Measurement confusion

[datune]

Hello!

I'm doing some measurements with Mackie HR824 MK2's and my Yamaha YST SW515 subwoofer, and I get some very confusing results.

Here's what I did (note, the measurement microphone was placed at groundlevel, since I have the subwoofer where I would be sitting):

1. Mackie's: Set the cutoff to 47Hz, and measured using ONLY the Mackie's, and I get the following (confusing) result:



2. I then measured ONLY the sub, with it's cutoff set to about 47Hz



Now my questions are this:

1. How come it peaks at 47Hz from the Mackie?

2. How come the subwoofer has so much going on ABOVE 47Hz, i set the cutoff, so I expected to see it going downhill around 47Hz, yet it peaks again at around 65Hz

What am I missing here? Has this to do with room modes?

FYI: the room is 13.12ft x 10.5ft x 8.2 ft, and has some acoustical DIY treatment

http://www.mackie.com/products/hrmk2series/downloads/HR824MK2_OM.pdf


http://www2.yamaha.co.jp/manual/pdf/av/english/sub/YST-SW515_en_UB_rev-1.pdf

[datune]

And now I am completely lost, here's the frequency response of ONLY the Mackie's with the cutoff set to 80Hz





How can this be????

How is it possible that there is a PEAK at 40? I'm really confused right now...

[Ethan Winer]

Originally Posted by datune

What am I missing here? Has this to do with room modes?

Mackie 824 speakers are very flat, and all the peaks and nulls are due to your room. You need bass traps. Lots of bass traps. All rooms need lots of bass traps. What you measured is not only common but typical. All small rooms are like this.

--Ethan

[datune]

Ethan, what I ment is, how can there be so much output below 80Hz fom the mackies, even though I have set the cutoff switch to 80hz?

I have 10 basstraps, 3 corners are treated from floor to ceiling, and I have two "RFZ" panels above my head (all DIY, basically just fiberglass with fabric covering it, no frames...)

There is something really weird in my room, namely that there always seems to be this peak at 40Hz.

I remember from another post that someone said:

It looks like you have a primary mode around 40Hz with predictable responses at the harmonics

Could this be, and if so, what can I do? (let me guess? MORE basstrapping? )

[jayfrigo]

You want to have the mic at least 1/2 wavelength away (for the lowest desired frequency to be measured accurately) from any boundaries (wall, floor, ceiling), so you don't want to lay the mic on the floor. You can't get numbers you can trust that way. Also, are you doing one speaker at a time, or the pair?

Looks like you have some ringing, and plenty of room modes as is to be expected. Though again, with your described mic placement, I wouldn't consider these graphs the last word. Also, did you alter the gain? The first Mackie graph looks lower in level than the second. And, what does everything look like combined?

As for why the satellites alone show so much output below the supposed filter turnover frequency, I couldn't say. That's a legitimate question. It might be informative to know the filter type and slope in use, but still, it doesn't look like much of anything is filtered. Did you take a measurement with the filter in the off position to compare? And again, get 1/2 WL away from boundaries, rinse, repeat...

Room and mic/boundary anomalies aside, the sub response alone looks like it makes sense (as far as one can tell from what's provided). Looks like the selected frequency is the minus 3 dB point, and the filter is a reasonably steep 24 dB per octave. That sounds about like what one would expect from this kind of sub.

[datune]

Thanks for the replies.

Just for clarification, I had the mic a few inches away from the floor.

Still, I turned the subwoofer off, and placed the mic (Behringer ECM8000) at the mixer position and measured

Red line is with NO lowcut
Green line is with lowcut set to 80Hz

The mackie website seems to be down atm, so I jsut took a picture of the manual, showing the slopes.

I don't understand what's going on, to me it seems like the frequency response is EXACTLY the same, except a few db's less.

[avare]

What are the dimensions of the room, the locations of the speakers in inches or mm, the location of the microphone in inches or mm and what is the construction of floor?

Andre

[datune]

The dimensions of the room are listed int he first post; but since I am currently floored anyway due to not handling the lugging of the speakers well (and thus hurting my back), I figured I draw it all out, if more information is required let me know.

The floor consists of tiles, but I am planning to put a rug (1m x 1.5m where the mix position is)

[datune]

I also computed the Axial modes, and it confirms the peaks at 43Hz, but obviously doesn't explain the other issue.

[avare]

Sorry I missed the room dimensions.

Got it. The mic is against the floor where height modes are strongest. That is the 70 Hz peak. Try locating the mic at normal listening ear level.

Andre

[datune]

Erm, no, the mic is at normal ear height, it's not against the floor, and what 70Hz peak are you talking about? You ment 40 i assume?

[datune]

This is really driving me nuts. I don't understand how this is possible.

How can the EQ curves be the same (only db difference) whether the 80Hz cutoff is set or not on the Mackie's?

I have taken at least 70-80 measurements, at numerous spots, and it's always the same story, no cutoff present, only a difference in db.

And the mackie site still seems to be offline!

[DanDan]

Hi Datune, well, don't go nuts, your room hardly warrants that!
Measurements look deceptively simple. Without context they have little meaning. Would I be right in guessing that there is concrete under your floor tiles? Are several of the walls concrete block? Such a room would be acoustically 'hard' even at low bass frequencies. I would not be surprised to measure modal peaks and dips up to 25dB or more. Such variations would overide your nice crossover slopes. Furthermore those crossover curves are of electronic filters. A graph of output of the speaker unit would not look at all like that. Don't sweat it. It will take very very serious large bass treatment to approach your problem. I suggest you go for the mostest. 34 inch wide Superchunks , MondoTraps etc. Consider floor and ceiling 'corners'. Can you leave that door open when you are mixing? Why the SubWoofer? It is common practice in such rooms to chose a speaker which does not go low enough to strongly provoke the first mode. Your Mackie is a fine example of just that. Experiment with distance from the wall and speaker height. Such changes are little or no cost and can make an impossible situation tolerable. My CR is of a similar size to your and is mostly concrete. The room from hell. I have achieved a very workable acoustic in there by using massive amounts of Absorbtion. See the attached pic. There is a Cloud of four MiniTraps omitted for visibility.



Good Luck with it,
Dan FitzGerald AMIOA
SoundSound - Homepage

[datune]

Hey DanDan!

Yes, the floor below the tiles is concrete, the walls however are made of bricks.

To this day I could not figure out why the Mackie's show the same freq. curve whether the cutoff is set or not.

The reason for the sub is because the music I tend to mix is bass heavy, and because the mackie's together with the sub simply sound better as opposed to just the mackies (though I realize this might be deceptive). The funny thing is, the mackie's together with the sub produce a reasonable flat response, but the decay time is still too long, which potentically makes it sound boomy (probably because the sub stimulates the room modes even more).

What I have done meanwhile is studying acoustics A LOT more, trying to figure out the math behind it all. But it's much more complex as I hoped it to be. The more I learn the more I think that today's acoustical products that are available are not really all that expensive ;-)

Another thing I have done is asking a top mastering studio whether they would allow me to visit them and get a quick tour, just so I can experience for myself how MUCH difference there REALLY is between a moderate setup, and a perfect setup with the best equipment available, and built, measured and treated by the world's top acousticians.

And unless I find the difference to be really significant (compared to my room that is), I'm going to give myself a break visa vi acoustical treatment ;-), be happy with what I have, and..well...you know...make music!

[DanDan]

Hi Datune, Don't sweat the Mackie, it's probably working properly, the curve won't change much in the context of enormous room issues. If you really want to know, measure it outdoors!
What is your ceiling made of? More concrete? If so you are in hell.
It looks like you have done your work with 10 traps, are they Superchunks? Filled with? You should be able to find an optimised spot for the Sub using the sub. There is a very good demo of such a procedure at Welcome to ETF Acoustics
I noticed a pipe resonator/trap on page 229 of Alton Everest's Master Handbook of acoustics. This is a low Q Helmholtz device which very effectively shortened a 47 Hz Boom. This seems to be a very simple device made from readily available materials. I am about to build one myself to try to tame my 37 Hz problem?
Regards, DD

[datune]

Hey DanDan,

yeah, the ceiling is made of concrete.

The traps are just fiberglass covered with fabric, no wooden frames, nothing.

Keep us posted whether you got rid of the 37Hz issue.

[Ry-Fi]

Just came across this thread in a search and found it odd that no one had alleviated the poor OP's confusion and frustration by pointing out that the 80 Hz high-pass IS doing exactly what it's supposed to be doing. It appears to be a 12 dB/octave filter with a -3 dB point at 80 Hz. 80 Hz looks about 3 dB down, and 40 Hz is about 13 or 14 dB down. It's doing what it should be doing.

I just needed to say this, because I really thought it was strange that no one else did.

The real issue is the mode-induced peak in the response, not any kind of problem with the filter...

Originally Posted by datune

Thanks for the replies.

Just for clarification, I had the mic a few inches away from the floor.

Still, I turned the subwoofer off, and placed the mic (Behringer ECM8000) at the mixer position and measured

Red line is with NO lowcut
Green line is with lowcut set to 80Hz

The mackie website seems to be down atm, so I jsut took a picture of the manual, showing the slopes.

I don't understand what's going on, to me it seems like the frequency response is EXACTLY the same, except a few db's less.

[nosebleedaudio]

Your mix position is off center, have you done any tests in the center(left/right) ?
With ridgid boundaries your walls/ceiling are reflecting very well...