This is a theoretical question.
I did some more vented box simulations in WinISD Pro. It seems that when you add a rumble filter to a closed system, you increase the group delay almost as much as making the same box a vented one without the filter.
So, if a well designed vented system provides much more SPL at the lower frequencies, AND group delay effects are present anyway with filters, then why is closed box preferred? Surely if the ports are flared enough you can have a nice sounding sub without paying excessive amounts for better drivers?
Secondly, if filters can affect group delay, would it be possible to design a filter to offset the group delay of a vented system so then it sounds like a closed system? The point is, if this is possible, you would end up with more response below 20Hz without the need for more amp power or a better driver, just a smarter box.
I did some more vented box simulations in WinISD Pro. It seems that when you add a rumble filter to a closed system, you increase the group delay almost as much as making the same box a vented one without the filter.
So, if a well designed vented system provides much more SPL at the lower frequencies, AND group delay effects are present anyway with filters, then why is closed box preferred? Surely if the ports are flared enough you can have a nice sounding sub without paying excessive amounts for better drivers?
Secondly, if filters can affect group delay, would it be possible to design a filter to offset the group delay of a vented system so then it sounds like a closed system? The point is, if this is possible, you would end up with more response below 20Hz without the need for more amp power or a better driver, just a smarter box.
You've got the right ideas but the application is slightly wrong. The rumble filter would normally be used on the vented box not the sealed box, because below the tuning frequency the vent unloads the cone like it's just waffling in free air, whereas a sealed box keeps it under control all the way down to DC.
This unloading problem is also the reason why you can't really equalise (amplitude or group delay) or do much to a vented box below Fb.
This unloading problem is also the reason why you can't really equalise (amplitude or group delay) or do much to a vented box below Fb.
Just checked out Paul Spencers wiki on group delay:
http://www.diyaudio.com/wiki/index.php?page=Group+Delay
Sure you can equalise group delay with a digital filter... You just delay the signal entering the amp for different frequencies to offset excessive delay caused by a port.
So, perhaps the top end speaker systems use vented boxes with digital delay filters to match the delay to the main speaker system at the crossover point!
This could be an interesting solution for diy.
How feasible would it be to implement a variable digital delay filter into a subwoofer plate amp, as an addon diy kit?
http://www.diyaudio.com/wiki/index.php?page=Group+Delay
Sure you can equalise group delay with a digital filter... You just delay the signal entering the amp for different frequencies to offset excessive delay caused by a port.
So, perhaps the top end speaker systems use vented boxes with digital delay filters to match the delay to the main speaker system at the crossover point!
This could be an interesting solution for diy.
How feasible would it be to implement a variable digital delay filter into a subwoofer plate amp, as an addon diy kit?
Dunno about digital filters. but I believe an "all-pass" filter changes phase while leaving amplitude alone, which I think could be used to get rid of group delay. And you wouldn't have to digitize the signal.
Better ask the solid state guys about "all-pass" filters if you're interested, though. I really don't know that much about them.
Better ask the solid state guys about "all-pass" filters if you're interested, though. I really don't know that much about them.
There are several factors that differentiate a sealed sub's sound from a ported sub's sound, one of which is group delay.
Group delay is related to frequency response. Hard knee, high GD, soft knee, low GD.
If you place a HP filter in line to protect a sealed sub from itself, you increase GD, thus defeating the purpose for wanting to build a sealed sub vs a ported sub, regarding only GD.
Still, as has been pointed out in an earlier post, it is the ported sub which requires a HP filter to protect the driver, not (well, it shouldn't be necessary, in a proper design) the sealed sub. The HP filter increases GD in the ported sub, widening the gap between the 2 alignments.
Design your sealed sub with these general guidelines in mind:
1. Design the box size to allow for excursion headroom. A rule of thumb is to shrink the box to constrict the excursion to 60% of available X-Max.
2. Use EQ (L/T) to extend the anechoic response, using enough boost to be able to also design the system 'Q', which will control GD. The boost, or amount of extension factored into a Linkwitz/Transform is directly related to how much you can affect the resulting 'Q' of the sealed sub.
If you shrink the box to acheive a mechanical protection for your driver(s), you will usually end up with a system 'Q' that is not optimal, or higher than .707. You then use the L/T to dial in your desired 'Q'. Lower is better, IMO. I prefer a 'Q' of .5, so-called critically damped, but there is at least one study that concluded that the majority preferred a 'Q' of .65.
3. If you have a ported design configured with a given driver, a general rule is that you keep the same size box, eliminate the port(s), use 2 of the given driver and double the amplifier power. This will net you output that is within 1-2 dB of the ported sub at tune, give 4-6 dB headroom above tune and destroy the HP filtered ported sub's output below tune, when using an L/T EQ circuit.
4. Use an amplifier that has it's DC protection circuit as low as possible. Some amps have a relatively high corner on their HP (20 Hz.), and they vary wildly. The Adire ADA plate amps (for example) have a 2 Hz. protection circuit. This will keep the DC protection from negatively influencing your system GD, after all your hard work to keep GD low.
There are not many sealed systems that take this sort of care to keep GD down, and therefore it follows that there are not many (if any at all) examples of a sealed subwoofer system that is available to directly compare to a ported sub that utilizes a HP filter for protection, which is unwittingly cascaded with the amplifier's HP filter.
It is audible.
Bosso
Group delay is related to frequency response. Hard knee, high GD, soft knee, low GD.
If you place a HP filter in line to protect a sealed sub from itself, you increase GD, thus defeating the purpose for wanting to build a sealed sub vs a ported sub, regarding only GD.
Still, as has been pointed out in an earlier post, it is the ported sub which requires a HP filter to protect the driver, not (well, it shouldn't be necessary, in a proper design) the sealed sub. The HP filter increases GD in the ported sub, widening the gap between the 2 alignments.
Design your sealed sub with these general guidelines in mind:
1. Design the box size to allow for excursion headroom. A rule of thumb is to shrink the box to constrict the excursion to 60% of available X-Max.
2. Use EQ (L/T) to extend the anechoic response, using enough boost to be able to also design the system 'Q', which will control GD. The boost, or amount of extension factored into a Linkwitz/Transform is directly related to how much you can affect the resulting 'Q' of the sealed sub.
If you shrink the box to acheive a mechanical protection for your driver(s), you will usually end up with a system 'Q' that is not optimal, or higher than .707. You then use the L/T to dial in your desired 'Q'. Lower is better, IMO. I prefer a 'Q' of .5, so-called critically damped, but there is at least one study that concluded that the majority preferred a 'Q' of .65.
3. If you have a ported design configured with a given driver, a general rule is that you keep the same size box, eliminate the port(s), use 2 of the given driver and double the amplifier power. This will net you output that is within 1-2 dB of the ported sub at tune, give 4-6 dB headroom above tune and destroy the HP filtered ported sub's output below tune, when using an L/T EQ circuit.
4. Use an amplifier that has it's DC protection circuit as low as possible. Some amps have a relatively high corner on their HP (20 Hz.), and they vary wildly. The Adire ADA plate amps (for example) have a 2 Hz. protection circuit. This will keep the DC protection from negatively influencing your system GD, after all your hard work to keep GD low.
There are not many sealed systems that take this sort of care to keep GD down, and therefore it follows that there are not many (if any at all) examples of a sealed subwoofer system that is available to directly compare to a ported sub that utilizes a HP filter for protection, which is unwittingly cascaded with the amplifier's HP filter.
It is audible.
Bosso
"an "all-pass" filter changes phase while leaving amplitude alone, which I think could be used to get rid of group delay. "
Wouldn't you need a filter that gives group advance to compensate for group delay (seriously)?
Also, while a VB needs it a lot more, a sealed box can also benefit from a HP filter.
What's the point in pushing the driver further into nonlinear operation and thermal compression by feeding it low freq that it can't produce at a useful audible level?
Wouldn't you need a filter that gives group advance to compensate for group delay (seriously)?
Also, while a VB needs it a lot more, a sealed box can also benefit from a HP filter.
What's the point in pushing the driver further into nonlinear operation and thermal compression by feeding it low freq that it can't produce at a useful audible level?
Thanks guys for all your input. I'm beginning to steer towards the sealed box arrangement for my XLS at the moment because:
(a) Flatter delay curve
(b) easier to build
(c) sounds better for music?
(d) smaller box is more portable
(e) can't hear 20Hz anyway.
The amp and driver have arrived and I have some 18mm mdf ready to go!
Building a 30L Q=0.515 box. If I can be bothered after that I'll find a pcb for a linkwitz transform to boost it a little. But the cone excursion in WinISD goes crazy with the LT in place, so I'll see how it goes.
(a) Flatter delay curve
(b) easier to build
(c) sounds better for music?
(d) smaller box is more portable
(e) can't hear 20Hz anyway.
The amp and driver have arrived and I have some 18mm mdf ready to go!
Building a 30L Q=0.515 box. If I can be bothered after that I'll find a pcb for a linkwitz transform to boost it a little. But the cone excursion in WinISD goes crazy with the LT in place, so I'll see how it goes.
Wow, I did the measurements and realized the box i'm building is just over a foot cube! I'm doing a math degree so I should know that 3 cubed is 27, but I just had a feeling that the box would be more rectangular! Anyway... should sound good tomorrow.
Is it better to acoustically isolate the amp from the actual box, ie make the box slightly longer and put a wooden partition in between with a hole for the speaker wires? Or doesn't make any difference? My plate amp has a cover on the inside so there are no exposed parts and has a rubber o-ring to seal it, so leakage shouldn't be an issue.
PS. Let's just say I live close to the city, it's handy to ride my bike to uni.
Is it better to acoustically isolate the amp from the actual box, ie make the box slightly longer and put a wooden partition in between with a hole for the speaker wires? Or doesn't make any difference? My plate amp has a cover on the inside so there are no exposed parts and has a rubber o-ring to seal it, so leakage shouldn't be an issue.
PS. Let's just say I live close to the city, it's handy to ride my bike to uni.
I tend to disagree with some of the posts here so I'll put forward my view to balance, then you make the choice. The posts are valid but here's an alternative view ...
Actually I think a sealed sub should have a rumble filter also for home theatre use, for reasons stated in wiki pages. My own simulations and experience with my current sealed subs confirms this. I just think its a waste to have to allow so much more headroom.
I'd expect that it would be better due to lower distortion above tuning. However, I think better performance could be had with two vented subwoofers. For a given output, cone excursion will be lower, and excursion related distortion is much lower. This includes distortion related to BL and suspension linearity. You could argue that the vented subwoofer only has an advantage near tuning, in the bottom octave, since the vented and the sealed box achieve their output above say 40 Hz from the driver output and not from vent reinforcement. However, distortion products created by the bottom octave will create harmonics above the passband of the subwoofer where the ear is more sensitive.
There was a comparison of a number of commercial subwoofers, including a Velodyne 18" serfvo subwoofer, an Adire Sadhara, Servodrive Contrabass and a number of others. The sub which was rated as the best was a Genelec which had four 12" drivers in vented configuration. A number of the subwoofers were sealed and I don't recall any of them being rated as superior in any area of performance.
I recall a discussion on Sound Illusions forum in which Dan Wiggins mentioned his preference for vented subs.
I think vented subs get bad press and I suspect it often has a lot to do with integration in many cases. I'd like to see if people can really tell the difference between a vented and sealed sub where the inroom response is the same. I have heard of a study in which it was concluded that below 50 Hz the difference is not audible. Who has actually done a test with the same driver in both configurations where the response is calibrated the same?
I've heard a vented tumult with very impressive results, but I did note that it had a sohpisticated calibration system consisting of Behringer Ultradrive and Ultracurve. It had digital phase correction, and was all calibrated flat in-room, and was linked with some other subs. Very deep and very tight. It removed any doubt in my mind as to how good vented subs can sound, certainly better than mine currently in sealed boxes!
If a sealed LT sub has the same response as a vented sub they will have the same GD. It's the steeper rolloff of a vented sub that gives it higher GD. But is it an audible difference? If you say "yes" then please show some kind of evidence, as I'd like to see it.
Actually I think a sealed sub should have a rumble filter also for home theatre use, for reasons stated in wiki pages. My own simulations and experience with my current sealed subs confirms this. I just think its a waste to have to allow so much more headroom.
I'd expect that it would be better due to lower distortion above tuning. However, I think better performance could be had with two vented subwoofers. For a given output, cone excursion will be lower, and excursion related distortion is much lower. This includes distortion related to BL and suspension linearity. You could argue that the vented subwoofer only has an advantage near tuning, in the bottom octave, since the vented and the sealed box achieve their output above say 40 Hz from the driver output and not from vent reinforcement. However, distortion products created by the bottom octave will create harmonics above the passband of the subwoofer where the ear is more sensitive.
There was a comparison of a number of commercial subwoofers, including a Velodyne 18" serfvo subwoofer, an Adire Sadhara, Servodrive Contrabass and a number of others. The sub which was rated as the best was a Genelec which had four 12" drivers in vented configuration. A number of the subwoofers were sealed and I don't recall any of them being rated as superior in any area of performance.
I recall a discussion on Sound Illusions forum in which Dan Wiggins mentioned his preference for vented subs.
I think vented subs get bad press and I suspect it often has a lot to do with integration in many cases. I'd like to see if people can really tell the difference between a vented and sealed sub where the inroom response is the same. I have heard of a study in which it was concluded that below 50 Hz the difference is not audible. Who has actually done a test with the same driver in both configurations where the response is calibrated the same?
I've heard a vented tumult with very impressive results, but I did note that it had a sohpisticated calibration system consisting of Behringer Ultradrive and Ultracurve. It had digital phase correction, and was all calibrated flat in-room, and was linked with some other subs. Very deep and very tight. It removed any doubt in my mind as to how good vented subs can sound, certainly better than mine currently in sealed boxes!
If a sealed LT sub has the same response as a vented sub they will have the same GD. It's the steeper rolloff of a vented sub that gives it higher GD. But is it an audible difference? If you say "yes" then please show some kind of evidence, as I'd like to see it.
Well, I built the 30L sealed XLS, 200W with HP filter at 17Hz... and it sounds AWESOME!!
It can shake windows with the accurate deep bass present in dvd movies! Also, I am very happy with the musical sounds it can produce, such as low pipe organ music!! Acoustic bass guitar sounds full with my midi setup.. What a difference it makes over crappy computer speakers!
I'll post some pics when i find the digicam.
Time to dance salsa with this sub@!
PS. I agree Paul, the pipe organ pushes the limit of the excursion even with the rumble filter so I'm glad the amp has one. Even if that means less SPL I'd rather not have to worry about the cone tearing itself apart.
It can shake windows with the accurate deep bass present in dvd movies! Also, I am very happy with the musical sounds it can produce, such as low pipe organ music!! Acoustic bass guitar sounds full with my midi setup.. What a difference it makes over crappy computer speakers!
I'll post some pics when i find the digicam.
Time to dance salsa with this sub@!
PS. I agree Paul, the pipe organ pushes the limit of the excursion even with the rumble filter so I'm glad the amp has one. Even if that means less SPL I'd rather not have to worry about the cone tearing itself apart.
paulspencer said:
Paul,
If a sealed L/T sub has the same response as a ported sub, you designed and built a poor version of a sealed L/T subwoofer.
Siegfried Linkwitz himself has said that placing a HP filter in the signal chain defeats the purpose of a sealed L/T sub. I agree.
A sealed sub should not only be designed to retain it's 2nd order roll off advantage, the L/T should be used to control the width of it's F6 bandwidth ('Q').
It's not just the GD that's affected by the shape of the knee, but also stored energy release and ring time. The ported sub also allows for the noise occuring inside the box to be much more audible because, well...it has a big hole in it. IOW, there is more to the audible differences than GD.
Comparing a 1X15" sealed sub to the gigantic 4X12" ported Genelec is a bit of a stretch, still it beat the Genelec at 15 Hz and below, and two of the Kleiss 1X15" sealed subs would be a good match above 20 Hz and leave the Genelec in the dust for the bottom 2 octaves (5-10, 10-20 Hz), cost 1/2 the price and use up a little over 1/2 the real estate.
I use a 4X15" sealed L/T system powered by 5,600 watts, 2 cabinets of 2X15" each, combined, taking up less than 10 cubes of space. Believe me they would hang with the Genelec.
Now, as to the differences in sound quality, I have compared quite a few ported and sealed subs and I'll say that there is a difference in sound between the 2 alignments. I'm not going to say that my studies are valid for anyone but me, and I won't say that you'll find many commercial sealed EQ'd subs with multiple drivers to compare, as I said in the earlier post.
The Gotham looks interesting. Four ACI Maestros would be interesting. But, this thread isn't about finding a commercial sealed L/T sub that has low GD, it's about building one, DIY...and it's very doable.
The Genelec has very high THD when pushed. SVSubs have very low THD numbers, as a ported sub should...that's it's single advantage, as you point out...at it's tune point. Below it's tune point, THD and excursion skyrocket to the point of it's offering nothing below tune.
To combat this, the ported sub should be tuned low. That means a bigger box. It will be more efficient and have an overall lower distortions number at tune, or between impedance peaks. It will have not much of anything below tune. It will have multiples higher GD. It will have higher stored energy release. It will have higher ring time. It will have a hard knee corner. These differences will be audible.
I'm surprised you haven't done the comparo yourself.
Bosso
Bosso, I see a few should's in your post, but in order for a should to have some meaning, it needs qualification.
ie. in order to achieve XYZ, one should do ABC
Hence I don't necessarily agree or disagree on those statements, except to point out that it depends on what you are trying to achieve. You may in fact want to achieve max SPL and extension out of a sealed LT sub, in which case this means the same response as a vented sub. Not that I see myself ever designing a sub that way, but for some with certain goals and restraints, this might be appropriate.
I'll agree that the Genelec had an unfair advantage over most of the subs compared, however the Contrabass has two 15" drivers with passive radiators as well. IIRC there was a sealed 21" driver (Bagend) and that represents to me a sealed driver with eq pushed to an extreme, with very limited output (considering VD), and high distortion. My point here, was that here is an example where a vented subwoofer wins over a number of sealed subs, not only due to high SPL but where all aspects of performance are considered.
I've have comparisons to draw from, including my own subwoofers, using both sealed and vented for the same driver. I have the means to do a blind AB test with a more scientific approach than I've done so far, but there are a lot of other things on my to do list! I'd like to organise some kind of diy event this summer, so I may do something like that and post the results, see what a group of diyers think when they don't know what they are hearing ...
ie. in order to achieve XYZ, one should do ABC
Hence I don't necessarily agree or disagree on those statements, except to point out that it depends on what you are trying to achieve. You may in fact want to achieve max SPL and extension out of a sealed LT sub, in which case this means the same response as a vented sub. Not that I see myself ever designing a sub that way, but for some with certain goals and restraints, this might be appropriate.
I'll agree that the Genelec had an unfair advantage over most of the subs compared, however the Contrabass has two 15" drivers with passive radiators as well. IIRC there was a sealed 21" driver (Bagend) and that represents to me a sealed driver with eq pushed to an extreme, with very limited output (considering VD), and high distortion. My point here, was that here is an example where a vented subwoofer wins over a number of sealed subs, not only due to high SPL but where all aspects of performance are considered.
I've have comparisons to draw from, including my own subwoofers, using both sealed and vented for the same driver. I have the means to do a blind AB test with a more scientific approach than I've done so far, but there are a lot of other things on my to do list! I'd like to organise some kind of diy event this summer, so I may do something like that and post the results, see what a group of diyers think when they don't know what they are hearing ...
Good morning gentlemen,
Ported versus sealed is a classic case of tradeoff in subwoofer design. Ported subwoofers have some advantages in that there tends to be less dynamic distortion in the low bass since the porting assists in boosting output. Also, in a ported system, driver excursion demands in the low bass are reduced, which means less IM distortion. Ported subwoofers also have some disadvantages in that there is much less controlled behavior of the driver below tuning. Also, enclosure size tends to be larger in order to achieve a given leven of extension, and there is a chance of port turbulence-related noise at very high output levels.
A ported subwoofer that is tuned very low tends to act much like a sealed box subwoofer in the range where bass is transient. Due to larger enclosure size and less driver excursion demands, there is also less need to use ultra high excursion drivers and very high powered amplifiers. This usually means that there is some cost savings to the consumer.
DIYers should choose their design based on that which compromises in the least number of areas that they consider important, including size, cost, features, flexibility, extension, output, etc.
Sincerely,
Ported versus sealed is a classic case of tradeoff in subwoofer design. Ported subwoofers have some advantages in that there tends to be less dynamic distortion in the low bass since the porting assists in boosting output. Also, in a ported system, driver excursion demands in the low bass are reduced, which means less IM distortion. Ported subwoofers also have some disadvantages in that there is much less controlled behavior of the driver below tuning. Also, enclosure size tends to be larger in order to achieve a given leven of extension, and there is a chance of port turbulence-related noise at very high output levels.
A ported subwoofer that is tuned very low tends to act much like a sealed box subwoofer in the range where bass is transient. Due to larger enclosure size and less driver excursion demands, there is also less need to use ultra high excursion drivers and very high powered amplifiers. This usually means that there is some cost savings to the consumer.
DIYers should choose their design based on that which compromises in the least number of areas that they consider important, including size, cost, features, flexibility, extension, output, etc.
Sincerely,
Peter Marcks said:
Peter,
Great to have you here! I've followed your icareer from it's onset with great admiration, and your posts always add to the conversation.
I singled out this sentence because I believe it's the single reason that here are many ported subs to choose from and few all out sealed design choices, and of those that are available commercially, you'll shell out a good bit more $$.
Just as Dr. Hsu has chosen the ported alignment and taken it to the Nth degree, some DIYers have chosen the sealed, EQ'd sub and explored the possibilities.
The OP has chosen a ported alignment and is apparently very happy, and the original question regarded GD differences.
...Any thoughts?
Bosso
Ron E said:
I for one would love to see your proof of such a statement. I suppose it would depend greatly on how one defines "extension".
Extension in a ported subwoofer is a function of internal volume and port length. There is only so much port length that can fit inside any given box. Port length requirements can increase dramatically as one lowers the tune, even to the point where the port will no longer fit inside the box. Ported subwoofers also tend to roll off very quickly below the tuning point, so that even in-room extension falls off relatively sharply below tuning point.
The most obvious example where a sealed subwoofer would typically have an advantage in extension versus a similarly sized ported subwoofer would be when looking at relatively small box sizes. With, say, a subwoofer enclosure that is 12" cubed, very low extension is not a very realistic goal for one who wants to design a ported subwoofer. It would be much more realistic to try to achieve that in-room extension in a sealed box subwoofer or a subwoofer with PR(s).
Cheers
bossobass said:
Thank you for the warm welcome, Bosso! I enjoy the perspective that you are bringing to various forums on sealed versus ported subwoofer design!
Great question about GD differences. Very difficult to conclusively answer. There is always debate about audibility at very low frequencies, and it is very difficult to isolate just that one variable without considering all the other variables in play when comparing different alignments. There are definitely some seriously good arguments to be made in favor of a sealed design, and same with a ported design.
Take care!
Sincerely,
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.