation.
Regardless, it's always better to measure with the best instrument you can.
Very best,
Welcome to the world of DIY loudspeakers. Myth is rampant. Real engineering knowledge is usually rare, and tends to get lost among all the nonsense posts.
This is an exception - an excellent thread IMO. You got some great responses!
(I learned a few new things about choosing port size, too.)
With the air-conditioner or room heating running? With the kitchen 'fridge running? Traffic noises outside? Jet planes flying overhead? Kids playing? Dogs barking? Cats meowing?
Maybe you have unusually quiet housing. (A peaceful cottage out in the country, far from roads and busy air traffic lanes?)
I live in an apartment with neighbours above and on both sides. There is a fairly busy street not far away. There's almost always some white noise from fans, and deep rumbles and hums from our 'fridge. No way I'm near 35 dB, except maybe late at night when traffic has stopped and the neighbours are all asleep.
-Gnobuddy
Yeah, I live on the edge of a small town. The **** was even running when I got the 35db measurement! The cat is pretty quiet, most of the time. There is a highway not far away, but noise isn't that bad.
Having less noise has it's advantages, but at the same time, a lower noise floor makes speaker/ amp self-noise more noticeable. I had issues with the high sensitivity horn drivers I'm using...it took a 20dB pad to get them under control!
Here's the design I'm working. It has the 2 ports that were discussed above. The box I'll be building right away will be cheap and crude, so not quite like the drawing, but I just need to test dimensions and port configuration. If it works, then I'll make it pretty.
I have the 2 ports exiting a cabity in the back of the cabinet. I needed to mount them vertical due to their length, and exiting out of the top or bottom of the cabinet isn't an option. So yeah, I made a notch in the bottom of the cabinet.
I am thinking this will work. I have read a bunch about flared ports, and the concensus is that a port's effective length is measured from where the flare starts. I'm sure reality is not quite that simple, but it's a starting point for me. All I need from these subs is a little of that 20hz.
I can explain more if you want. I drew this more to represent the finished cabinet as I currently envision it. I'll definitely do a build thread once they're built to share precisely what I'm doing.
I have the 2 ports exiting a cabity in the back of the cabinet. I needed to mount them vertical due to their length, and exiting out of the top or bottom of the cabinet isn't an option. So yeah, I made a notch in the bottom of the cabinet.
I am thinking this will work. I have read a bunch about flared ports, and the concensus is that a port's effective length is measured from where the flare starts. I'm sure reality is not quite that simple, but it's a starting point for me. All I need from these subs is a little of that 20hz.
I can explain more if you want. I drew this more to represent the finished cabinet as I currently envision it. I'll definitely do a build thread once they're built to share precisely what I'm doing.
But now I'm second guessing myself a little as I'm typing this.
The theory I've picked up along the way is that low frequencies from ~20-40hz are absorbed by typical house walls, so they aren't the problem frequencies. The bulk of bass problems are more like ~40hz to 100hz. This seems to prove out in my testing.
I'm just realizing that the direct radiation of the ported rear subwoofer is almost exactly the same frequency range as the front facing sealed woofer. Both have an F3 of ~44hz. And all the DSP tuning is designed to work in the range where the 4 woofers overlap.
But does it matter where the ports radiate from? They are producing frequencies below what the 4 woofers can do, so maybe it doesn't matter that they radiate from a completely different part of the cabinet?
Do you guys understand what I'm getting at? I was assuming all the sub frequencies should radiate from approximately the same place. But there is no front port radiating 20hz frequencies, so I don't need to balance it with the rear port? I guess I could describe the port as the 5th driver, So maybe it doesn't matter where it is located.
So should I try to port out the top? It's a more of a sure bet as far as tuning goes...
The theory I've picked up along the way is that low frequencies from ~20-40hz are absorbed by typical house walls, so they aren't the problem frequencies. The bulk of bass problems are more like ~40hz to 100hz. This seems to prove out in my testing.
I'm just realizing that the direct radiation of the ported rear subwoofer is almost exactly the same frequency range as the front facing sealed woofer. Both have an F3 of ~44hz. And all the DSP tuning is designed to work in the range where the 4 woofers overlap.
But does it matter where the ports radiate from? They are producing frequencies below what the 4 woofers can do, so maybe it doesn't matter that they radiate from a completely different part of the cabinet?
Do you guys understand what I'm getting at? I was assuming all the sub frequencies should radiate from approximately the same place. But there is no front port radiating 20hz frequencies, so I don't need to balance it with the rear port? I guess I could describe the port as the 5th driver, So maybe it doesn't matter where it is located.
So should I try to port out the top? It's a more of a sure bet as far as tuning goes...
Hi,
For location purposes, as long as the time delay difference of sound from the driver vs sound from the port arrive at your ear at a close enough timing, it's fine. You'd have to make an incredibly big and long port for that to even come into play.
Very best,
For location purposes, as long as the time delay difference of sound from the driver vs sound from the port arrive at your ear at a close enough timing, it's fine. You'd have to make an incredibly big and long port for that to even come into play.
Very best,
I don't mean to argue, so here is some data to show why I half agree and half disagree.
These are individual measurements of the 4 woofers in my system. The measurement mic is at the main listening position, and was not touched between measurements. And there has been no smoothing applied!
Keep in mind the 4 woofers are contained within 2 cabinets. The Subs are ported and rear facing (rear ported too), and the Woofers are sealed and front facing. All are playing the same summed signal. The distance between the two woofers is 2-3ft. An the two cabinets are ~6' apart center to center.
The black line is the result of combining these 4 individual drivers. The only thing I did was apply delays to them. I did not apply any EQ to individual drivers to fix peaks and nulls.
I have used some EQ to shape the frequency response, but I used the same filters on both left and right. I used a combination of high shelf + low shelf filters to boost the Subs 20-30hz response while reducing the 60-100hz response. I put a small peak filter on the woofers at 40hz to extend them. I should add a notch filter to the woofers at 80hz, because the Lab 12 woofer response curve peaks at 80hz.
I have found that setting delays and inverting phase + delay can work really well, and REW does a good job of calculating what delays I need. I can ask it to calculate a delay to time align 2 drivers at a specific frequency, or I can just drag the delay slider back and forth until the combined frequency response is a close to being a flat line.
Other people use different methods. I watched a presentation by Geddes about his method of tuning multiple subwoofers, where he explained that he only uses EQ. He said that he used to use delays, but then he decided that he didn't need delays to achieve his desired result.
So yeah, this subject of using multiple subwoofers to control bass response in a small room is not solved, or atleast there is more than one way to do it. I guess that's what got me to take the plunge.
Last edited:
A great example and results in your measurements.
This is not what I was speaking to however.
I was speaking to the question regarding if it matters were the port is located and sound radiating from there, relative to the driver (example does it matter if the driver is front facing and the port is rear facing). Normally this would not matter much. But in my suggestion, incredible length ports, it can. If your sub was long and driver nearer to you and rear ported, and say it's a 36" or 40" long cabinet front to back, the distance of the driver to the listener could be significantly less than the port output location's distance, and so as you transitions from frequencies primarily driven by the driver to the port frequencies, a new delay is noted. Normally this doesn't matter on shorter distances--but as I noted, it can on longer distances. Obviously this is not a common sub configuration, but it was a plausible example to the question.
Very best,
One might add that if the distance (between port and woofer) is much less than a wavelength of the frequencies over which the port is in operation, any delay is entirely irrelevant. It cannot influence the sound.
Ports normally operate only over a small range of frequencies at the bottom end of the speaker's response curve. Let's say the port is tuned at 50 Hz (it will often be lower, unless you're dealing with a very compact speaker).
With sound travelling at about 340 m/S in room-temperature air, the wavelength of a 50 Hz acoustic wave is 6.8 metres (over 22 feet).
If the port is less than, say, one quarter of this distance from the woofer, then any time delay from it becomes irrelevant.
In other words, you only need to worry if your port openings are at least 2 metres / 6 feet away from the woofer. This is incredibly unlikely to happen.
The story is very different at much higher frequencies, where acoustic wavelengths are short, and our ears and brain are able to sense small time differences between multiple otherwise identical time signals (which evolution programmed our brains to recognize as reverberation, echo, sense of ambient space, etc).
For instance, let's assume a crossover frequency of 3 kHz from midrange to tweeter. At 3 kHz, both the midrange and the tweeter are emitting sound.
The wavelength of a 3 kHz acoustic wave is only 11.3 centimetres (about 4 1/2 inches).
One quarter of that is around one inch.
So if the midrange is acoustically set back one inch further than the tweeter, it will have the same impact on the sound as a 6-foot set-back on the woofer/ port tube!
Short version: short time delays between midrange and tweeter do matter.
Short time delays between woofer and bass-reflex port do not matter.
-Gnobuddy