Wierd Null

[mdrake]

I am getting a wierd null at 140hz and can not seem to pin point the problem. Below is a screen shot. I tried messing with the phase and it did not seem to help at all. The Maggies and sub are crossed over at 120hz which may account for some of the null. Any ideas to the cause of this?

[thewire]

How big is your room? I have about a 120 null off of my back wall in my 20.5' X 13' room. It is mostly a problem in my back row that is 4.5 feet away from the back wall.

[thewire]

Here is an example of having traps on my back wall and without.. This in an old measurement before I got my BFD. Placing panels at first reflection points also seemed to help this area.

traps on back wall.jpg

[jerome]

Before looking at your wierd null at 140hz, I would recommend you to check why you are measuring SPL over 104dB over most of your frequency range.
Did you perform the calibration correctly or are you simply playing too loud?

When it comes to your null, could it be standing waves? Check out your room modes and try to mode you sub around to see if you can remove this null.

[Wayne A. Pflughaupt]

Looks a bit large (wide) for a true null. Could be caused by any number of things, such as interaction with nearby boundaries, your listening position, etc.

Just thinking out loud here, but 120 Hz – that’s kind of a high crossover point. Maybe the Maggies aren’t extending down to the crossover point? Don’t those speakers radiate from the rear as well as the front? If so, that means you have all kinds of phase (i.e. delayed signal-related) things happening, which probably explains why the phase control on the sub didn’t do anything (assuming its variable). It may be helpful to take some more measurements, such as mains only, and only of the mains with the sub.

P.S. I'm with Jerome - why are your SPL levels so high?

Regards,
Wayne

[jerome]

Matt: could you post the same measurement with the correct logarithmic scales? Look at Steven's graph to understand what I mean...

This will probably help other people to give to better feedback !

[mdrake]

Not sure what you mean but I will read up and see how to setup the mic. If I use my Audix mic it looks the same only the SPL is accurate. I am really having a hard time understanding how a high spl "more sensitive mic" would cause an inaccurate measurement. I was just wondering what might cause my dip at 140 and that happens not matter which mike or software I use. I was also wondering how to track it down.

Matt

[brucek]

I was just wondering what might cause my dip at 140 and that happens not matter which mike or software I use. I was also wondering how to track it down.

A calibrated mic and the use of standard scaling helps with accuracy and proper interpretation.
A change to LOG scale from the LIN scale you're using will change the look of the dip quite a bit, (although of course it won't get rid of it).
Best to do your testing at 75dBSPL rather than such a high level.
Use a vertical scale of 45dB-105dB and a horizontal scale of 15Hz-200Hz.
To save your graphs, click the small floppy icon in the bottom left hand corner of your graph and it will save a 800bit wide jpg image.

That said, it is easiest to track down a problem of a dip like that (around your crossover) by taking a seperate measure of your sub and then a seperate measure of your mains and then use the All Measured tab to lay them on top of each other. It can show if one or both are at fault. Often the sub distance setting (in the receiver) can have a nice impact on dips that the phase control might not.

brucek

[mdrake]

WOW, that was really helpful!!! Thanks for all the tips. I will give them a try.
Does anyone know what mic comes with the Onkyo 805 and how accurate it is?

Matt

[JohnM]

Dips are generally caused by phase cancellations, which means the sound reaching the measurement point is a combination of sounds which have a phase difference of half a wavelength at the frequency the dip occurs. Adding two signals at the same frequency which are half a wavelength out of phase results in cancellation at that frequency and a sharp null or dip, the closer the signals are in level the more severe the cancellation.

With a sub and main speaker this can be caused by the sub phase adjustment being out compared to the main speaker. It most often occurs at or close to the crossover frequency.

Another cause of dips is sound reaching the listening position by 2 different paths, a direct path from the loudspeaker and a delayed path reflecting from one of the surfaces of the room. How sharp and deep the dip is depends on how close the level of the reflection is to the level of the direct signal, the lower the level of the reflection the shallower the dip. The frequency at the lowest point of the dip, 130Hz in the case of your plot, gives a clue to the other path. At 130Hz the wavelength of sound is about 2.6m or 8.7 feet (wavelength = 343m/frequency or 1125 feet/frequency). To have a half wavelength phase difference the reflected path would be travelling half that distance further than the direct path, so 1.3m or 4.35 feet. That rules out reflection from a very nearby surface (e.g. a speaker very close to a sidewell) but is fairly typical of the differences that occur from a floor reflection or a sidewall a few feet from the speaker. It could also occur if your measuring position is a couple of feet from the rear wall. To work out which surface the reflection is coming from, move the measuring position and see what happens to the dip. If the dip gets lower in frequency you have moved closer to the surface the reflection is coming from, if it gets higher you have moved further away. When you know what surface is the cause of the problem you can try some acoustic treatment on it (e.g. a thick rug on the floor if the problem is a floor bounce) or move your listening position or speakers to somewhere the effect is lessened.

[rafaelsmith]

JohnM wrote:

Dips are generally caused by phase cancellations, which means the sound reaching the measurement point is a combination of sounds which have a phase difference of half a wavelength at the frequency the dip occurs. Adding two signals at the same frequency which are half a wavelength out of phase results in cancellation at that frequency and a sharp null or dip, the closer the signals are in level the more severe the cancellation.

With a sub and main speaker this can be caused by the sub phase adjustment being out compared to the main speaker. It most often occurs at or close to the crossover frequency.

Another cause of dips is sound reaching the listening position by 2 different paths, a direct path from the loudspeaker and a delayed path reflecting from one of the surfaces of the room. How sharp and deep the dip is depends on how close the level of the reflection is to the level of the direct signal, the lower the level of the reflection the shallower the dip. The frequency at the lowest point of the dip, 130Hz in the case of your plot, gives a clue to the other path. At 130Hz the wavelength of sound is about 2.6m or 8.7 feet (wavelength = 343m/frequency or 1125 feet/frequency). To have a half wavelength phase difference the reflected path would be travelling half that distance further than the direct path, so 1.3m or 4.35 feet. That rules out reflection from a very nearby surface (e.g. a speaker very close to a sidewell) but is fairly typical of the differences that occur from a floor reflection or a sidewall a few feet from the speaker. It could also occur if your measuring position is a couple of feet from the rear wall. To work out which surface the reflection is coming from, move the measuring position and see what happens to the dip. If the dip gets lower in frequency you have moved closer to the surface the reflection is coming from, if it gets higher you have moved further away. When you know what surface is the cause of the problem you can try some acoustic treatment on it (e.g. a thick rug on the floor if the problem is a floor bounce) or move your listening position or speakers to somewhere the effect is lessened.

Good info...I am glad I read this..

I finally got around to trying out a BFD to EQ my sub...which turned out pretty well...but when I do a measurement with my main I have a huge dip at around 130 that I had no clue what was causing it. This info probably explains it. I experimented with speaker distances and phase but the dip was always there.

[mdrake]

John I can not thank you enough for that information!!!! It was VERY helpfu. I had forgotten the formula for calculating the wavelength, THANKS. The informaiton on how to track down the NULL was very well explained. THANKS AGAIN!!!

Matt

[tonyvdb]

mdrake wrote:

Does anyone know what mic comes with the Onkyo 805 and how accurate it is?

Matt

Thats a very good question Mike, I wonder if a person can use it with REW?

[Wayne A. Pflughaupt]

Does anyone know what mic comes with the Onkyo 805 and how accurate it is?

No cheap mic is accurate. As such any response-measuring device will have built-in calibration to compensate for the mic’s inadequacies, be it an expensive stand-alone RTA or a home theater receiver. REW has essentially the same thing with the calibration files used for the RS SPL meter.

That said, the Onkyo mic could be used, if you could come up with a graph of its frequency response. From that you could create a calibration file. But then you also have the issue that the mic probably has a low level signal instead of the line-level signal the RS meter’s mic element has; if that’s the case a separate pre-amp is needed, as is used with the Behringer ECM 8000 mic.

Regards,
Wayne

[mojogoes]

Could you use the mic8000 through a preamp instead of the receivers mic to run Odyssey....if you can/could would the mic8000 be more accurate.

[mdrake]

Thanks for respones on the mic. It is funny how the Behringer ECM 8000 is the recommended mic. Behringer is not exactly a high-end brand.

Matt

[jerome]

mdrake wrote:

Thanks for respones on the mic. It is funny how the Behringer ECM 8000 is the recommended mic. Behringer is not exactly a high-end brand.

Matt

You're correct. But this mic doesn't cost much and has a very good frequency response for sub measurements. I use it too for full range measurements but not everybody recommends it...

Most of us are only a bunch of amateurs, I guess that the pros use other (better) mics to calibrate their systems.

[Wayne A. Pflughaupt]

mdrake wrote:

Thanks for respones on the mic. It is funny how the Behringer ECM 8000 is the recommended mic. Behringer is not exactly a high-end brand.

Matt

This may surprise you…

Regards,
Wayne

[mdrake]

WOW, that is surprising. I figured there had to be a good explanation due to expertise of the people using these microphones but to see a professional using them in a mission critical situation is really remarkable. Thanks for the link!!

Jerome why is it not recommended for full range measurement?

Matt

[Wayne A. Pflughaupt]

Matt,
The mic in the link is a different model from the one Jerome's talking about, that we usually use with REW. Not sure it is who "doesn't recommend" the ECM - it's pretty popular around here.

Regards,
Wayne