Do you correct above Schroeder/transition frequency? If yes, how?

[thcdru2k]

I'm interested in how people decide the upper correction limit for room correction. Above the Schroeder/transition how do you distinguish between room correction, speaker correction, L/R matching, and mic position artifacts.

How high do you allow correction (up to 500 Hz, up to 1khz, or full range)? Above transition do you use broad shelves/tilt only or detailed correction? Do you try to preserve the speaker's response or do you correct it fully for a neutral response? Does your answer change for music production vs listening? What sounds best to you?

 

[Curvature]

For room measurement, I use the moving mic method or a 2-cycle frequency-dependent window per speaker.

For filters, I use low Q bandpass or shelves.

I check speaker measurements to assess directivity and make sure the filters won't produce strange results, and then listen to dial in a tilt. I generally leave it alone after that. I don't like the habit of adjusting EQ for different music.

 

[YouDontHaveEnoughBass]

No, I've got a filter to bring down a bump in the speakers on-axis response at 500hz but that's using anechoic data; the only room-correction filter I have is a room mode at 46hz (and even then I only cut 3db I don't try to chop the peak entirely off the graph, doing so makes the system sound hollow)

I've tried all sorts of room correction but most filters do more harm than good in most cases I've tried; I also value having the speakers sound good in a broad area and when you actually take measurements across a broad area and check peaks for stability across the lot you'll generally find most aren't stable across the area so shouldn't be fixed

This is for a living room hifi system; if you're sat at a desk your mileage may vary

 

[Snarfie]

I'm interested in how people decide the upper correction limit for room correction. Above the Schroeder/transition how do you distinguish between room correction, speaker correction, L/R matching, and mic position artifacts.

How high do you allow correction (up to 500 Hz, up to 1khz, or full range)? Above transition do you use broad shelves/tilt only or detailed correction? Do you try to preserve the speaker's response or do you correct it fully for a neutral response? Does your answer change for music production vs listening? What sounds best to you?

A lot is discussed in this topic.

Should we correct to Schroder, or full range?

I would like to gather opinions from ASR members as to whether we should apply room correction to the Schroder frequency, or whether we should do full range correction from 20Hz - 20kHz. There appear to be two schools of thought: - Correction to Schroder: aims to only correct bass peaks and...

www.audiosciencereview.com

 

[gnarly]

I feel the track by track tonality differences are strong enough to require the occasional use of high-shelving or an equivalent form of EQ.
So room EQ above Schroder is kinda a secondary consideration.

As far as any form of permanent, higher frequency room EQ....if room sounds off after work below Schroder...
I first like to get decay time a constant .3-.4 across the mid & upper spectrum. Then if the room is still tonally off, I will put in some permanent EQs that remain separate from the track-by-track tone control EQs mentioned above.

Or I just say screw room correction totally...I'm going to always have low and high tone controls ....they and automatic brain-to-room-adjustment is all I need.

 

[Ultrasonic]

I do in my current setup. I used time-gated measurements to isolate the direct signal from the speakers and have then applied EQ to reduce a frequency response dip at the tweeter/mid crossover frequency, and slightly flatten the response at higher frequencies. This is importantly using measurements at the same distance and height from the speakers as my listening position is.

I have tried using Dirac Live at higher frequencies but don't like the results and so only use this below 200 Hz currently. Some like this though, and I guess it may vary between rooms how well the averaging across multiple points helps to reduce the room influence at higher frequencies and so make this approach more viable.

 

[JohnnyNG]

I use Dirac Live below 500Hz (so a bit above Schroeder I guess) which covers my midbass speakers and sub. Outstanding.

 

[thewas]

Usually below approximately 300-500 Hz I correct based on moving mic measurements around the listening position separately for the left and right channel avoiding high Q positive filters and above that on a kind of listening window which is based on MMM and/or not too long gated measurements around 1-2 meters in front of the loudspeakers.

 

[Duke]

No, I've got a filter to bring down a bump in the speakers on-axis response at 500hz but that's using anechoic data;

^^^This^^^

Above Schroeder any EQing should be intelligently based on anechoic data. That doesn't mean you never make trade-offs between the direct sound and the off-axis sound, but it does mean you do so wisely, taking into account everything you can.

 

[john61ct]

Good points here too

Room Equalization: From Brute-Force Flattening to Perceptual Optimization

I worked hard on a room EQ system recently and I summarised what I learnt in this article. I tried to explain concepts as simply as possible with references when the reader wants to understand in more details. All feedback and corrections are welcome. The software that implements the room EQ...

www.audiosciencereview.com

A Light Touch is always emphasized by those members whose DSP assertions I pay most attention to

 

[Scott Borduin]

In my primary system, below Schroeder only. But that is not a hard and fast rule. My second home system has a couple of small Genelecs sitting at the front of a low table, 30" or so from the wall. There are resulting boundary reinforcement peaks around 440 and 1000 hz, which are quite audible and also need EQ. Not all significant resonances are modal.

 

[NTK]

I recently came across this interview of Tomlinson Holman.

Tomlinson Holman Interview | Life, Career, and the Future of Sound

Tomlinson Holman, film theorist, audio engineer, and inventor of Lucasfilm THX sound system, takes us on a journey through his life...

sound.krotosaudio.com

In the interview, Holman said this, which seemingly directly contradicted the commonly held ASR belief. As this is Holman, who in my book when says something related to audio, automatically deserves serious considerations, he got me thinking.

The 60+ years old BBC report Holman quoted (its link and a closely related one are below) were about L/R channel differences. I don't think a lot here will argue against L/R speaker matching. However, what if the room is not perfectly symmetrical and there are differences in the early reflections of the different channels? Do they and how much do they matter? Early vs not so early and how much the reflection delay time matters? Is EQ an appropriate remedy? I don't have answers to any of these. Just Holman got me thinking.

https://downloads.bbc.co.uk/rd/pubs/reports/1964-67.pdf (On L/R channel amplitude differences)
https://downloads.bbc.co.uk/rd/pubs/reports/1964-08.pdf (On L/R channel phase differences)

 

[krabapple]

Sorry, is he saying someone measured 32,000 room curves in situ at the MLPs in the 1960s?

 

[thcdru2k]

I recently came across this interview of Tomlinson Holman.

Tomlinson Holman Interview | Life, Career, and the Future of Sound

Tomlinson Holman, film theorist, audio engineer, and inventor of Lucasfilm THX sound system, takes us on a journey through his life...

sound.krotosaudio.com

In the interview, Holman said this, which seemingly directly contradicted the commonly held ASR belief. As this is Holman, who in my book when says something related to audio, automatically deserves serious considerations, he got me thinking.
View attachment 539322

The 60+ years old BBC report Holman quoted (its link and a closely related one are below) were about L/R channel differences. I don't think a lot here will argue against L/R speaker matching. However, what if the room is not perfectly symmetrical and there are differences in the early reflections of the different channels? Do they and how much do they matter? Early vs not so early and how much the reflection delay time matters? Is EQ an appropriate remedy? I don't have answers to any of these. Just Holman got me thinking.

https://downloads.bbc.co.uk/rd/pubs/reports/1964-67.pdf (On L/R channel amplitude differences)
https://downloads.bbc.co.uk/rd/pubs/reports/1964-08.pdf (On L/R channel phase differences)

None of my rooms are perfectly symmetrical and I see significantly different measured responses from speakers even when they are swapped. He brings up a good point about phantom center. Having a true phantom center measurement wise in most rooms seems challenging also makes me wonder where is the line where the result becomes acceptable or how much correction is necessary.

 

[NTK]

Sorry, is he saying someone measured 32,000 room curves in situ at the MLPs in the 1960s?

I wish Holman had said more, especially on how to EQ above Schroeder and based on what measurements and criteria. In the interview, he said "we measured over 32,000 responses", so I'd assume those were his and his company's measurements.

 

[Matias]

On my HT LCR I measured REW + UMIK-1 MMM, and corrected above 500 Hz with minimalist 4 band independently per speaker, using conservative PEQ gains and Qs, nothing too drastic. But it makes all the difference IMO.

Sure, better yet would have been replacing my LCR speakers with top measuring speakers like AsciLab, but...

 

[10PSPMemSticksForMiku2007]

The 60+ years old BBC report Holman quoted (its link and a closely related one are below) were about L/R channel differences. I don't think a lot here will argue against L/R speaker matching.

The thread title is about room correction, but as I supposed, people are also talking about speaker correction here. Of course you should equalize your speaker if there is an imperfection. For example, if you have an metal tweeter that breaks up around 15 khz you might correct that with DSP.

Not that I'll tell you never to do room correction in higher frequencies either. It's just DSP, so try whatever you want.

 

[LTig]

I don't, as my speakers (studio monitors) are very neutral. Upper limit is about 300 Hz.

 

[Keith_W]

In the interview, Holman said this, which seemingly directly contradicted the commonly held ASR belief. As this is Holman, who in my book when says something related to audio, automatically deserves serious considerations, he got me thinking.

There is a difference between "WE measured 32,000 loudspeakers in homes" and "we have 32,000 USER measurements of loudspeakers in homes". I think the former sounds VERY unlikely, he would have to be extremely well resourced to do that.

And if it's 32,000 USER measurements ... you run into all the usual problems. Uncalibrated / poorly calibrated microphones, poor measurement technique, poor signal-noise, incorrect measurement volume, etc. I have examined dozens of measurements posted on ASR. I usually compare them to the anechoic measurements if they are available. I can tell you there is a lot of variance from home measurements taken by unskilled users.

Anyway I have no doubt that some companies have thousands of user measurements on their database, especially companies like Dirac, Audyssey, and Genelec.

 

[Duke]

I recently came across this interview of Tomlinson Holman.

Tomlinson Holman Interview | Life, Career, and the Future of Sound

Tomlinson Holman, film theorist, audio engineer, and inventor of Lucasfilm THX sound system, takes us on a journey through his life...

sound.krotosaudio.com

In the interview, Holman said this, which seemingly directly contradicted the commonly held ASR belief. As this is Holman, who in my book when says something related to audio, automatically deserves serious considerations, he got me thinking.
View attachment 539322

Well, OF COURSE if you do "room EQ" to speakers, the in-room frequency response measurement (at the microphone location) will be improved!

But that's fixing the wrong problem, ESPECIALLY if it degrades the first-arrival sound! I'm not saying there are NEVER situations where EQ north of Schroeder makes a worthwhile net improvement while simultaneously degrading the first-arrival sound, but that juggling of tradeoffs should be done with a fairly complete knowledge of the off-axis response. The in-room response curve alone is an inadequate metric.

I don't think a lot here will argue against L/R speaker matching. However, what if the room is not perfectly symmetrical and there are differences in the early reflections of the different channels? Do they and how much do they matter? Early vs not so early and how much the reflection delay time matters? Is EQ an appropriate remedy?

I would never use EQ to deliberately introduce a discrepancy between the first-arrival sound of the left and right speakers, but I can see using EQ to CORRECT a small discrepancy in the first-arrival sound of the two.