Waterfalls

[Wayne A. Pflughaupt]

Nick,

brucek’s graphs were presented as a visual aid for his narrative. No, the Behringer does not add ringing to an electrical signal, nor does any other equalizer. It’s the room that does that.

Don’t let all the theory confuse you. It’s easy enough to take your own REW readings and compare equalized to baseline and see what you get.

Finally, should I not be using a waterfalls plot in determining how bad the room accoustics are for my living room?

REW can certainly show you the effects of equalizing and treatments, if you apply them.

Regards,
Wayne

[Blaser]

Just check my small experiment supported with graphs at the beginning of the thread ...This is not a simulation but true measuements

[PeteD]

Wayne A. Pflughaupt wrote:

I believe that John clarified here that you are essentially correct. Boosting a null shouldn’t result in ringing that’s any (or at least much) worse than it would be if the null wasn’t there.

Thanks Wayne. I guess my point was that if cuts help with an area of higher SPL and resonance, shouldn't boost help with the opposite situation? It could be the room or a lack of low frequency capability of the subwoofers, but sometimes a little boost is useful and I do not think it would introduce resonance.

Now, there are of course other reasons to use boost judiciously or not at all, including maximizing the signal to noise ratio through the BFD (although I believe one has to look at the entire system's S/N ratio, also), and the fact that boosting a null eats up your sub amp's headroom (needlessly, if you really have a null that does not respond). However, a little boost can be used to treat a low line level output from the BFD, which was my problem with using just cuts.

Unfortunately, I think there needs to be a case-specific evaluation by each person as they tweak their system. Luckily, there forum is available to guide people along the way.

Pete

[Blaser]

PeteD wrote:

However, a little boost can be used to treat a low line level output from the BFD, which was my problem with using just cuts.

Unfortunately, I think there needs to be a case-specific evaluation by each person as they tweak their system. Luckily, there forum is available to guide people along the way.
Pete

I agree with you about boosting to get back the signal that was reduced by cuts. This will have its problems but the benefits are greater IMO.

[MakeFlat]

Wayne A. Pflughaupt wrote:

Are you talking about phase or polarity? I believe phase is more-or-less a time alignment issue – not sure you one could make it “oppose.”

Regards,
Wayne

Perhaps polarity is the right way. Here's what I'm thinking. Run the second sub through a bandpass filter with center frequency the same as the room resonance. The second sub would have the polarity inverted. Sorry I don't have a second sub to try.

[DrWho]

I should mention that there is absolutely no doubt that the BFD is NOT introducing any kind of ringing when the filter is boosted. What we're seeing is precisely the effects of the window used to generate the waterfall...not what's happening in real life.

I brought it up to point out that changes in amplitude can pretend to show differences in ringing (as so clearly demonstrated for us). Since we know that no ringing is added when a frequency range is boosted, we can use the graphs to normalize what constitutes no change in decay when amplitude is changed.

In other words, any EQ added to your subwoofer is going to show the exact same difference on the waterfall that the EQ made....and it's not because the EQ magically made the room ring less. Likewise, any natural amplitude variations with the subwoofer will also show up as ringing (and dips will show up as decaying faster).

[JohnM]

DrWho wrote:

I should mention that there is absolutely no doubt that the BFD is NOT introducing any kind of ringing when the filter is boosted. What we're seeing is precisely the effects of the window used to generate the waterfall...not what's happening in real life.

That's not correct. The windowing effects are the cause of the response across the freq band not dropping sharply to the noise floor when no filters are active, but every EQ filter (gain or cut) rings at its centre frequency, the higher the Q (the narrower the bandwidth) the longer it rings. The ringing is easily observed in the impulse response by setting up the BFD in loopback and making a measurement with a sharp filter, it is more obvious with boost than cut but both ring. It is because of this time domain behaviour that EQ filters are able to counteract the ringing of modes, the decay of the EQ filter's attenuation over time matches the decay of the mode's gain, when the filter is properly set to match the mode's bandwidth and gain.

[PeteD]

JohnM wrote:

That's not correct. The windowing effects are the cause of the response across the freq band not dropping sharply to the noise floor when no filters are active, but every EQ filter (gain or cut) rings at its centre frequency, the higher the Q (the narrower the bandwidth) the longer it rings. The ringing is easily observed in the impulse response by setting up the BFD in loopback and making a measurement with a sharp filter, it is more obvious with boost than cut but both ring. It is because of this time domain behaviour that EQ filters are able to counteract the ringing of modes, the decay of the EQ filter's attenuation over time matches the decay of the mode's gain, when the filter is properly set to match the mode's bandwidth and gain.

So, can I assume that would part of the key to using the waterfall to tweak your response...adjusting the bandwidth of the filters to modify the waterfall plot?

What is the ideal decay rate (or realistic desired decay rate where sound quality is not compromised - i.e. we obviously aren't listening in anechoic rooms), or do we just want it to be relatively even across the frequency response?

I guess could also be said that boost in particular should be performed at as wide a bandwidth as possible, with subsequent cuts to included peaks, instead of boosting individual dips - which is more likely to introduce ringing?

This is a very informative thread - it motivated me to pick up a sound card yesterday.

Pete

[JohnM]

PeteD wrote:

So, can I assume that would part of the key to using the waterfall to tweak your response...adjusting the bandwidth of the filters to modify the waterfall plot?

Exactly, though centre frequency and gain might also need tweaking for best results.

PeteD wrote:

What is the ideal decay rate (or realistic desired decay rate where sound quality is not compromised - i.e. we obviously aren't listening in anechoic rooms), or do we just want it to be relatively even across the frequency response?

Ideally decay would be uniform across the band, for domestic rooms the ideal is thought to be around 250-300ms, though there is more tolerance of extended decay times at low frequencies than high. Look for discussions about RT60, a measure of decay time (more relevant for large venues but often discussed in the context of domestic rooms also).

PeteD wrote:

I guess could also be said that boost in particular should be performed at as wide a bandwidth as possible, with subsequent cuts to included peaks, instead of boosting individual dips - which is more likely to introduce ringing?

Right again

[Wayne A. Pflughaupt]

John, where do phase changes from filtering fit into the picture? I lifted this from Rane’s Exposing Equalizer Mythology note. The bolded text (emphasis mine) pretty much reads like what you’ve often stated here at HTS, except he’s talking about phase shift, not modal ringing. Comments?


Phase shift is not a bad word. That it has become a maligned term is most unfortunate. This belief stands in the way of people really understanding the requirements for room equalization.

The frequency response of most performing rooms looks like a heart attack victim's EKG results. Associated with each change in amplitude is a corresponding change in phase response. Describing them as unbelievably jagged is being conservative. Every time the amplitude changes so does the phase shift. In fact, it can be argued that phase shift is the stuff that causes amplitude changes. Amplitude, phase and time are all inextricably mixed by the physics of sound. One does not exist without the others.

An equalizer is a tool. A tool that allows you to correct for a room's anomalies. It must be capable of reproducing the exact opposite response of the one being connected. This requires precise correction at many neighboring points with the associated phase shift to correct for the room's opposing phase shift. It takes phase shift to fix phase shift. Simple as that.

One way people get into trouble when equalizing rooms is using the wrong type of equalizer. If an equalizer is not capable of adding the correct amount of phase shift, it will make equalizing much more difficult than it has to be. The popularity of the many constant-Q designs has come about because of this phenomenon. Equalizers that produce broad smooth curves for modest amounts of boost/cut make poor room equalizers, and good tone modifiers. They lack the ability to make amplitude and phase corrections close together. Lacking the ability to make many independent corrections with minimal interference to neighboring bands restricts their usage primarily to giving a shape to an overall response rather than correcting it. Serious correcting requires sharp constant-Q performance, among many other things.

Only by adding many precise, narrow phase shift and amplitude corrections do you truly start equalizing a system's blurred phase response. You do not do it with gentle smooth curves that lack the muscle to tame the peakedness of most rooms. Broad smooth curves do not allow you to correct for the existing phase shift. Its just that simple, you must pre-shape the signal in both amplitude and phase. And that requires narrow filters that preserve their bandwidths at all filter positions.

Regards,
Wayne

[atledreier]

That seems to somewhat contradict your general advice to use broad filters, Wayne. I use your advice with good results, but whatever works.. I'd like to know more about this.

[JohnM]

The amplitude and phase response of EQ filters is inextricably linked. What the passage says is use narrow filters to counter narrow effects, i.e. modal resonances. To counter a mode the corresponding filter needs to match it precisely, which is why smoothing should not be used when viewing responses with the aim of addressing room modes. The waterfall plot is a good indicator of whether a filter's bandwidth and centre frequency are correct to deal with a mode - it is difficult to do this with the initial response alone as the overlapping effects of the various modes can make it hard to separate them, as the waterfall progresses the strongest modes (which most need correcting) stand out as the response around them decays, making it easier to work out the filter settings needed to address them.

All that is not an argument against broader filters for response shaping, that is another tool in improving the overall response and the only one that can be usefully applied above about 200Hz as modal eq filters would only be effective over a very small area at such frequencies. Both should be combined with careful placement of speakers and listener to start with the smoothest response that can be practically achieved in the space and whatever room acoustic treatment can be accommodated, even sympathetic choice of furnishings, drapes etc can be very beneficial.

[Wayne A. Pflughaupt]

Okay, so basically phase shift and modal ringing are the same thing. So in essence what the Rane man is saying here...

In fact, it can be argued that phase shift is the stuff that causes amplitude changes. Amplitude, phase and time are all inextricably mixed by the physics of sound. One does not exist without the others.

...is that change in modal ringing (phase) from an equalizer is no big deal. I agree.

atledreier wrote:

That seems to somewhat contradict your general advice to use broad filters, Wayne. I use your advice with good results, but whatever works..

I get the impression he’s mainly talking theory more than practical application, because...

Only by adding many precise, narrow phase shift and amplitude corrections do you truly start equalizing a system's blurred phase response.

...for one, I don’t know of anyone who makes an equalizer like that – i.e., allows for numerous highly-precise filters.

Second, I assume he would know that it would be impossible to achieve this for every location in a room.

Third, from what I understand, room modes as we know them in our little rooms are not a significant problem in the large rooms you typically see in Rane’s world.

Regards,
Wayne

[Wayne A. Pflughaupt]

JohnM wrote:

To counter a mode the corresponding filter needs to match it precisely...

Am I the only one who has “repeatability” problems? A measurement I take say, today is a little different than what I got back in the summer, or even last week - different enough for REW to generate slightly different modal filters. Which are the precisely-perfect filters for the room mode?

Regards,
Wayne

[terry j]

Wayne A. Pflughaupt wrote:

Am I the only one who has “repeatability” problems? A measurement I take say, today is a little different than what I got back in the summer, or even last week - different enough for REW to generate slightly different modal filters. Which are the precisely-perfect filters for the room mode?

Regards,
Wayne

How much natural variation would you get between measurements over a period of time, given that the mic is probably not exactly in the same spot, small changes may or may not have been made in the room, different ambient conditions etc ?

I know the bass is not as twitchy as the mids and highs etc when the mic moves just a bit, but are the minor changes in the bass from different mic positions enough to account for what you are talking about?

[atledreier]

I'd say screw EQ for anything over deep bass, and treat the room! That's what I do, and I'm very happy. I do one filter at 33Hz and that's it!

[Blaser]

atledreier wrote:

I'd say screw EQ for anything over deep bass, and treat the room! That's what I do, and I'm very happy. I do one filter at 33Hz and that's it!

You are one lucky man!

[JohnM]

Wayne A. Pflughaupt wrote:

Am I the only one who has “repeatability” problems? A measurement I take say, today is a little different than what I got back in the summer, or even last week - different enough for REW to generate slightly different modal filters. Which are the precisely-perfect filters for the room mode?

Filter settings are likely to need tweaks to get the best match, the waterfall plot is helpful in assessing what tweaks are improving the result. For good results the centre frequency needs to be within 1% of the mode's frequency, the bandwidth settings is less sensitive. Once the right values have been found they won't change unless the room is changed in some way (altering the modes), or the sub is moved, or you measure at a different position.

[Wayne A. Pflughaupt]

John,

Is there any particular reason why these graphs are displayed in a 300 ms window and not some other value?

Regards,
Wayne

[JohnM]

That's just a convenient starting point that is often suitable for looking at low frequency behaviour, can dial in whatever window width and time span you find appropriate, bearing in mind that shorter windows decrease frequency resolution (the resolution is shown next to the window control).