[MetropolisLake]

So basically I got a flat response with minidsp, but, I pulled down my peaks so much that it squished the entire output. Now, my subs are really quiet. Tried to increase gain on the amp and increase output on the receiver but its not enough plus its picking up noise. I thought I could bump up the overall output but I guess not, can only cut it I guess.

Am I missing something or do I need to redo everything and bring levels up rather than cut my peaks down to the baseline on the low end of the spectrum? Attached is my before and after response. The flat line is with REW 3 db higher and I think a couple of clicks on my amp. I just want to make this line be at a higher output level.

 

[JohnM]

Your target level was set too low so the filters ended up being used as a volume control. Probably need a target level around 62 dB with that response, but difficult to be sure without seeing what it does a little further up. It also looks like a sub + main measurement rather than sub only, unless there is no crossover applied. Bear in mind you can't use filters to make a sub that rolls off at 35 Hz into one that is flat to 10 Hz, filters are really only good for dealing with the peaks above the sub's low end roll-off - to address the low end try different positions (e.g. sub in a corner), but ultimately for very low extension a sub that has a very extended low end response is required, which typically means big.

 

[MetropolisLake]

Your target level was set too low

I had no target level set period as far as I know. I just plugged it in, played it at a comfortable level that didn't wake anybody up, then adjusted the response manually. I didn't export any curves or anything from REW and as far as I know MiniDSP has no target SPL or anything of the sort.

looks like a sub + main measurement rather than sub only, unless there is no crossover applied.

This is my laptop being plugged straight into the sub amp. I didn't want to blend in the mains yet. The receiver wasn't even on.

Bear in mind you can't use filters to make a sub that rolls off at 35 Hz into one that is flat to 10 Hz, filters are really only good for dealing with the peaks above the sub's low end roll-off - to address the low end try different positions (e.g. sub in a corner), but ultimately for very low extension a sub that has a very extended low end response is required, which typically means big.

These are four 18's sitting in a granite covered box that weighs 800 pounds. That big enough? They're just not spread out, they're all under the screen so I still have nulls and peaks due to room modes, and no I'm not going to run sweeps 100 times at full blast so that's why it's not 115 db.


It seems that my mistake was to bring everything down to the level of what it is around 20 hz but the more I pulled down, the more the entire volume was pulled down, and now that I'm done there's just no saving it. Looking for specific alternatives and advice to this situation because I'm afraid I'm going to have to start over and adjust with a different strategy.

 

[JohnM]

These are four 18's sitting in a granite covered box that weighs 800 pounds. That big enough?

Certainly should be hard to reconcile that with the posted response though.

I'm not going to run sweeps 100 times at full blast so that's why it's not 115 db.

The level of the test was OK, it's the level aimed for during the EQ that wasn't. Need to understand why the response looks like that in the first place before applying the EQ. Difficult to play with placement when the sub is that heavy, but something is off. Do you have any measurements of the sub before it was installed?

 

[MetropolisLake]

Need to understand why the response looks like that in the first place before applying the EQ. Difficult to play with placement when the sub is that heavy, but something is off.

What exactly looks off? I can shut half of them down and the response looks exactly the same, just lower output. It's just four sealed ultimax's in a little over 4 cubic feet after displacement. The 37 and 53 hz boosts and the 82 hz null is expected if you run a room mode calculator for 15x20x9.

hard to reconcile that with the posted response though.

The downward slope on the low end is pretty typical with sealed subs that have zero EQ or DSP, adding more drivers right beside each other isn't going to fix itself, just going to make everything louder.

Do you have any measurements of the sub before it was installed?

Should look very similar to the measurement on data-bass, it's the same driver and about the same size box. No I didn't test my box outside though. It's a heavily braced cabinet grade MDF box stuffed with R-19 insulation.

 

[Wayne A. Pflughaupt]

What exactly looks off? I can shut half of them down and the response looks exactly the same, just lower output. It's just four sealed ultimax's in a little over 4 cubic feet after displacement.

What’s “off” is that an 18” sub that’s “done” at 40 Hz. Most decent 8-inchers can do that good!

Are you saying you have four of those speakers in a single enclosure? If so that’s a huge part of the problem. According to Parts Express, four cu. ft. is the right size for one of them. If they’re all sharing the same enclosure then each one is “seeing” only 1 cu. ft. That would account for the poor extension.

Regards,
Wayne

 

[MetropolisLake]

Are you saying you have four of those speakers in a single enclosure? If so that’s a huge part of the problem. According to Parts Express, four cu. ft. is the right size for one of them. If they’re all sharing the same enclosure then each one is “seeing” only 1 cu. ft. That would account for the poor extension.

 

[MetropolisLake]

something is off

Actually something was off. My amp seems to be fried. It sounds ok down low but up high it distorts badly with peaks. Kick drums sound like the box is made out of sheet metal and gives a weird percussive clapping sound rather than a low thud. It does this in or even out of the box, with or without the MiniDSP, all four subs do it even if it is just running stereo into one, my amp clipping indicators is not getting past the second light and it still does it. A few days ago I ran the new subs hard at 18 hz in free-air to break them in, music hasn't sounded right since then, thought it just needed EQ but I guess not, maybe I damaged the amp.

I did redo the EQ on miniDSP and got the levels to a good place, it is flat all the way down without wimping out, but it sounds like up high on kick drums. Will have to ship this amp off for replacement or repair I guess, it is still under warranty. Supposedly the turnaround is only 10 business days or less. Probably will just sell it and get a speakerpower if they replace it, getting tired of these stupid chinese class D's.

 

[MetropolisLake]

What’s “off” is that an 18” sub that’s “done” at 40 Hz. Most decent 8-inchers can do that good!

I guess I'm at a loss to figure out what is appropriate and expected. Amp issues aside, this is in-room and:

1. I have a big predicted boost due to room modes at 37 hz which is what is making that peak. Pull up a room mode calculator and enter 15x20x9 and you'll see it. Interestingly, there's a big bump at this frequency.

2. Using the method of taking the speed of sound divided by frequency and assuming room gain starts at half this wavelength, I should get a boost starting at 28 hz on down. Interestingly, there's a bump there.

3. I should have another huge boost at around 57 hz. It showed up on the graph as 53 for some reason, but sure enough, it is there.

4. If you stay above 28 hz and ignore the 37 hz room mode and look at the general expected curve through 44 hz before you get to the mid-50's room mode issue, there's about a 4 db drop in this frequency range.

5. If you look at the raw response of the ultimax on data-bass.com in the same size box that I have, interestingly, it is identical to this:

http://www.data-bass.com/data?page=system&id=116&mset=128


So I guess my question is why anybody would expect to see a completely non-EQ'ed sealed sub with zero DSP magic going on, all sitting in the same position and not spread around the room, have a much flatter response, then blame it on the subs if it doesn't? I have no doubt that if you get a ported sub with a built in amp that has a built-in DSP the graph would be much different. I can't tell that DIY sealed subs with no DSP magic is going to ever be much different without spreading them around the room though. I'm actively trying to find graphs that prove this to be wrong and it seems that everybody else has about the same issue going on. Sealed subs in the same location is going to need a bunch of EQ to be flat, I can't see that there is any way around this. Even the SI 24" in a huge 16 cubic foot box has a raw response that falls off a cliff in the same manner, it's a 10 db/octave drop below 37 hz.

 

[willis7469]

Where is the LP, and have you measured other spots around the LP just to see how the room responds? If the subs can't move, try moving the LP. (If only to see). Mode calculators leave out a couple things(IME), and those are where the subs are and where the listener is. For example, you may have subs that are flat to 7hz if you're 2' off the back wall, but might have a huge peak or null if you're more towards the canter of the room. (That part was arbitrary so don't hold any numbers against me lol)


Sent from my iPhone using Tapatalk

 

[MetropolisLake]

Where is the LP, and have you measured other spots around the LP just to see how the room responds? If the subs can't move, try moving the LP. (If only to see). Mode calculators leave out a couple things(IME), and those are where the subs are and where the listener is. For example, you may have subs that are flat to 7hz if you're 2' off the back wall, but might have a huge peak or null if you're more towards the canter of the room. (That part was arbitrary so don't hold any numbers against me lol)

Room is 15x20 with 9' ceilings. Subs are on one long wall. Seats are 4' from the back of the other long wall. Eventually there will be two more there for nearfield but not for now.

To visualize what I was saying about room gain vs. raw response, you can see it here, second one down on the right.

http://www.data-bass.com/data?page=content&id=80#!prettyPhoto

That's a DIY sealed 12" with a raw frequency response rolloff similar to most other sealed subs, even very large ones. Notice what happens when you put it in a room. There are multiple 10-13 db peaks as well as a 25 db null, with the overall response vaguely following the ground plane response. Perfectly normal and expected in my opinion. Yeah ideally you'd spread the subs around and also use EQ but to just look at that response in the room and say something is wrong with the sub, just isn't reality. If that's all that four 18's could put out, sure, but that's not the conversation here. Every sealed sub or even multiple sealed subs with no DSP magic is going to have similar issues.

 

[MetropolisLake]

As a follow up, this is my last sweep. Almost 115 db at 12 hz. It's usually flatter but I was kind of cranking it, probably kicked in some soft limiters on the amp. Got it to look this good by taking the MiniDSP out of the equation. I'll keep studying on it but at this point it is doing more harm than good.