Bump at 40Hz despite a 80Hz cross over… (Classe Sigma SSP weirdness?)

[Alexandre]

Hi there,

I finally got around to trying out measuring the room response with Room EQ Wizard. The results show a rather large bump around 40Hz and another one around 100Hz…

I have a Classe Sigma SSP Preamp, a single JL Audio sub, a pair of Magellan Duetto and an Emotiva 5 channel amp.

After seeing those results, I started experiment with some settings but none of it made any difference on the curve… I could change the amplitude of the overall curve but not the peaks which I assumed came from the sub. So, in order to experiment some more, I turned the sub off and tried again. The curve did change drastically but there was one very very strange thing: I still got a peak at 40Hz! I don't understand how that's possible. The speakers aren't even supposed to emit at that frequency. Could it be purely the room making this bump appear? Or did I forget some magical setting on the preamp to actually turn on all those little settings?

I am measuring at the central listening position, which, since this is the living room and not a dedicated home theater room, is a sofa, across the room and against the back wall.

Any suggestions on what to try next?

Thanks!
Alex.

 

[KEW]

Triangle's website spec's 38Hz for the -3db point, so not sure why you believe the Duettos can't produce that frequency:
http://www.triangle-fr.com/magellan_duetto_en.php

I think it pretty much has to be the room. But a good test would be to measure your sub with the speakers disconnected and confirm that your room is resonating this frequency.

PS - I am not familiar with your pre, so consider that! I suppose if it offers PEQ, it could be causing it.

Post your measurements if you can!

 

[Alexandre]

Triangle's website spec's 38Hz for the -3db point, so not sure why you believe the Duettos can't produce that frequency:
http://www.triangle-fr.com/magellan_duetto_en.php

I think it pretty much has to be the room. But a good test would be to measure your sub with the speakers disconnected and confirm that your room is resonating this frequency.

PS - I am not familiar with your pre, so consider that! I suppose if it offers PEQ, it could be causing it.

Post your measurements if you can!

Thanks Kurt,

I will post my measurements once I reboot the mac mini into OS X (right now it's OpenElec/Movie time). The reason I am thinking it shouldn't be the speakers is because crossover is way higher than the 40Hz bump. So I'm wondering if somehow the pre-amp's settings are being bypassed…

I was also thinking that I might try to do a sweep of measurements with the microphone right by the speaker to see if I can measure direct speaker output rather than room acoustics. I'll test that out tomorrow when the baby's not sleeping in the next room.

Alex.

 

[KEW]

If your pre has some type of "Pure Direct" setting, that is usually the most certain way to make sure the pre is not altering the signal.
You can also stuff socks or something in the ports to restrict flow and dampen any port tuning. I mention this in case you begin to suspect the measurement method - plugging the ports should show a major reduction in the bass.

 

[TheWarrior]

Room modes!

Meaning that specific room measurements are of equal length to one quarter of the diameter of a 40 hz or 100 hz sound wave. They get 'excited' at those frequencies and add their own sound track.

Try moving your sub (inches matter!) closer/further from walls. Also consider the location of your mains in relation to the rear wall. Those are front ported, which means they should be as close to the wall as possible, otherwise you will suffer a midrange suck out. Despite being front ported, energy still leaves the back of the cabinet and reflects off the wall, if it is too far from that wall, when the wave rebounds it will be out of phase with the direct sound from the front of the speaker, and therefore will cancel each other out.

 

[Alexandre]

Here are a couple of measurement of the speaker, with a 0-300Hz sweep, the microphone is placed right in front of the speaker (6-7 inches away), I've turned off the SUB, and unplugged the other speaker. I was hoping to minimize room modes that way.

Socks out:

Socks in the vents:

It still looks to me like the crossover is effectively 40Hz instead of the 80Hz crossover of the preamp, no?

Cheers,
Alex.

 

[TheWarrior]

That is way too close. The drivers don't even fully sum together at that distance! Move the mic back to 1m or 39-3/8" (with just a hair over the tick mark, most people just have tape measures and measure again.

 

[Alexandre]

Oh, good to know (did I mention this was the first time I did any sort of measurements? .

Here is a slightly more "controlled" – as in I probably did something else terrible – measure of the effects of the crossover on the preamp. Each time, I've just tried to only change the cross over.

4 measurements:
- Crossover OFF (speaker set to full)
- crossover at 80Hz with 12dB slope
- crossover at 80Hz with 24dB slope
- crossover at 140Hz with 24dB slope

Each time the speaker is in the same spot, same volume and this time the microphone is a little over a meter away. I don't mean this to be an accurate measurement of the speaker itself, rather, I'd like to understand if this looks like the crossover is doing its job properly. It seems to me that the slope makes nearly no difference and I would have expected to have significantly less sound below 140Hz with a 24dB slope crossover set to 140Hz, but again, maybe I am measuring the room in this case (in which case I guess I'm going to have to start investing in a lot of absorption panels).

Note that I did one more measure with 140Hz crossover with a 12dB slope which matched nearly exactly the 24dB slope one so I didn't include it here.

And for those of you looking for a laugh, this is the wonderful "setup":

Happy to take advice on things to look into, the 40Hz bump clearly seems to be in large part caused by wall reflection behind the speaker so I'll play with that but first, I'd like to figure out if the crossover settings are working as they should.

 

[andyblackcat]

If you think your taking readings partly from the room's boundary reflections then get laptop and microphone and long cable for the speaker and test it outside during maybe early evening when its a bit quieter from noise, and get some frequency graphs and compare them. Also get a long cable for the sub and do that as well outside. Outside will have far less reflections.

 

[Alexandre]

If you think your taking readings partly from the room's boundary reflections then get laptop and microphone and long cable for the speaker and test it outside during maybe early evening when its a bit quieter from noise, and get some frequency graphs and compare them. Also get a long cable for the sub and do that as well outside. Outside will have far less reflections.

That should be a fascinating experiment…

 

[andyblackcat]

Make sure you do your first test at same volume level and speaker distance from the ground and microphone distance the results should look lot a different. Then just freely experiment, but bare in mind there is outdoors traffic/airplane noise about and that will get added onto the graph so make sure to run the sweep when your sure its okay. I have traffic outside, buses and bus stop across the road and the engine noise is GREAT and the mic picks up it easily so I run any sweeps early AM hours.

 

[KEW]

First off, I am no expert at measurement, so don't take what I say as gospel. Hopefully someone more experienced will be along before too long, but, in the mean time, I will stick with you and learn as we go!

I think you are expecting a more vertical slope than you should be. If your horizontal axis is spread out, the visual angle of the slope is less.
Here is a graph of the theoretical crossover:

Note the curve is 6dB down at the 80Hz XO point and has a 12dB/octave roll-off (you can see it goes from ~63dB @ 40Hz to ~51dB @ 20Hz).
The 12dB roll-off can only be evaluated where the slope is steady (not where there is any curvature to the chart).

Real world measurements are much more complicated!

However, I can think of no explanation for your 12dB and 24dB curves to be so close other than either the pre-amp is not doing what it should, or you are missing some control which overrides whatever "switch" you are using. You might contact Classe and email them your charts. There ought to be a way to measure the pre-amps output to get the curves before the complications of speaker/room/microphone issues. It seems like you could feed your pre-amp output into you mic input, but for all I know, that might cook the circuits in teh laptop (or whatever you are using).
BTW, I really like you akimbo shelf/boxes! Great idea!!!

 

[Alexandre]

Make sure you do your first test at same volume level and speaker distance from the ground and microphone distance the results should look lot a different. Then just freely experiment, but bare in mind there is outdoors traffic/airplane noise about and that will get added onto the graph so make sure to run the sweep when your sure its okay. I have traffic outside, buses and bus stop across the road and the engine noise is GREAT and the mic picks up it easily so I run any sweeps early AM hours.

Cool, I will report when I get to it (but it's complicated enough for me to move all this stuff into reach of the outdoors that it might take some time).

First off, I am no expert at measurement, so don't take what I say as gospel. Hopefully someone more experienced will be along before too long, but, in the mean time, I will stick with you and learn as we go!

I think you are expecting a more vertical slope than you should be. If your horizontal axis is spread out, the visual angle of the slope is less.
Here is a graph of the theoretical crossover:

Note the curve is 6dB down at the 80Hz XO point and has a 12dB/octave roll-off (you can see it goes from ~63dB @ 40Hz to ~51dB @ 20Hz).
The 12dB roll-off can only be evaluated where the slope is steady (not where there is any curvature to the chart).

Oh, you make a good point there, I keep forgetting the really heavy scaling of the horizontal axis on those measurements! I also had forgotten what an octave represented on the frequency chart so after looking at that closer I guess it's not all that unexpected.

Real world measurements are much more complicated!

The difference between theory and practice: in theory there is none!

However, I can think of no explanation for your 12dB and 24dB curves to be so close other than either the pre-amp is not doing what it should, or you are missing some control which overrides whatever "switch" you are using. You might contact Classe and email them your charts. There ought to be a way to measure the pre-amps output to get the curves before the complications of speaker/room/microphone issues. It seems like you could feed your pre-amp output into you mic input, but for all I know, that might cook the circuits in teh laptop (or whatever you are using).

I will contact Classe and ask them if they have some suggestions.

BTW, I really like you akimbo shelf/boxes! Great idea!!!

Are you talking about the bookshelves in the back? It's a designer from columbia, I have 2 of his bookshelves and that's usually the first thing anyone sees when walking into the living room and then starts asking how they magically float atop one another, they look pretty incredible: http://malaganadesign.com

Cheers!

 

[KEW]

Are you talking about the bookshelves in the back? It's a designer from columbia, I have 2 of his bookshelves and that's usually the first thing anyone sees when walking into the living room and then starts asking how they magically float atop one another, they look pretty incredible: http://malaganadesign.com

Cheers!

I assumed you just mounted separate boxes on the wall that way!
As a stand alone piece they are quite intriguing. I'm not sure how you would do that and it be reasonably sturdy! Nice trick!

 

[andyblackcat]

A table set up in the garden don't you have a garden with patio door windows leading to the garden.

Place laptop on table with a chair that you can sit on and make it comfortable for yourself.

Use that grey box to place the speaker.

Look for some extra speaker cable that's in a box and connect the ends to existing cable to extend the length to garden.

Use that little stool to place the microhone or buy a boom stand that are reasonable cheap on ebay that way you can alter the angle or height of the mic on the boom stand.

 

[TheWarrior]

Alexandre, your crossover is working just fine. That speaker is not designed to play in the middle of a room. And the room, is not an anechoic chamber. So move it back to its intended place. Take a measurement 1m back, and at ear height at your preferred listening position.

I notice in your pic that the top of those speakers is level with the door handle. Are your seats that low to the ground?

 

[KEW]

Alexandre, your crossover is working just fine. That speaker is not designed to play in the middle of a room. And the room, is not an anechoic chamber. So move it back to its intended place. Take a measurement 1m back, and at ear height at your preferred listening position.

I notice in your pic that the top of those speakers is level with the door handle. Are your seats that low to the ground?

Are you sure about that? Did you see how closely the 24dB roll-off tracks the 12dB roll-off? Regardless of room issues, I would expect there to be two clearly distinct curves instead of "one".

 

[TheWarrior]

Are you sure about that? Did you see how closely the 24dB roll-off tracks teh 12dB roll-off? Regardless of room issues, I would expect there to be two clearly distinct curves instead of "one".

Yes. But ask Alexandre, he's the one who gets to take these measurements! (and hopefully ends up with even better sound when we're done!)

 

[TheWarrior]

Oh and I still don't know where I saw the graph we discussed earlier, but this site came up in searching. http://arqen.com/acoustics-101/speaker-placement-boundary-interference/


Unrelated to the thread, but relevant nonetheless!

 

[TheWarrior]