[platoon2063]

I constructed a new enclosure for my now dual opposed ported pa460. Added it to my ported UXL and two ported Dayton 460HO's. My room is 20 x 21 square-ish so, can be challenging. After taking measurements for several hours today this is the result when I called it quits for the day.

pa460 HPF @ 35hz, other subs @ 20hz, no boosts, two cuts centered around 58hz and 83hz

Time-aligned the best I could. Right side of the room just would not align with the left side.

Pretty big drop off from 30hz to 20hz. Could not get the large peak area centered around 82 ish to diminish much; sounds a little boomy probably because of that.

Waterfall looks terrible to me. I don't understand how to decipher it or what to do to clean it up.

Where do I go from here?

 

[lknhomeaudio]

So, three separate ported sub systems. Are they even tuned to remotely the same frequency? I'm betting not!

 

[smcmillan2]

I constructed a new enclosure for my now dual opposed ported pa460. Added it to my ported UXL and two ported Dayton 460HO's.

More information on each please. Tunes, measurements, location, etc.

Waterfall looks terrible to me. I don't understand how to decipher it or what to do to clean it up.

Looks like a lot of ringing to me, but I'm no expert. What type of room are you in? Sealed off from others?

 

[platoon2063]

So, three separate ported sub systems. Are they even tuned to remotely the same frequency? I'm betting not!

Yes, the UXL and two Dayton 460HO's are all in the exact same size (approx. 8 cuft) enclosure all tuned to 18hz with a HPF in the Inuke amps set @ 20hz. The pa460 is tuned to 30hz with a hpf in the Inuke set @ 35hz.

W/O the pa460 I had very little midbass, thus the reasoning for adding.

 

[platoon2063]

More information on each please. Tunes, measurements, location, etc.



Looks like a lot of ringing to me, but I'm no expert. What type of room are you in? Sealed off from others?

I don't quite have an understanding of ringing. Room is carpeted. Measurements were taken in an absolute quiet environment.

The room can be sealed; just one door opens to the rest of the home. I have taken measurements with the door closed and opened. Measurements look the same with the exemption of approx.

 

[Wayne A. Pflughaupt]

Ringing simply means that it takes a long time for the signal to fade down to the noise floor.

The reason your waterfall looks so bad is that it was taken at an extremely high SPL. Obviously it takes a loud signal longer to fade away than a signal at a lower level. More about waterfalls here.

Thirty Hertz seems like poor extension for subs HP’d at 20 Hz. I’d suggest taking individual measurements of the best sub at all viable locations.

Regards,
Wayne A. Pflughaupt

 

[lknhomeaudio]

Yes, the UXL and two Dayton 460HO's are all in the exact same size (approx. 8 cuft) enclosure all tuned to 18hz with a HPF in the Inuke amps set @ 20hz. The pa460 is tuned to 30hz with a hpf in the Inuke set @ 35hz.

W/O the pa460 I had very little midbass, thus the reasoning for adding.

Okay, cool. Can you post a graph without the PA460 hooked up?

And so we're clear, you're concerned with how your response rolls off at 30hz... correct?

 

[platoon2063]

Okay, cool. Can you post a graph without the PA460 hooked up?

And so we're clear, you're concerned with how your response rolls off at 30hz... correct?

First graph is each sub solo and then combined -green. Second graph is original 3 combo-green, pa460 substituted in for a 460HO-red, then all 4 subs-black.

*15db drop from 30hz to 20hz confuses me (4 18" ported subs should be strong to at least 20hz)?

*large peak in the 60's might come down with peq cuts, correct?

*crazy dip @ 82hz - ish (not sure how to deal with that)

 

[lknhomeaudio]

Outside of the issue of your subs not playing more "flatish" down to 20hz, or so, it appears you have some distance/time/alignment/phase issues going on. Even in the first graph, before the PA460 is added in, the combined response is falling below the individual responses and that should never happen. Either something needs it's polarity reversed, or moved further away, or something!

I guess the first thing to check is make sure they are all moving in and out at the same time. I've found the simplest way to do this is use REW signal generator and play a 2-3hz tone and watch them. Start off with the volume low, then just turn it up enough to where you can see the movement pretty good.

 

[Jedi940]

Are all the subs up front are are some positioned nearfield?

Did you follow mtg90's multiple sub integration guide? If not, what was your process for adding these subs.

 

[LastButNotLeast]

Where do I go from here?

You go to the REW thread
https://www.avsforum.com/forum/91-a...surement-techniques-how-interpret-graphs.html
and you read Jerry's guide.
Then you take measurements with the correct settings and post them correctly to that thread and you will get tons of useful information.
Michael

 

[platoon2063]

Are all the subs up front are are some positioned nearfield?

Did you follow mtg90's multiple sub integration guide? If not, what was your process for adding these subs.

Per AustinJerry, all subs should line up in a line on the impulse measurement to show they are time aligned. I have done this per the first graph. BTW, not sure why the impulse graph looks like it was written in crayon. I also level matched them @ 82db using the REW generator pink noise.

Attached is a pic of the current layout in my room.

 

[Jedi940]

Per AustinJerry, all subs should line up in a line on the impulse measurement to show they are time aligned. I have done this per the first graph. BTW, not sure why the impulse graph looks like it was written in crayon. I also level matched them @ 82db using the REW generator pink noise.

Attached is a pic of the current layout in my room.

Well, since none of them are nearfield, you shouldn't need to filp the polarity of any of them although if you have tried everything else, that may be a last resort.

As for the impulse graph, I don't see how that is showing them time-aligned. If anything, it is showing that they are way off from my understanding. You have way too much activity happening before time zero and there are 3 distinct peaks after time zero that are no where close to each other. One is at 4ms, one is at 11ms and one is at 15ms. I'm not sure why the orange line is showing the largest peak before time 0. That seems odd to me.

0 along the x axis is when the sound occurs. Anything to the right of that represents delay. The highest peaks of each sweep do not line up at all which will result in cancellation and a poor response.

EDIT: I'd be curious to see all the settings you used for that impulse graph because it doesn't look like what I'm used to seeing so there is a chance I could be interpreting it incorrectly.

 

[platoon2063]

Well, since none of them are nearfield, you shouldn't need to filp the polarity of any of them although if you have tried everything else, that may be a last resort.

As for the impulse graph, I don't see how that is showing them time-aligned. If anything, it is showing that they are way off from my understanding. You have way too much activity happening before time zero and there are 3 distinct peaks after time zero that are no where close to each other. One is at 4ms, one is at 11ms and one is at 15ms. I'm not sure why the orange line is showing the largest peak before time 0. That seems odd to me.

0 along the x axis is when the sound occurs. Anything to the right of that represents delay. The highest peaks of each sweep do not line up at all which will result in cancellation and a poor response.

EDIT: I'd be curious to see all the settings you used for that impulse graph because it doesn't look like what I'm used to seeing so there is a chance I could be interpreting it incorrectly.

Impulse axis limits are as followed.

left-0.051, right 0.140, top 110, bottom -110. set for % instead of dbfs

 

[BassThatHz]

Step 1 should be measure each sub nearfield with no EQ or XO's.
Then add the XO's and re-measure nearfield.

Then apply the necessary individual delays for each spot (or pairs if symmetrical),
if you don't have enough discrete DSP outs then that could be part of your problem.

This is assuming you've already found the X best spots for each sub. (Which must be measured individually in isolation.)

Then you level-match them all, if you want to (which you probably will...)
1:1 octave smoothing helps with that.

Only after all of that is DONE. Then apply polarity (so that they sum in pairs or groups to a louder SPL).

Once they are summing properly, THEN apply EQ for each sub in each spot.

Generally you have two choices here: optimize the golden-seat via delay/EQ or average the EQ adjustments with multi-point mic'ing across all the seats. It doesn't make much sense to apply delay in a multi-mic average since every spot will require different values (which is impossible).
So basically: selfishly-optimized to perfection for 1 seat OR all the seats being ho-hum porridge, nothing spectacular but nothing horrible either (hopefully...)

That's all that can be done.

There are 3 types of room-nulls:
-Seat to wall.
-Sub to seat
-Sub to wall to seat
All rooms have this, the remaining question is: at what frequencies and how badly.

To solve the 15-30hz problem, the only other thing you can try, is different slopes of the PA's HPF. Instead of say 12db/oct try 18 or 24. That should shift the phase near those frequencies and "hopefully" bring all the subs into coherence and thus avoid the SPL drop.
Phase issues are huge problem when you start mixing different ported/horned boxes. Sealed and IB isn't nearly as bad in this regard, it is more forgiving of different driver-types and box sizes. Especially in the mid-bass (like mains to MBM and MBM to LFE overlaps etc.)

 

[BassThatHz]

There is no completely-free lunch.

One of several reasons why I sealed my LCR's and added more/beefier sealed woofers in them, and more sealed MBM's, and more sealed dedicated-LFE subs.
It is easy to build, compact, deep extension, chuffing-free and phase tolerant.
It's just not very efficient or cheap, you need lots of drivers and power to make it work.

Unless you are powering a football stadium (which you aren't ) then as the quantity of woofers climbs, ported-only rigs make less-and-less sense. By the time you hit 32 18's it's gonna be hella-loud no matter WHAT you do. The big difference is that with sealed and IB you'd be GETTING the 0-10hz extension, as where ported and horned would STILL be rolling off like a rock at 15-20hz. (Which will sound WRONG\OFF when playing stuff like BassILoveU or F-Irene which goes to down (flat) to 5-7hz. i.e. Where is the beef/meat/high-fidelity! hehe! )

 

[andyc56]

First graph is each sub solo and then combined -green. Second graph is original 3 combo-green, pa460 substituted in for a 460HO-red, then all 4 subs-black.

*15db drop from 30hz to 20hz confuses me (4 18" ported subs should be strong to at least 20hz)?

*large peak in the 60's might come down with peq cuts, correct?

*crazy dip @ 82hz - ish (not sure how to deal with that)

Is it possible that you might have put a lot of damping material in the vented boxes that are tuned to 18 Hz, partially blocking the airflow to the ports? These subs measure like their port outputs have been suppressed. Vented boxes should have only enough damping material to prevent standing waves in the box, and the path to airflow out the ports should be kept free.

 

[platoon2063]

Is it possible that you might have put a lot of damping material in the vented boxes that are tuned to 18 Hz, partially blocking the airflow to the ports? These subs measure like their port outputs have been suppressed. Vented boxes should have only enough damping material to prevent standing waves in the box, and the path to airflow out the ports should be kept free.

I pulled the UXL driver out this morning because I hadn't done so in a while. Had one pillow in there that is now removed. The walls are lined just like the Dayton 460HO enclosures. 8 ft3 boxes w/3 x 9 x 23 ports in the 460HO boxes. The port in the UXL box is 2" longer. Boxes are made of 3/4 birch, wood glued and pocket screwed together.

 

[Jedi940]

I still think you do not have your subs properly time aligned. THIS documentation on REW shows what I'm used to seeing when looking at an impulse graph. That is why I am confused looking at yours.

HERE is another article from GIK acoustics talking about the impulse response as well.

The peaks highlighted on your graph should all line up and should all be after the 0 ms mark.

I think individual measurements of each sub showing the frequency response and the impulse response graphs would be helpful with no corrections at all. Remove all audyssey, miniDSP settings, amp settings, etc except a highpass filter to protect the ported subs. This will give us a base line and help understand what a summed response should look like. Then, following mtg90's guide listed below, you start with the sub furthest from the listening position and measure it. Then add a second sub and measure the combined response. Adjust delay of the second sub until the response sums appropriately. Wash, rinse, repeat.

mtg90’s multiple subwoofers + mains integration How To thread

 

[sdurani]

*crazy dip @ 82hz - ish (not sure how to deal with that)

Are you willing to move your subwoofers and/or seating?