Value of room correction?

[Kal Rubinson]

The first minute or two sets up a couple of red herring arguments that led me to believe that he does not really understand REQ or he chooses to to mislead the viewer for his own reasons.

 

[QMuse]

I won't bog down the thread with a lengthy response. Wrt your last point, actually Dirac (and the other software I am referring to), does indeed correlate the phase response of the left and right speaker: "the latest Dirac Live 2.x algorithm might even slightly compromise on FR response of one of the speakers in favor of the best possible phase coherence between them that is important for imaging."

Nice to hear that. Feature liket hat would cetrainly come handy, especially in multisub scenario.

What hasn't been addressed and my point with a few commercial software being used is "excess phase" correction at low frequencies. I have tried MMM technique several times and it does not deal with excess phase at low frequencies. I have also tried MMM with rePhase for excess phase correction and is not the same, but closer, to the commercial software. What I am discussing isn't "magic." In this article, one can see in the time domain a maximum phase peak caused by low frequency room reflections, that are corrected in the time domain. It certainly is audible to my ears.

I used the term "magic" to allow for Dirac to get as good spatial information from 9 sweeps (per channel) as RTA MMM is getting with 50+ samples using some "magical" algorithm. I also said that such "magic" is simply not possible with a single sweep as it doesn't contain any spatial information.

When doing manual correction I'm using MMM RTA with pink noise as a basis for frequency correction. Once I fix that I take one or more sweeps (with vector averaging) to gather phase response which I then correct with rePhase.

While doing phase correction I'm also trying to get excess phase close to zero as possible without getting pre-ringing. As rePhase is a manual tool anything that commercial software does can be done with rePhase as well assuming you have a knowledge of how to do it properly.

Here is how phase response (violet) and excess phase (black) are looking for my left speaker:

And here is the overlay of both phaser responses:

Phase wiggles at 90-Hz, 190-220Hz and around 420Hz are due to less than optimal speaker positioning, but I managed to get them under control as unfortunately I cannot move my speakers.

 

[QMuse]

The first minute or two sets up a couple of red herring arguments that led me to believe that he does not really understand REQ or he chooses to to mislead the viewer for his own reasons.

As he's coming from the room/studio/churche acoustic treatment company I would put my money on the latter.

 

[mitchco]

@QMuse Yup, that is similar to what I get too using the same technique...

Curiously, remove the FDW setting, use REW's default 500 ms window and 1/12 oct smoothing and set the vertical scale (+- 180 degrees) and horizontal scale (10 Hz to 200 Hz) like in the screen shot below and lets see the phase response below 200 Hz. This is what mine looks like with large double 15" woofers in ported cab per side plus dual sealed 18" subs crossed at 46 Hz in a room that sucks, plus an asymmetric setup:

 

[QMuse]

@QMuse Yup, that is similar to what I get too using the same technique...

Curiously, remove the FDW setting, use REW's default 500 ms window and 1/12 oct smoothing and set the vertical scale (+- 180 degrees) and horizontal scale (10 Hz to 200 Hz) like in the screen shot below and lets see the phase response below 200 Hz. This is what mine looks like with large double 15" woofers in ported cab per side plus dual sealed 18" subs crossed at 46 Hz in a room that sucks, plus an asymmetric setup:

View attachment 52730

LOL Remove FDW, you say?

I'm from EU and walls in my appartment are made of 20cm thick hollow bricks filled with concrete while floor and ceiling are made of solid concrete re-inforced with iron net. I don't think you are aware of the differences in typical EU home vs USA home which affect reflections.

If I remove FDW this is what I get:

But I can always send you my uncorrected measurements so we can compare if one of your EQ toys can do a better job than my manual EQ.

 

[mitchco]

@QMuse As expected. And while room construction has some effect, at these wavelengths it is simply the dimensions of the room that define room modes. I wouldn't call commercial room correction software toys... but I forgive you

Let's have some fun and send me your uncorrected measurements and take this OT offline. I can't get to it right away, but will over time.

I think @Alexanderc already sees the value in room correction

 

[QMuse]

@QMuse As expected. And while room construction has some effect, at these wavelengths it is simply the dimensions of the room that define room modes. I wouldn't call commercial room correction software toys... but I forgive you
Let's have some fun and send me your uncorrected measurements and take this OT offline. I can't get to it right away, but will over time.

I wouldn't agree - density of wall material and its thickness are crucial with reflections. Dimensions of the room simply define the frequencies at which modes will happen but they don't tell you anything about the ratio of reflected vs absorbed sound which define how large the effects would be.

When I said "toys" I didn't mean it in any negative context.

Sure, thank you!

 

[Soniclife]

When doing manual correction I'm using MMM RTA with pink noise as a basis for frequency correction. Once I fix that I take one or more sweeps (with vector averaging) to gather phase response which I then correct with rePhase.

Interesting. You measure the frequency corrected output with sweeps, at the LP or near field, and does it matter.

 

[QMuse]

@QMuse As expected. And while room construction has some effect, at these wavelengths it is simply the dimensions of the room that define room modes. I wouldn't call commercial room correction software toys... but I forgive you

Interesting. You measure the frequency corrected output with sweeps, at the LP or near field, and does it matter.

I measure it at the LP as I need to correct phase at the LP. It does matter, as phase response at LP may, and typically will, differ at LP vs near field. As my LP is almost 4 meters from speakers it does matter as phase response is not the same as when measured near field.

 

[Snarfie]

Room correction is definitely always worth it but it can be done for much less than $1100.

I'm using Mathaudio room eq with a NAD C370 for FREE with foobar2000 did compare it to a Lyngdorf system cost a round 4000,- did exactly the same correction IMO. Suggest download it for free buy for less than 100,- a measuring mic & mic stand put it on a Mac OS or Windows OS an compare bypass & roomcorrection setup. Than decide to Stay with the used setup or use dirac Lyngdorf minidsp or other solutions. Despite a small investment in a measuring mic & mic stand there is 0 investment risk in trying room correction this way.

 

[NoMoFoNo]

To dumb it down for a moment for the more basic among us, how worthy (or not) is the room correction embedded in AVRs from Pioneer, Denon, Yamaha? How much more 'correct' correction can be had with external units and/or taking measurements from multiple positions?

 

[QMuse]

To dumb it down for a moment for the more basic among us, how worthy (or not) is the room correction embedded in AVRs from Pioneer, Denon, Yamaha? How much more 'correct' correction can be had with external units and/or taking measurements from multiple positions?

Here's my opinion on this.. When you listen to uncorrected response even with the speakers with superior spinorama measuremetns chances are you are listening to +/-15dB amplitude swings in the 20-300Hz range, make it +/-10dB if you are lucky - and that results in pretty much horrible sound quality.
Expect half of that trouble within 300-900Hz range.

Premium room EQ system, assuming well executed, can reduce that to +/-3db, but it is safe to assume every one of them will achieve +/-6dB, which is still a HUGE improvement over the uncorrected response.

You should be aware that every room Eq system is sensitive to "garbage in-garbage out" effect so following the instructions when taking measurements is essential, but the conclusion is that even the basic room EQ will significantly improve SQ of your system.

It is hard to estimate the improvement of AVR built-in room EQ systems vs the stand-alones alternatives - probably the latter are slightly better, but the important message here is that any room EQ system is BY FAR better than uncorrected response you're gettng with your speakers, and that is true no matter how superior their anechoic response looks.

 

[NoMoFoNo]

Excellent post and explanation @QMuse. Thanks. I've been pleased with my latest iteration of Pioneer AVR correction (MCACC Pro) and when I toggle between Pure Direct and Auto the sound seems (subjectively) much more balanced with the DSP applied. Reading threads like this makes me wonder about other/better solutions so I'll continue to follow and learn more.

Here's my poinon on this.. When you listen to uncorrected response even with the speakers with superior spinorama measuremetns chances are you are listening to +/-15dB amplitude swings in the 20-300Hz range, make it +/-10dB if you are lucky - and that results in pretty much horrible sound.
Expect half of that trouble within 300-900Hz range.

Premium room EQ system, assuming well executed, can reduce that to +/-3db, but it is afe to assume every one of them will achieve +/-6dB, which still a HUGE improvement over the uncorrected response.

You should be aware tha tevery room Eq system is sensitive to "garbage in-garbage out" effect so following the instructions when taking measurements is essential, but the conclusion is that even the basic room EQ will significantly improve SQ of your system.

It is hard to estimate the improvement of AVR built-in room EQ systems vs the stand-alones alternatives - probably the latter are stasticaly better, but the important message here is that any room EQ system is BY FAR better than uncorrected response you're gettng with your speakers, and that is true no matter how superior their anechoic response looks.

 

[Snarfie]

Looking to mine frequency response i can assure you that Room correction improved the new (found) sound dramatically. At least 70% for the better. No cable new amp or DAC could make this difference doesn't come close compared to Room EQ. At some points there is a difference of 15db you need a lot of absorbing or diffuser panels to over come this. In 50 years time i never encountered this dramatical improvement with a simple software solution as this. Basically buying an decent measuring mic was the best audio investment i ever did.

White line is my preferd/subjective target curve (flat) i did tried a lot Harman. B&K etc etc.

 

[Unclevanya]

The first minute or two sets up a couple of red herring arguments that led me to believe that he does not really understand REQ or he chooses to to mislead the viewer for his own reasons.

I have to agree. His comments about altering the intent of the artist/engineer are just completely off base. We have zero knowledge of the room the recording was done in, but we can generally assume the room is better treated than our homes and has less compromise. The room treatment advocate in him is missing the point that those treatments alter the sound heard also, if purity required listening unaltered then no treatments would be preferred to treating the room! But he seems to only apply purity to make the "anything but dsp is pure" argument conveniently forgetting that treatment of the room alters sound with a similar outcome. If we measure both and obtain similar results at the listening position - how we get there isn't important.

Also, let's be clear, electronic room correction as implemented typically isn't correcting the listening for a large group of listening positions. Physical room treatment can be more effective than the electronic correction in a church or other large room to correct multiple listening positions. The size and complexity of dsp processing and measurements and output sources that would be required to render many hundreds of listening positions corrected is far more expensive and challenging. But in our homes with one or two people listening in a set position, dsp becomes more accessible and easier to A/B compare than direct treatments - not to mention more WAF friendly.

 

[SIY]

I have to agree. His comments about altering the intent of the artist/engineer are just completely off base.

Playing that video where my wife could hear it was a hilarious experience. "What the f--- is that guy talking about? Is he an idiot?" over and over.

Come to think of it, we were near the kitchen.

 

[Igor Kirkwood]

[QUOTE = "Unclevanya, poste: 345400, membre: 12439"] Je dois accepter. Ses commentaires sur la modification de l'intention de l'artiste / ingénieur sont tout simplement hors de propos. Nous n'avons aucune connaissance de la pièce dans laquelle l'enregistrement a été fait, mais nous pouvons généralement supposer que la pièce est mieux traitée que chez nous [/ CITATION]
En cas d'enregistrement dans une église ou une grande salle de concert, celà n'a pas grand sens

 

[Unclevanya]

I'm sorry my French is tres mal. But if I read you correctly I'm confused. I know very specifically that many churches use sound treatment to improve the acoustics - so I'm not following what you mean. Apologies if I misread you.

 

[Igor Kirkwood]

[QUOTE = "Unclevanya, post: 345400, member: 12439"] Nous n'avons aucune connaissance de la pièce dans laquelle l'enregistrement a été fait, mais nous pouvons généralement supposer que la pièce est mieux traitée que chez nous et a moins de compromis. .[/CITATION]

Désolé j'essaye l'anglais!

En Europe, nous n'avons jamais de traitement acoustique dans l'église

Il est difficile mais intéressant de comparer le traitement d'une salle d'écoute et d'un lieu d'enregistrement.

Si vous avez, dans votre salle d'écoute, un bon RT 60 de 0,2 ou 0,3; pour un enregistrement acoustique le son sera mauvais pour RT 60 de 0,2 ou 0,3

 

[Unclevanya]

Your English beat my French. Most chicken in Europe are old and have excellent acoustics built in. Essentially sound treatment was a criteria in the design. The modern churches in the USA are often built in a way that requires drastic sound modifications and these are built into the design as well but using different materials and designs.

My main point was that dsp approaches are sufficient for the small complexity we have in our rooms and the perspective presented in the video might be influenced by this large scale sound engineering rather than the small scale.