Some thoughts on the ported vs sealed debate

[Steve81]

would I set the line at another frequency?

As you see, is there a diference between the two alignments or I’m being fooled by placebo effect?

thank your time so far

You don’t need to worry too much about the line; the waterfall tells the story just fine.

Yes, there is a measurable difference between the alignments, so there is a viable explanation for your observations.

Only way to fully confirm if your impressions are placebo or not, i.e. are these differences truly audible, is to run a blind trial. You’d need to pick a test track where you feel the difference is significant, and replay it a bunch of times, while a partner randomly plugs or unplugs the port, with you being unaware of whether or not the sub is running ported or not. Pick right in a statistically significant manner, and it’s not placebo.

 

[caioferrari]

You don’t need to worry too much about the line; the waterfall tells the story just fine.

Yes, there is a measurable difference between the alignments, so there is a viable explanation for your observations.

Only way to fully confirm if your impressions are placebo or not, i.e. are these differences truly audible, is to run a blind trial. You’d need to pick a test track where you feel the difference is significant, and replay it a bunch of times, while a partner randomly plugs or unplugs the port, with you being unaware of whether or not the sub is running ported or not. Pick right in a statistically significant manner, and it’s not placebo.

The difference you saw is where?
On frequency response or on waterfall?
id like to Learn how to read the graphs to see that too.

I cannot cont on a partner to help me on the test, and switch the two modes is difficult since I have to close the port and adjust the gain.
I consider myself a skeptical and I want to prefer the bass reflex alignment since the box was designed that way. Lol
So, I don’t think I’ll prefer the sealed mode that easy.

 

[Steve81]

The difference you saw is where?
On frequency response or on waterfall?
id like to Learn how to read the graphs to see that too.

Both the FR and the waterfall indicate a difference.

The FR shows the ported alignment is delivering more output in the 20-30Hz range (bottom of the graph tells you frequency, left side tells you output in dB). This results in a peak in response relative to the rest of the range.

For the RT60, just look at the final “slice” to see what’s left after 573 / 581 milliseconds for the vented and sealed alignments respectively. Left side again gives you the dB scale to work with, and bottom is the frequency.

 

[caioferrari]

Both the FR and the waterfall indicate a difference.

The FR shows the ported alignment is delivering more output in the 20-30Hz range (bottom of the graph tells you frequency, left side tells you output in dB). This results in a peak in response relative to the rest of the range.

For the RT60, just look at the final “slice” to see what’s left after 573 / 581 milliseconds for the vented and sealed alignments respectively. Left side again gives you the dB scale to work with, and bottom is the frequency.

The FR is pretty evident to me, but since there is not much information on 20 - 30 Hz range, I don’t think it explains the sonic difference. What do you think?

I saw the RT 60 and the 40-50 Hz on the ported design has longer decay time without having more SPL. Is that correct?
Does it explain the sound difference between the alignments ?

 

[Steve81]

The FR is pretty evident to me, but since there is not much information on 20 - 30 Hz range, I don’t think it explains the sonic difference. What do you think?

I saw the RT 60 and the 40-50 Hz on the ported design has longer decay time without having more SPL. Is that correct?
Does it explain the sound difference between the alignments ?

I believe you are on the right track, and I would consider both the FR and the RT60 differences taken in aggregate to cause your observations.

 

[caioferrari]

I believe you are on the right track, and I would consider both the FR and the RT60 differences taken in aggregate to cause your observations.

Thanks for your help.

I wasn't reading RT60 decay in a right way. Now its clear there is a measurable diference between the two alignment. The sound from the sealed box is more pleasant to hear.

May be I should try a longer port to reduce Fb and get a better response and lower the group delay.

 

[Steve81]

May be I should try a longer port to reduce Fb and get a better response and lower the group delay.

Vented alignments are tricky to get right. @TLS Guy could possibly work something up for you; the only time I’ve bothered designing a vented system was for a set of custom bookshelves I use, and that was nigh on a decade ago. They work pretty well though In the future, it’s going to be a KEF concentric and a matter of finding an appropriate driver to get me an F3 of 40-50Hz in a 0.707 Qtc box with loads out output capability (i.e. a big effing driver or two) in an active setup.

 

[TLS Guy]

Thanks for your help.

I wasn't reading RT60 decay in a right way. Now its clear there is a measurable diference between the two alignment. The sound from the sealed box is more pleasant to hear.

May be I should try a longer port to reduce Fb and get a better response and lower the group delay.

Is this the driver you are still talking about? If so you need reliable models. The Fs of the driver is just below 30 Hz, so you will always end up with an F3 somewhere a little above Fs. Fs pretty much sets the possible bass extension. Your decay observations are related to group delay. It is true that a vented alignment will have greater group delay than the same driver in a sealed alignment. But with proper design, the group delay can be kept within acceptable limits.

I think you said somewhere in this thread that you had modelled the driver, but I have not seen the model, or models here. Do you have them?

 

[caioferrari]

Is this the driver you are still talking about? If so you need reliable models. The Fs of the driver is just below 30 Hz, so you will always end up with an F3 somewhere a little above Fs. Fs pretty much sets the possible bass extension. Your decay observations are related to group delay. It is true that a vented alignment will have greater group delay than the same driver in a sealed alignment. But with proper design, the group delay can be kept within acceptable limits.

I think you said somewhere in this thread that you had modelled the driver, but I have not seen the model, or models here. Do you have them?

yes, this is it.
I don’t remember the exact box properties, but the total internal volume is about 45L and Fb is around 30 Hz. This is the box I block the port to test on sealed mode. I don’t have the graphs, they died with my old PC

I would like to have a proper driver, but in Brazil we just can‘t find speakers for home use. The best we can do is use some car audio speakers for DIY projects.

 

[TLS Guy]

yes, this is it.
I don’t remember the exact box properties, but the total internal volume is about 45L and Fb is around 30 Hz. This is the box I block the port to test on sealed mode. I don’t have the graphs, they died with my old PC

I would like to have a proper driver, but in Brazil we just can‘t find speakers for home use. The best we can do is use some car audio speakers for DIY projects.

You certainly have a good command of the English Language. I will see what can be done to optimize that driver.

 

[TLS Guy]

yes, this is it.
I don’t remember the exact box properties, but the total internal volume is about 45L and Fb is around 30 Hz. This is the box I block the port to test on sealed mode. I don’t have the graphs, they died with my old PC

I would like to have a proper driver, but in Brazil we just can‘t find speakers for home use. The best we can do is use some car audio speakers for DIY projects.

It certainly is a car driver but not hopeless.

It is certainly designed as a car driver. You can see it is designed to be used in very small boxes. It is clearly designed for sealed application.
However, it does have reasonable response in a 1.3 cu.ft box. The vent would need to be a tube with an elbow or a slot vent. The group delay is right at the top end of what is acceptable. That is the best vented design I can do.

For this driver the sealed alignment is the best performer overall, as you would expect from a car driver. The driver will require a 12db/octave boost starting at 50 Hz, and a 12db per octave high pass at 25 Hz to prevent driver damage. That is the best you can do with that driver.

Is it difficult to import drivers from the US to Brazil?

 

[Verdinut]

It certainly is a car driver but not hopeless.

It is certainly designed as a car driver. You can see it is designed to be used in very small boxes. It is clearly designed for sealed application.
However, it does have reasonable response in a 1.3 cu.ft box. The vent would need to be a tube with an elbow or a slot vent. The group delay is right at the top end of what is acceptable. That is the best vented design I can do.

For this driver the sealed alignment is the best performer overall, as you would expect from a car driver. The driver will require a 12db/octave boost starting at 50 Hz, and a 12db per octave high pass at 25 Hz to prevent driver damage. That is the best you can do with that driver.

Is it difficult to import drivers from the US to Brazil?

For the ported design, on their site JBL recommends a net internal volume of 35.4 liters (1.25 cf) with a tuning at 33 Hz. It gets a little awkward to tune that low with a volume that small, one logical solution would be to have a pipe installed outside the box with the use of an elbow.

 

[Steve81]

It certainly is a car driver but not hopeless.

For this driver the sealed alignment is the best performer overall, as you would expect from a car driver. The driver will require a 12db/octave boost starting at 50 Hz, and a 12db per octave high pass at 25 Hz to prevent driver damage. That is the best you can do with that driver.

My 2 cents… given Dr. Mark’s analysis, I would ditch the vent entirely, and get a second woofer. Turn your vented enclosure into a dual opposed sealed design, wired in parallel (assuming your amp can handle this, mum is the word on what you’re using there). Use a healthy amount of bracing material to drop the internal volume to double what the good doctor specified for his single-driver sealed build.

Assuming your amp can double down into a lower impedance load, you’ll gain 6dB of output over the sealed model. This will meet or exceed the vented system above ~35Hz, without falling too far behind in the lowest audible octave.

Now you can start to see why I think sealed has some untapped potential compared to what the market has brought us for the most part.

 

[TLS Guy]

My 2 cents… given Dr. Mark’s analysis, I would ditch the vent entirely, and get a second woofer. Turn your vented enclosure into a dual opposed design, wired in parallel (assuming your amp can handle this, mum is the word on what you’re using there). Use a healthy amount of bracing material to drop the internal volume to double what the good doctor specified for his single-driver sealed build.

Assuming your amp can double down into a lower impedance load, you’ll gain 6dB of output over the sealed model. This will meet or exceed the vented system above ~35Hz, without falling too far behind in the lowest audible octave.

That will take an awful lot of power. The F3 of the sealed design is 53 Hz. Remember every 3db of boost doubles the power. So there will need to be a 10 fold boost in power going from 50 Hz to 30 Hz. However those drivers can't take the power, they are out of xmax at 30 watts with 60 watts of power.

That is the Achilles heel of sealed designs, they take an awful lot of power to reproduce the last octave. That explains why powerful sealed subs are so costly. It takes robust costly drivers with enormous motor systems and very powerful costly amps to drive them. A car is a small space, so you can get away with it. A room is a totally different reality.

So the design is fine at modest volumes, but not for impressive floor shaking.

The trade off here is that the sealed design will be higher quality in the deeper ranges, but the ported design will play much louder, and will actually be twice as loud by perception.

It is just too bad that the OP is limited in his driver selection by geographic location.

The most cost effective subs in db per watt in the last octave are ported, horns or pipes. With sealed you are fighting the laws of physics all the way, and paying through the nose for the privilege!

 

[Steve81]

That will take an awful lot of power. The F3 of the sealed design is 53 Hz. Remember every 3db of boost doubles the power. So there will need to be a 10 fold boost in power going from 50 Hz to 30 Hz. However those drivers can't take the power, they are out of xmax at 30 watts with 60 watts of power.

Power is cheap these days, though I'd prefer not to drop system impedance down to 2 ohms, as that makes things a little costlier. It's only $340 to buy a professional power amp that's rated to deliver 300W x 2 into 4 ohms. $565 gets you 750W x2 into 4 ohms.

Obviously this isn't a driver I would prefer to work with for a sealed design... I'd be looking at no less than a pair of 12" drivers that have a reasonable Vb (balancing size and efficiency) with a Qtc of 0.707 and an F3 of ~40Hz under those circumstances, and enough linear displacement to reach the SPL's I want in the bottom audible octave. Add EQ, DSP limiters, a healthy supply of power, and a bit of end user common sense just for good measure.

So the design is fine at modest volumes, but not for impressive floor shaking.

Imagine 4x Dayton Reference 18” HO drivers optimally placed for improved in-room response (and a bit of additional overall output hopefully), and a pair of those Crown 750W amps. That's not an excessive outlay by high end standards (comparable in cost to a single PB16U), and will provide substantial output capability in a domestic environment. The build is a lot less complex than your TL's, probably requires a good deal less volume in aggregate than all your TLs (74.5L per box according to WinISD), and will sound pretty darned good by any objective qualitative standard. The only real issue is that momentary peaks on the low end will require relatively high momentary peaks of amplifier power. Of course, unless you plan on running 20Hz sine waves at full power through the rig 24/7, the system will still be quite frugal in relation to your electric bill. To wit, with 100W to a single driver, WinISD predicts ~95dB @ 20Hz @ 1m (110dB @ 80Hz!!!). Cone excursion at 100W is ~5mm (xmax is 12.75mm) @ 20Hz, and you're going to have 4 of these (+12dB with perfect summing), and gain from being in an actual room. That's floor shaking potential right there, on 400W split among 4 drivers.

And a few measurements from Josh (CEA tests were taken in his ~4 cubic foot test enclosure.

Just wish we had the full stable of measurements to see what distortion looked like at lower drive levels.

The trade off here is that the sealed design will be higher quality in the deeper ranges, but the ported design will play much louder, and will actually be twice as loud by perception.

Don't forget that the sealed design will play 6dB louder at 50Hz and above, while requiring less power on a dB for dB basis there due to the efficiency gained by adding a second driver. The vented design, in addition to needing more Vb than a DO sealed design, has its advantage over a very narrow bandwidth. Outside of that range around port tune, the DO sealed design is simply better.

If caioferrari's amp CAN double down into 2 ohms (or close enough for government work), it's just the cost of the driver, a little wiring, plus the effort to re-work the cabinet (which would happen either way).

It is just too bad that the OP is limited in his driver selection by geographic location.

Pretty much.

PS: you can see what I did with a couple of 6.5" Seas drivers in vented enclosures here. It's not half bad, even if I do say so myself, considering it's in an average living room that's a part of a massive volume (open to dining room, kitchen, stairwell to upstairs, hallway to other rooms on the main level), with basically zilch for EQ. So I don't hate vented by any means. Given the design constraints I had to deal with (no interest in active back when those were built, limited by available pre-fabricated cabinetry, etc.), they work extremely well for my purposes. 10 years later, with less aversion to an active rig, a better UniQ driver as a basis, and a few more dollars in my bank account, I'm willing to bet I could do even better...That's down the road though.

 

[caioferrari]

It certainly is a car driver but not hopeless.

It is certainly designed as a car driver. You can see it is designed to be used in very small boxes. It is clearly designed for sealed application.
However, it does have reasonable response in a 1.3 cu.ft box. The vent would need to be a tube with an elbow or a slot vent. The group delay is right at the top end of what is acceptable. That is the best vented design I can do.

For this driver the sealed alignment is the best performer overall, as you would expect from a car driver. The driver will require a 12db/octave boost starting at 50 Hz, and a 12db per octave high pass at 25 Hz to prevent driver damage. That is the best you can do with that driver.

Is it difficult to import drivers from the US to Brazil?

I’m at the bus using my smartphone. When I get home, I’ll see the pdfs with attention.
But, the 37 L box you’ve designed is at the top end of group delay, I can assume my design is way above what is acceptable! So, we found the problem, right? My ears are not that crazy.

it’s a little hard to import drivers here. The importing fees are to high, the prices double

 

[caioferrari]

For the ported design, on their site JBL recommends a net internal volume of 35.4 liters (1.25 cf) with a tuning at 33 Hz. It gets a little awkward to tune that low with a volume that small, one logical solution would be to have a pipe installed outside the box with the use of an elbow.

Yes! And if you pay attention on the specs, that tube dimensions don’t hit the given fb. It needs to be longer

 

[caioferrari]

That will take an awful lot of power. The F3 of the sealed design is 53 Hz. Remember every 3db of boost doubles the power. So there will need to be a 10 fold boost in power going from 50 Hz to 30 Hz. However those drivers can't take the power, they are out of xmax at 30 watts with 60 watts of power.

That is the Achilles heel of sealed designs, they take an awful lot of power to reproduce the last octave. That explains why powerful sealed subs are so costly. It takes robust costly drivers with enormous motor systems and very powerful costly amps to drive them. A car is a small space, so you can get away with it. A room is a totally different reality.

So the design is fine at modest volumes, but not for impressive floor shaking.

The trade off here is that the sealed design will be higher quality in the deeper ranges, but the ported design will play much louder, and will actually be twice as loud by perception.

It is just too bad that the OP is limited in his driver selection by geographic location.

The most cost effective subs in db per watt in the last octave are ported, horns or pipes. With sealed you are fighting the laws of physics all the way, and paying through the nose for the privilege!

I have enough power here since I drive my sub with a crown xls 2500 in bridge mode. The power is limited by cone displacement as you said.
Looking for the alignments you’ve suggested, seems that I messed up trying to extended the driver response with a very large enclosure.
But when I listen to it in sealed, I see that a higher frequency cut is a fair price to pay for Quality

 

[Steve81]

I have enough power here since I drive my sub with a crown xls 2500 in bridge mode. The power is limited by cone displacement as you said.
Looking for the alignments you’ve suggested, seems that I messed up trying to extended the driver response with a very large enclosure.
But when I listen to it in sealed, I see that a higher frequency cut is a fair price to pay for Quality

I’ll see what I can come up with for a vented alignment, though it’ll take a little time (traveling today). Dr. Mark also used a pretty large diameter port in his vented spec, and I’m willing to play a little more fast and loose with port velocity A shorter 2.5” flared port could do just fine.

 

[TLS Guy]

I’ll see what I can come up with for a vented alignment, though it’ll take a little time (traveling today). Dr. Mark also used a pretty large diameter port in his vented spec, and I’m willing to play a little more fast and loose with port velocity A shorter 2.5” flared port could do just fine.

Well don't! You absolutely do not want a high port velocity. A turbulent noisy port is a distorted output port. You need to keep vent air velocity no higher than 20m/sec. Flaring the ports is fine, but it makes the port significantly longer and the max air velocity still stands.