Paul Hales on room equalization on Soundstage! Network

[DaveBoswell]

Attached you may find a recorded kick drum transient (duration roughly 200 ms).

This is a REW scope capture at MLP (3,5 meter distance) of my system reproducing this signal:

View attachment 556684

For the sake of simplification, this is the impulse of the signal downmixed to mono:

View attachment 556685

And this is system's impulse response from the scope data:

View attachment 556686

Overlay of the impulse responses:

View attachment 556687

Phase around 44Hz fundamental:

View attachment 556695

Overlay of ETC plots:

View attachment 556688

Wavelet 1/1 of the signal:

View attachment 556689

Wavelet 1/1 of the system response:

View attachment 556690

This is a rather crude method, measured at a single point. However it can give you a rough idea on the potential signal demand and some idea about in room reproduction of it.

Very interesting analysis - thanks for sharing. I have been meaning to try something similar in my system for some time so this is helpful.

Did you by chance think of measuring the time domain response at the MLP of the drum beat signal before and after the application of PEQ? This would give some more credence potentially to the premise in this thread from the Video that PEQ reduces tactile impact. I would have thought that they would not be as dramatically different as your "Overlay of the Impulse Responses" plot.

Also - IMO I think the overlays of Impulse and ETC might be a little misleading as the vertical scale may not be directly comparable between the two: if both impulse responses result in the same SPL response in the frequency domain, then the sum of the area under the ETC curve over the time window must be equal - the only way this could be true is if the initial rise of the Signal is much higher than that of the in-room MLP measured result.

Is there a way to corelate the REW Scope vertical axis measurements to actual SPL?

 

[Few]

Hales also argues for wide directivity speakers because (he argues)

His https://www.proaudiotechnology.com/s2115sm above. Nothing wide, normal 90x60º.

The video included in post #137 in this thread makes clear very wide directivity (he talks about 180 degrees) is something he has worked on extensively.

 

[audiofooled]

Very interesting analysis - thanks for sharing. I have been meaning to try something similar in my system for some time so this is helpful.

Did you by chance think of measuring the time domain response at the MLP of the drum beat signal before and after the application of PEQ? This would give some more credence potentially to the premise in this thread from the Video that PEQ reduces tactile impact. I would have thought that they would not be as dramatically different as your "Overlay of the Impulse Responses" plot.

Also - IMO I think the overlays of Impulse and ETC might be a little misleading as the vertical scale may not be directly comparable between the two: if both impulse responses result in the same SPL response in the frequency domain, then the sum of the area under the ETC curve over the time window must be equal - the only way this could be true is if the initial rise of the Signal is much higher than that of the in-room MLP measured result.

Is there a way to corelate the REW Scope vertical axis measurements to actual SPL?

These are all fair questions and I wish I had an answer. The fact that we are talking about fundamentally different sources in different rooms would be complicated enough. If I were to simplify a bit and pretend that this is just an impulse signal of "that system in that room at that distance", vs. "this system in this room at this distance", and apply the same MTW window to both, further simplified with just mono "that" and both channels summed "this" You get the picture

Then the group delay of the kick drum over there would be this:

And group delay of my system would be this:

Time alignment across frequency is important for perceived "impact", perhaps this post of mine could be relevant:

PUNCH! How do we define it?

Basically, time alignment across frequency is a "thing" as far as the auditory system is concerned. Phase does matter too, at least for fundamental frequencies? I.e. in terms of "phase lock"? Of course, what people are trying to describe here would potentially be contained in signals with...

www.audiosciencereview.com

I don't know if this is helpful, but, as far as "Not all kick drums are the same" I highly recommend this paper: https://adamjhill.com/AJHILL/wp-content/uploads/2017/05/Hill-Hawksford-AES-130.pdf

 

[LPH]

My take: This the sort of argument that is made from the lack of understanding of how room modes work. Room modes are a phenomenon that appears below the transition frequency, when wavelengths are on the order of the dimensions of the room. In particular, the traveling wave analogy breaks down. Room modes, standing waves, resonances - are collective phenomena. One or two cycles of sound is irrelevant. A peak in the frequency response is not created by «a reflection». It is built up by several repeated cycles.


What is being said here?

janbth - you've made my point for me. "One or two cycles are irrelevant...it is built up by several cycles". What I'm saying is for concussive sounds like drums, guns, explosions,etc. the enegy in the signal does not last more that one or two cycles. The bulk of the impact of a kick drum, for example, lasts just 20ms - about the duration of a *single* cycle at 50Hz. The "standing waves" don't get enough time to "stand". If you EQ for standing waves occuring from several cycles building up, you do so at the expense of the kick drum.

 

[LPH]

I think what's largely missing from this discussion, is the duration of transient impact from explosive bass.
In the Hales interview, starting at about 34 min in he talks about this, saying 90% of a kick drums energy happens in 20ms.

A kick drum hit, or a big bass drop, are short term events, more full-range than not.
To the extent room-correction of modal behavior alters the steady state anechoic performance of the speaker,
it also alters the short term transient behavior of the speaker, the 20ms response, before modal behavior begins.
Transient gets sacrificed for steady state, imo/ime.

gnarly - you're exactly right

 

[LPH]

And then again, it may.
We would not EQ a Piano, but we will tune it. But not for a particular room, the tuner does not seat in the room when tuning, he seat at the piano. The room tuning is done at the room level, not the instrument.
So, what about speaker, it is tune as well, by the speaker designer, hopefully in an anechoic room or using a device mimicking an anechoic room, and hopefully not sitting in a room. Same as for the instrument, the room tuning should be done at the room level.
Many, here love applying EQ, not a bad idea, but not a good idea, in my opinion, if applied on a different recording basis, or even worst on a song basis. One who not detune a piano when changing player or music sheet. Or change acoustic of the room in the same circumstances.
EQ the room trough the speakers, is one way to do it, it works, but it also change the tune of the speaker, their is no way around it, is that a good idea? To me, it is not.
I want my speaker to play how they are designed to, if I buy a neutral speaker, I want to listen to a neutral speaker, not EQ to compensate for bad placement or inadequate room. Same I would not detune a Piano to make it sound "better"
If the room as an (many) issue, fix the room.
In these days of Digital Signal Processing, applying band aid seems favored over finding a cure, how ever strong and efficient the band aid is.
All that said, if any one like to adjust treble or bass on their equipment on a recording basis, to suit their taste, I have nothing against it, I just prefer not to and have not touch those in ages.

well said. the piano needs to be tuned to itself and not the room to sound right; the speaker needs to be tuned to itself and not the room, to sound right too.

 

[kemmler3D]

I think there's two things missing as we're discussing what happens to the impact of a single kick drum.

First, the transient might by short but the decay time of low frequencies in-room will not be as short. Bass tends to keep moving around the room for 250ms and often much longer. This is something that EQ helps with.

Second, it's rare to just have isolated drum hits in music for very long. Generally you're going to have other instruments playing and activating room modes too.

So the question becomes: is the masking from bass decay time and room modes in an un-EQ'd setup worse than the loss of impact from EQing the bass? I would hazard that in most rooms and systems it is. In large rooms that don't have bad peaks and with very dynamic speakers, maybe the trade-off flips, no idea really.

 

[Titurel]

well said. the piano needs to be tuned to itself and not the room to sound right; the speaker needs to be tuned to itself and not the room, to sound right too.

BTW, a piano needs to be tuned for the music that will be played on it (temperament). For example, the (historically accurate) tuning for Chopin is not the correct tuning for Bach nor for Bacharach. BUT, a particular room may not work well for any of them. Tuning the room to the piano is sometimes the more sensible solution or (in some contexts) picking the right piano for the room.

I believe tuning is an ill chosen term in this thread's context. It causes confusion (scrambled tropes) and it obscures the real question. Optimized is probably a better choice, but it raises further questions of optimized for what and for where. For example speaker Alpha, optimized to measure flat in acoustical environment Z, may not measure flat in acoustical environment Y. AND may not sound (be perceived as) flat in either. (10ms later, all bets are off.)

 

[Lars Risbo]

janbth - you've made my point for me. "One or two cycles are irrelevant...it is built up by several cycles". What I'm saying is for concussive sounds like drums, guns, explosions,etc. the enegy in the signal does not last more that one or two cycles. The bulk of the impact of a kick drum, for example, lasts just 20ms - about the duration of a *single* cycle at 50Hz. The "standing waves" don't get enough time to "stand". If you EQ for standing waves occuring from several cycles building up, you do so at the expense of the kick drum.

that argument is mathematically wrong: even the shortest possible pulse (delta pulse) triggers a mode to ring. it’s true tjat energy builds up if the source drives the mode at its mode frequency for longer time. the bottom line is that these high Q bass modes blurrs the envelope of transient signals like kick drums and plugged strings.

 

[Lars Risbo]

If you EQ for standing waves occuring from several cycles building up, you do so at the expense of the kick drum.

That is wrong. a correct EQ removes the envelope blurr.

 

[LPH]

That is wrong. a correct EQ removes the envelope blurr.

correct EQ where in the room? everywhere? no. so "correct EQ" for a spatially variant signal is nonsensical.

 

[kyuu]

correct EQ where in the room? everywhere? no. so "correct EQ" for a spatially variant signal is nonsensical.

For the listening position(s), obviously.

You really don't know what you're talking about here. I suggest you learn something about the topic rather than persisting in arguing from a position of ignorance.

 

[singular]

well said. the piano needs to be tuned to itself and not the room to sound right; the speaker needs to be tuned to itself and not the room, to sound right too.

I understand your points but it sounds more like your beliefs.

That is not how loudspeakers and pianos are designed to play music in enviroment.

You, and Cote Dazur, glued many things together,.

We tune the speakers and instruments to human ear first - ( the idea of flat ramp-like FR response is completely wrong in many ways ).
That is in fact your need to preserve "itself tune" by not EQ speakers.

We can discuss EQ for many different important FR range, and it will always be different approach for the same stereo set - with final better result than raw.

The thing is how well EQ is done to preserve "itself tune".

 

[LPH]

For the listening position(s), obviously.

You really don't know what you're talking about here. I suggest you learn something about the topic rather than persisting in arguing from a position of ignorance.

My knowing that a theater room, or any room for that matter, that has 4 or 6 or 8 seats requires different EQ settings to “correct” the different seats is fact. The soundfield within a room is chaotic and spacially variant. You feel differently and that’s ok. You believe that a single set of “room correction” filters can fix all seats simultaneously. Let’s agree to disagree. Now I will take my ignorance elsewhere and we can all be happy.

 

[singular]

My knowing that a theater room, or any room for that matter, that has 4 or 6 or 8 seats requires different EQ settings to “correct” the different seats is fact. The soundfield within a room is chaotic and spacially variant. You feel differently and that’s ok. You believe that a single set of “room correction” filters can fix all seats simultaneously. Let’s agree to disagree. Now I will take my ignorance elsewhere and we can all be happy.

You cannot listen to music on stereo set with other persons sitting around ( or wall behind, or head on large pillow ), and at the same time preserving "itself tune" of loudspeakers ( until it is included by designer ).

It is already in the envelope of what needs to be preserved.

PA is different than Home Stereo and both are different than Theater.

 

[Keith_W]

My knowing that a theater room, or any room for that matter, that has 4 or 6 or 8 seats requires different EQ settings to “correct” the different seats is fact. The soundfield within a room is chaotic and spacially variant. You feel differently and that’s ok. You believe that a single set of “room correction” filters can fix all seats simultaneously. Let’s agree to disagree. Now I will take my ignorance elsewhere and we can all be happy.

I think that is the point of contention - you have mischaracterized out position. We don't believe that. We agree that the sound field is spatially variant. We agree that EQ can not fix all seats simultaneously. But where we disagree is this: if you have more subs, you can fix more seating positions. That was shown by Olive and Welti in 2004. DSP and the science of psychoacoustics has gotten considerably more sophisticated since then, and these days you don't need as many subs. Or you can use the same number of subs to fix a larger area. This doesn't change the basic fact that more subs = larger area that can be optimized.

 

[SigbergAudio]

janbth - you've made my point for me. "One or two cycles are irrelevant...it is built up by several cycles". What I'm saying is for concussive sounds like drums, guns, explosions,etc. the enegy in the signal does not last more that one or two cycles. The bulk of the impact of a kick drum, for example, lasts just 20ms - about the duration of a *single* cycle at 50Hz. The "standing waves" don't get enough time to "stand". If you EQ for standing waves occuring from several cycles building up, you do so at the expense of the kick drum.

This is really very simple: Due to the room mode, the sound of the kick drum is too loud compared to the source material. The EQ does rob the kick drum of it's original impact, it removes excessive impact.

 

[SigbergAudio]

well said. the piano needs to be tuned to itself and not the room to sound right; the speaker needs to be tuned to itself and not the room, to sound right too.

I don't know a whole lot about piano tuning, but isn't that mainly about ensuring the correct pitch? EQ is not used to adjust the pitch of the music.

What you are saying about speakers is correct above the room transition frequency, but not below. A speaker clearly needs to be tuned to the room to sound right in the bass. The room is a part of the sound system, and the part that impacts the results the most in addition to the speakers themselves. Audible differences in any other system component are immaterial in comparison.

 

[Lars Risbo]

correct EQ where in the room? everywhere? no. so "correct EQ" for a spatially variant signal is nonsensical.

i was falsifying your previous statement. it’s true that such EQ is only correct for a single pair of speaker - listener positions at a time. By that does not make your previous statement true

 

[goat76]

My idea of how things work when it comes to bass frequencies in a room has always been the common one where the interaction between the loudspeakers and the room can be seen as one “unified” thing, making it possible to rely on an in-room measurement when adjusting the overall bass response. Still, the topic of this thread has got me thinking that the initial bass impact, which should be happening about the same time (without too much delay) as the transient response, may be affected by treating the direct sound and the room acoustics as the “same” thing.

I would like to see a study on how the bass impact is perceived in a room where some acoustic treatment has been used to address the bass problems, compared to the same room where the bass problems have been addressed using EQ. Will the sensation of bass impact in particular be perceived the same, granted that the bass sounds and looks equally flat in the measurements, or will the small “kick-back” delay that the room adds to the equation of the overall bass response be enough to lessen the sense of bass impact?