Hey marty,

The only thing is it doesn't seem to reduce ringing...

As you have discovered, an equalizer doesn’t reduce ringing in the strictest sense of the word. What it does is deprive a room mode of energy. That does have the effect of reducing the mode’s ringing, but it only brings it back in line with what the rest of the room is exhibiting. It won’t make any improvement beyond that.

If your intent is to reduce ringing, you need bass traps. These graphs show the effect of bass traps on ringing; you simply can’t get this with an equalizer.

Empty Room

With Four Bass Traps​

For a more dramatic demonstration of the above graphs, open up their direct links into two browser tabs and switch between them.

http://media.soundonsound.com/sos/sep04/images/realtrapsbefore.l.jpg
http://media.soundonsound.com/sos/sep04/images/realtrapsafter.l.jpg

it seems that with rew's own filters applied there seems to be an increase in ringing compared to my attempt. It may have something to do with the fact that rew applied some boosts where as I only applied cuts?

Any increase in signal level will show up on a waterfall graph as an apparent increase in ringing. It isn’t necessarily so; it’s a common mistake to confuse a higher signal with increased ringing, but one has nothing to do with the other: Ringing – improving it or making it worse - is about the rate of signal decay. Signal levels have no effect on that. For a better understanding of signal levels vs. rate of decay, see this post.

Regards,
Wayne

 

Thanks Wayne, helpful as ever

I have gik acoustic bass traps pretty much everywhere, I'm getting a few more from Bryan soon so it will help some more.

The trouble is the ringing or the decay times are bigger low down where bass traps are not that effective so eq is really the only way to bring it under control. I did notice at the bottom of the filter tasks it has the modal resonances which I thought may indicate that rew will account for them and try and bring the decay time to below 300ms.

Marty

 

The trouble is the ringing or the decay times are bigger low down where bass traps are not that effective so eq is really the only way to bring it under control.

Unless you have some modal peaks down there, EQ is not going to help.

I did notice at the bottom of the filter tasks it has the modal resonances which I thought may indicate that rew will account for them and try and bring the decay time to below 300ms.

You might want to fully define your objectives, because “300 ms” is a pretty vague target. You can “achieve” 300 ms by reducing the signal level, or by raising the graph’s lower vertical value from 45 dB - which is already too high - to something even higher.

Also keep in mind that ringing accomplishments achieved via equalization are only “good” at the measured position today. Move the mic to another location and it's “gone.” Take some sweeps six months from now your measurements won’t look quite the same; neither will the waterfall graphs they generate.

Regards,
Wayne

 

To target filters at specific resonances you will need to do some manual work. Use 'Find Resonances' in the Modal Analysis part of the EQ window to generate the list of resonances in the measurement. Find one of the problem frequencies you see in your measurement in the list of resonances and note the exact frequency and T60 value. In the EQ filters add a filter of type 'Modal' and set the frequency and Target T60 values to those in the resonances list. Adjust the filter gain and check the 'Predicted' waterfall to see how the ringing has been affected. Try small changes in the Modal filter's centre frequency to make sure it has the correct value, you can also try changing the Target T60 value to see whether the resonance analysis found the best setting. It is a fairly interactive process, but it is straightforward by watching the effect in the predicted waterfall plot.

 

Thanks John I will give that a try.

Marty

 

midbass with no eq:

Above is waterfall of 10" driver with potential use as midbass. Not a pretty sight.

Simultaneous solution for equalizing amplitude and phase of all frequency components in FFT of impulse response results in the following:

Ringing at 90Hz is cutoff of high pass filter. Complementary ringing is set up by low pass filter of woofer, and it disappears when woofer is added to system. Much of decay is actually redistribution of driver harmonics energies into intermodulation components.

A bass trap is effectively an EQ with center frequency, a width, and a damping factor. This may be applied directly to a driver based on its performance in the room.

When too narrow a notch is applied to a peak, the result is two smaller peaks. When too broad a notch is applied to a peak, the peak is brought down, but so are neighboring frequencies.

Rooms are complex resonators, and so are speaker drivers. The room is passive and is pumped (illuminated) by the driver. The resonances common to room and driver from driver location are energized(unmasked). Balancing the frequency dependent energy at the driver required to get uniform response from the room is much easier than blasting the room with energy and shaping the room response with a large array of damped resonators.

Here is same process applied to 12" sealed woofer measured at 9':

Raw response spectrum:

Equalized response spectrum:

Woofer is driven by Hafler Pro 2400 bridged to mono 400watts. Without equalization the driver is prone to bottoming. With equalization voltage swings across driver terminals are greater, but excursion is tightly controlled. Articulation is super tight from upright bass to diesel locomotive. No room treatments, no bass traps and no smoothing.

Andrew

 

really curious about this eq for increasing decay slopes (thereby reducing decay times for a given SPL). The thread/post attached seems to contradict this thread about effectiveness of PEQ to increase decay slopes. In the thread someone tries to give evidence by using eq to bring down a specific peak, increase the overall volume and and keep the starting SPL at that freq equal and showing a reduced decay time which indicates a steeper slope since starting at the same SPl value.

Does this evidence contradict this thread? Not sure I am rreading this correct but if a PEQ can increase the rate of decay of <50HZ freqs it sure would be great vs trying to install 4 ft thick bass traps!!


http://www.hometheatershack.com/forums/rew-forum/7135-waterfalls.html

 

Not sure exactly what the contradiction would be, since there is no consensus in that thread. I mean, a thread doesn’t go on for 17 pages unless there is a lot of debate over conflicting ideas and interpretations, not to mention people just trying to get a handle on the topic. When someone says something like...

As I have frequently mentioned I boost at maximum at 20Hz to compensate for my high Fs 32Hz non-spec drivers. Here's a waterfall I generated from an older file. It seems that boosting with the BFD does indeed increase ringing.

I don't think anyone can argue that this filter is also operating in the time domain?
As a side note, you can see what a completely terrible idea it is to add a gain filter to boost the level of a sub at low frequencies. You do nothing more than emulate a room mode at the gain frequency.

Waterfall plot of a BFD using a single filter of (40Hz, Gain +15dB, BW 10)

...it's clear they don’t understand the relation between signal levels and decay times, and as such have a fundamental misunderstanding of the subject matter.

When the Waterfalls thread was opened, waterfall graphs and signal decay analysis was a new thing to most of the regulars here (JohnM being the obvious exception), and many of the posts in the thread are quite frankly in error or at least ill-informed. And I’d include many of my own to be in that category. I personally have learned a lot more about signal decay since then, and especially how to analyze the waterfall graphs.


That said, a parametric EQ will do a great job of taming the 50 Hz problem in your room and bring its decay time back in line with what the rest of the room is exhibiting, assuming it’s being caused by a mode (which typically a problem like that is).

Regards,
Wayne

 

Thank you for digging up Waterfalls thread. This is in complete agreement with my post #9, and both my post and Waterfalls thread contradict Wayne's view.

At low frequencies you are inside the speaker.

Andrew

 

Wayne,

Here you contradict yourself:

Post #8:

Note the dramatic difference above 140 Hz that absorption makes. You simply can't get this effect with an equalizer - again, it can't absorb acoustic energy.

Post #12:

That said, a parametric EQ will do a great job of taming the 50 Hz problem in your room and bring its decay time back in line with what the rest of the room is exhibiting, assuming it’s being caused by a mode (which typically a problem like that is).

Regards,

Andrew

 

Hey Andrew,

I’ll sum up Post #8 in a nutshell:

* An equalizer will reducing ringing in a room mode, but only to the point of (and not beyond) what the room is naturally exhibiting.
* If you desire a reduction in ringing beyond that point, absorption is required.
* An equalizer will not reduce ringing in areas of the frequency spectrum where there is no room mode. For that, absorption is required.

Regards,
Wayne

 

thanks guys for summing this up for me. Makes mrore sens and I totally understand that earlier people were going through a learning curve. Much like I am trying to do now with PEQ and FR and time domain equalization. I'll try to address the modes and bring it back to the room respnse. From there looks like absorption. Still have a lot more questions but I'll try to read more first before asking in case I can find the answers already posted.

cheers