View Full Version : "Bending" a Port

11-03-2016, 11:11 AM
I have an 11" x 4" port in 4 cu ft for my W15GTi. This tunes it to ~24.5Hz. But, the preferred tuning would be ~20 Hz with a port length of ~18".

The cabinet is not deep enough to accomodate this length and I was wondering if there would be any determent, or special "retuning" required if the port has a 90 degree bend in it. After all, we see many ports that are folded like a set of intestines and these seem to work fine. And when was the last time anyone saw a straight tuba?


11-03-2016, 12:26 PM
My L65s had bent ports. All you need to make sure of is that you measure it's length along it's center line.

11-03-2016, 12:40 PM
My L65s had bent ports. All you need to make sure of is that you measure it's length along it's center line.


I was seeing similar on other sites. Actually, I guess that I should only make it 16" total (currently 11" IIRC) so I don't exceed the 4:1 length/diameter "rule of thumb." And it needs a flare to keep the mach # in check.

Ed Zeppeli
11-03-2016, 03:08 PM
A 4" regular throw 90 will add 6" to your overall length.

11-03-2016, 03:45 PM

11-12-2016, 06:07 AM
Do you think the difference of 4.5 Hz would be audible ( I mean feelable ) ?

Regards, -D_E-

11-12-2016, 07:40 AM

My experience says that cabinets and woofers seldom behave in reality as they do in the SW calculations...

I would suggest to try different ports and just measure. I posted a simple way of doing that in the M2 thread and POS posten an even easier in the post after.


Kind regards

11-12-2016, 10:44 AM
Actually, my measurements with the 61-band RTA show that WinISD is very accurate if you get the parameters correct (e.g., "actual" cabinet volume vs volume of the "cavity").

And yes, as with the modified L200 cabinets, with the RTA you can see the "bump" caused by too short a port accentuating a frequency band. I prefer that the cabinets be tuned just a bit low to tighten the bass and remove tubbiness.

BTW, so far the final length ended up at 17.75". Yeah, it's a bit over the 4:1 rule of thumb, but it all fits fine and saves me trying to 1-3/4" off of the exisiting tube, which isn't so easy to do in the cabinet. WinISD shows similar results with a slightly smoother bass with additional length.

I also think that the added flare, sort of counteracts the 4:1 rule of thumb as it reduces the port velocity and chuffing.


11-13-2016, 02:39 AM
Hi again,

Cool setup with 90 degees elbow. Those woofers can really move air so flairing both ends is probably a good investment.

I have found it difficult to predict system fs by SW as the cabinet is rarely 100% resonance free and adding damping in the cabinet will change perceived volume. Geometric layout also effects the outcome. Maybe I'm not using the correct parameters.

Anyway, what is your procedure when using an RTA to measure port fs?

Kind regards

11-13-2016, 03:57 PM
Only flared on the inside. But one side is better than none.

I use the WinISD program combined with my Behringer 61-band RTA with pink noise to look at the response within the room. Sure the room skews things a bit, but the trends are fairly evident.

For example, WinISD projects a hump in the response of my modified L200 cabinets with both ports open (7.25" long), and this hump is evident on the Behringer. When a Zilch plug is put into one port, the hump goes away, but too much of the bottom end is sacrificed and your only real gain is in the deepest bass (which you can now hear because the bass above this range above this is reduced). This is also projected by WinISD and evident on the Behringer.

These ports will be extended to ~9.5" and Win ISD shows this to be about ideal..., and I've no reason to doubt it, especially when the single port used on the L300 was 4.75" long confirming this value.

11-14-2016, 08:13 AM
Hi Toddalin,

It may prove difficult to get the porting right by measuring via a microphone as you will get all sorts of room sounds influencing. -And mic’s & RTA’s are not really very accurate down there.

Maybe I got it all wrong here, but I do think it may be beneficial to get the cabinet/port right first and then try to fix the room.

It is rather easy to measure the port resonance and with your fine equipment I would certainly do an impedance measure just to get a “second opinion” on the system fs. A couple of Hz can make a real difference down there.

I did aquick-and- dirty in WinISD and I would be somewhat cautious to tune it too low. I have two W12GTI in small reflex cases and there the WinISD was off when compared to measurement with DATS.

JBL specifies that 4 cu ft box with a 6” port with 15,13” length would give a system fs of 28Hz, just above driver fs of 25,5. 11” of 4” port sounds like it will come in slightly lower in the same box. The only way to figure out what you prefer is to test. And maybe measure J

A simpleway to find a good place to put the sub is to place the sub where you normally sit and listen and the walk around the room to find the spot with the least room interference and then place your sub there.

Now I will shut up :-)

Kind regards

11-14-2016, 09:06 AM
What does that rule say? How is it derived?


11-14-2016, 10:56 AM
This shows the results of WinISD with the 14.5" port (green tuned to 23.3 Hz) that I just removed and 17.5" bent port (grey tuned to 21.5 Hz) that is in there now. Note that longer port has a no/less hump under ~90Hz, deeper extension, and better phase response and group delay within the audible range.



11-14-2016, 11:48 AM
Hi Toddalin,

I just don't get the same results.
Are your W15GTI Mk I or Mk II?
I'm using WinISD Pro. Do you?
http://www.linearteam.dk/default.aspx?pageid=winisdpro (http://www.linearteam.dk/default.aspx?pageid=winisdpro)

What basreflex alignment did you chose? I used BB4/SBB4 for best transient response.

(A basic description can be found here
http://audiojudgement.com/bass-reflex-alignments-explained/ (http://audiojudgement.com/bass-reflex-alignments-explained/)

MyT&S are as below

Volume 4 cu ft =113,28L
One Port 4” = 10,16cm inner diameter
Length 14,5” = 36,83cm, no flares

I get fs= 21,99Hz like this.

-And I still think you will get a different result if you measure the true impedance. That's what I get after running WinISD.

Now I really will shut up…

Kind regards

11-14-2016, 12:03 PM
Your impedience is wrong. I run in series at 12 ohms.

Also, "4 cu ft" is not "4 cu feet" once you remove the volume of the component displacement and space within the port. I used 100 liters. Your dB also looks off a bit.

I loaded WinISD Pro and it ran but would not stop running (couldn't turn it off other than through the Task Manager) on my system so I unloaded it.

BUT... From when I was playing with it, I think that the Pro version handles vent length differently when you've specified a flare. While you show a length of 38.05 cm, I'm thinking that the "end correction" is taking some of this length away to compensate for the flared section. My "length" is all tube before the flare so the actual length is longer than the 17.5" that I've specified.