Room correction/DSP

MartinC

Wammer
Wammer
Jul 29, 2005
8,793
4,984
158
HiFi Trade?
  1. No
Just skimmed the thread...

In the low bass region it's not obvious whether it's better to EQ to the sum of left and right, or each individually. The reason this is complicated is if you EQ each channel separately to be smooth, the sum almost certainly won't be due to phase differences. Most (but not all) low bass is played on both channels together so EQing the sum is probably what I'd go for if I had to guess.

The above dilemma is one I avoid by using a subwoofer fed the sum of left and right channels  :) .

Higher up the frequency range you would want to EQ the speakers separately though.

 
Last edited by a moderator:

lindsayt

Wammer
Wammer
Dec 19, 2009
616
382
78
Huddersfield, , Unit
May I suggest that you try EQ'ing only up to 500Hz an give it a listen?

To my eyes it looks like the upper-midrange and treble are somewhat exaggerated.

It's definitely a matter of preference but, unless you're listening nearfield, the response should show a tilt like this:

There's a paper that features that chart:

https://www.bksv.com/media/doc/17-197.pdf

The paper is mildly interesting reading.

In reference to that "optimum" curve, the paper said that for recordings made using farfield microphone positioning, it should be absolutely flat.

For recordings made with a mixture of nearfield and farfield microphone positioning, the curve should be as per that diagram.

That's fine for acoustic live recordings. Classical concerts.

What about rock and pop studio recordings? Where we are talking about very nearfield microphone positioning? Or electronic instruments plugged straight into the recording equipment. And where multi-track recordings are mixed down to 2 channel.

Also what about Fletcher Munson? Different playback volumes resulting in different perceived frequency responses.

I have no idea if my optimum curve is flatter than the B&K one, or if it is more steeply sloped, or if it is a more complex curve.

I have a feeling that a lot of pop music is mixed with a B&K curve built in. And it therefore seems daft to apply this curve a 2nd time.

 
  • Like
Reactions: flak monkey

flak monkey

Weirdo Mosher Freak
Wammer Plus
Dec 24, 2010
1,032
653
193
Norfolk, UK
AKA
Sera
HiFi Trade?
  1. Yes
Took a completely different approach this morning. Using full spectrum pink noise, generated by REW. Saved onto my media server so I could do proper before/after measurements.

I did check and the overall response shape is the same as using a sweep, or as near as makes no difference. Also, less smoothing on this than in previous charts.

Red is pre-EQ and blue is post-EQ. Took @tuga advice and applied more of a slope to the target EQ. Still boosted the HF a little, but that's entirely my personal preference.

Again this is both channels, I have no means by which to mess around with individual channel EQ at this point.

Before and after EQ REW.JPG

I think you are doing things in the right way and adding in your view what is needed then listening and adjusting accordingly . As it is your music with your system in your room then it has to please you and not anyone else's idea of what is right . Never had or used Roon so not sure what is and is not available . I use DIRAC Live and i use it across the full spectrum I tweak the filter from the one it suggests and listen as I alter things . I found it very helpful for me to make some really huge changes and listen to what they did (which was of course make it sound wrong and weired) then slowly bring the filter back up and listen to the music as it slowly changes . This gave me a bit of an idea of what changing each part sounded like and has enabled me to know very closely exactly what filter curve I like and suits my room . Not scientific at all I know but for me at least it works .
I like to use measurement as a starting point and then adjust to suit personal taste from there. :)  Partially scientific, partially subjective, but I prefer that method and it works for me!

Just skimmed the thread...

In the low bass region it's not obvious whether it's better to EQ to the sum of left and right, or each individually. The reason this is complicated is if you EQ each channel separately to be smooth, the sum almost certainly won't be due to phase differences. Most (but not all) low bass is played on both channels together so EQing the sum is probably what I'd go for if I had to guess.

The above dilemma is one I avoid by using a subwoofer fed the sum of left and right channels  :) .

Higher up the frequency range you would want to EQ the speakers separately though.
I also use a sub to boost below 40Hz, where my AE2 start to drop off quickly. 

There's a paper that features that chart:

https://www.bksv.com/media/doc/17-197.pdf

The paper is mildly interesting reading.

In reference to that "optimum" curve, the paper said that for recordings made using farfield microphone positioning, it should be absolutely flat.

For recordings made with a mixture of nearfield and farfield microphone positioning, the curve should be as per that diagram.

That's fine for acoustic live recordings. Classical concerts.

What about rock and pop studio recordings? Where we are talking about very nearfield microphone positioning? Or electronic instruments plugged straight into the recording equipment. And where multi-track recordings are mixed down to 2 channel.

Also what about Fletcher Munson? Different playback volumes resulting in different perceived frequency responses.

I have no idea if my optimum curve is flatter than the B&K one, or if it is more steeply sloped, or if it is a more complex curve.

I have a feeling that a lot of pop music is mixed with a B&K curve built in. And it therefore seems daft to apply this curve a 2nd time.
This is very interesting thanks. I will give the paper a proper read later.

 
  • Upvote
Reactions: tuga and lindsayt

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No
In reference to that "optimum" curve, the paper said that for recordings made using farfield microphone positioning, it should be absolutely flat.
I think that you misread, from my understanding they are referring to recordings not speakers:

When music is recorded under farfield conditions, it will contain a suit­ able mixture of direct and reflected sound, and the curve ought to be ab­solutely flat in that case. This is true for recordings, for instance, made with two B &K condenser mi­crophones in the farfield.

The axial anechoic response, usually measured at 1 metre, should indeed be flat.

.

Harman as also researched this subject and came up with its own curve:

bGMWcjI.png


.

And this is from the European Broadcasting Union:

https://tech.ebu.ch/docs/tech/tech3276.pdf

H9oBXpE.png


 
Last edited by a moderator:
  • Upvote
Reactions: flak monkey

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No
Most (but not all) low bass is played on both channels together so EQing the sum is probably what I'd go for if I had to guess.
Surely the balance between right and left will depend on the recording. Only when the instrument is reproduced in the centre will both speakers be reproducing the same "amount" of bass.

With most orchestral recordings the double bass section will be perceived as located very close to or actually on the right speaker (not my image).

noTWzsI.jpg


Below Schroeder we correct the speaker response in the room (at the listening spot); when the response if different for both speakers we correct each of them individually.

 

MartinC

Wammer
Wammer
Jul 29, 2005
8,793
4,984
158
HiFi Trade?
  1. No
Surely the balance between right and left will depend on the recording. Only when the instrument is reproduced in the centre will both speakers be reproducing the same "amount" of bass.

With most orchestral recordings the double bass section will be perceived as located very close to or actually on the right speaker (not my image).
Two thoughts:

  1. Other types of music are available  :) .
  2. I'm not sure what sort of microphone techniques are typically used for live classical recordings but I wouldn't be surprised if the low bass (let's say < 100 Hz) is generally of a pretty equal level on both recorded channels? Note that any localisation of where double-basses will be down to higher frequency content (harmonics or higher notes). Human hearing is not capable of localising the frequency range I've been focusing on. 

Below Schroeder we correct the speaker response in the room (at the listening spot); when the response if different for both speakers we correct each of them individually.
You might but there is no universal 'we' here. 

For info. I believe* Dirac Live factors in both the individual and summed response although quite how the balance is weighted when deciding on what adjustments to apply I have no idea.

*Edited as checking quickly I can't actually find anything to prove this. I thought it was referred to in a changelog but I can't find it. Possibly I'm mistaken on this this point...

 
Last edited by a moderator:

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No
Other types of music are available
Ok. Mono. Two speakers transduce exactly the same signal. One speaker has a different response below 200Hz than the other. Won't correcting the speakers individually produce a flatter response at the listening spot?

 

flak monkey

Weirdo Mosher Freak
Wammer Plus
Dec 24, 2010
1,032
653
193
Norfolk, UK
AKA
Sera
HiFi Trade?
  1. Yes
Ok. Mono. Two speakers transduce exactly the same signal. One speaker has a different response below 200Hz than the other. Won't correcting the speakers individually produce a flatter response at the listening spot?
They also react to each other with the standing waves in the room

From what little I understand there are benefits to EQ on each speaker and both, the interaction of both as sound waves meet at the listening position shouldn't be overlooked.

 

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No

MartinC

Wammer
Wammer
Jul 29, 2005
8,793
4,984
158
HiFi Trade?
  1. No
Ok. Mono. Two speakers transduce exactly the same signal. One speaker has a different response below 200Hz than the other. Won't correcting the speakers individually produce a flatter response at the listening spot?
Let me restate my point. If you EQ the amplitude response of each speaker individually to be flat but they have different phase responses, then the sum will not be flat. Your mono signal would be the summed situation. For signals that are predominantly 'mono' you are therefore arguably better off EQing the summed signal.

I would only consider EQing the sum in the region where the sound could not be localised though, not at 200 Hz.

 

newlash09

newlash09
Wammer
Aug 10, 2018
2,856
2,399
148
India
AKA
Y.Manohar
HiFi Trade?
  1. No
That interaction, called the crosstalk dip, takes place much higher in the audio band:



source: https://www.amazon.co.uk/Sound-Reproduction-Psychoacoustics-Loudspeakers-Engineering/dp/113892136X/
Have you tried dsp or drc @tuga :)

Atleast @flak monkeyshared her half boiled attempts at amalgamating the effects of her room. I do not debate that you are very knowledgeable.  But if you haven't given it a shot yourselves. Then please give one....so that you can be more condescending of others efforts to improve their sound, having tried atleast once :)

Posts like yours which look, like they are being posted from the imperial library, sitting atop a exhalted seat of all knowingness...without having tried anything in person...would probably discourage others to be open with their trails and experiments. Just my 2 cents....and if you've infact tried DSP or DRC, I would be all ears to know about that :)

 
Last edited by a moderator:
  • Upvote
Reactions: lindsayt

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No
Have you tried dsp or drc @tuga :)

Atleast @flak monkeyshared her half boiled attempts at amalgamating the effects of her room. I do not debate that you are very knowledgeable.  But if you haven't given it a shot yourselves. Then please give one....so that you can be more condescending of others efforts to improve their sound, having tried atleast once :)

Posts like yours which look, like they are being posted from the imperial library, sitting atop a exhalted seat of all knowingness...without having tried anything in person...would probably discourage others to be open with their trails and experiments. Just my 2 cents....and if you've infact tried DSP or DRC, I would be all ears to know about that :)
My first attempt was the KRK DAC with RoomPerfect in 2009. I wasn't impressed.

Have been making my own filters with REW and using them in HQPlayer since 2016.

It's not my fault if people don't (enjoy or want to) read about audio...

 

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No
Have you tried dsp or drc @tuga :)

Atleast @flak monkeyshared her half boiled attempts at amalgamating the effects of her room. I do not debate that you are very knowledgeable.  But if you haven't given it a shot yourselves. Then please give one....so that you can be more condescending of others efforts to improve their sound, having tried atleast once :)

Posts like yours which look, like they are being posted from the imperial library, sitting atop a exhalted seat of all knowingness...without having tried anything in person...would probably discourage others to be open with their trails and experiments. Just my 2 cents....and if you've infact tried DSP or DRC, I would be all ears to know about that :)
Also I do my best to help people, and that includes you. It's disappointing to read your comment.

That includes telling them things they don't want to hear instead of patting their backs.

 

newlash09

newlash09
Wammer
Aug 10, 2018
2,856
2,399
148
India
AKA
Y.Manohar
HiFi Trade?
  1. No
Also I do my best to help people, and that includes you. It's disappointing to read your comment.

That includes telling them things they don't want to hear instead of patting their backs.
It is about the way you put it across sir :D

 
Last edited by a moderator:

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No
Let me restate my point. If you EQ the amplitude response of each speaker individually to be flat but they have different phase responses, then the sum will not be flat. Your mono signal would be the summed situation. For signals that are predominantly 'mono' you are therefore arguably better off EQing the summed signal.

I would only consider EQing the sum in the region where the sound could not be localised though, not at 200 Hz.
I don't ever measure both speakers at the same time or we could test your hypothesis. From which source did you get the idea that the two speakers should be measured at the same time?

 

newlash09

newlash09
Wammer
Aug 10, 2018
2,856
2,399
148
India
AKA
Y.Manohar
HiFi Trade?
  1. No
Also I do my best to help people, and that includes you. It's disappointing to read your comment.

That includes telling them things they don't want to hear instead of patting their backs.
As I said.. you definitely are knowledgeable sir :)

But you don't have the art of masking it :D

Your posts, though being well intended, from your perspective. Always reek of superiority to me :D

I would rather go by the advise of someone who has tried stuff, and will speak from their experience in a humble manner. Than your incessant posting of a plethora of graphs to try and prove a point sir.

 
Last edited by a moderator:
  • Upvote
Reactions: rdale

tuga

. . .
Wammer
Aug 17, 2007
12,575
5,080
173
Oxfordshire, UK
AKA
Ric
HiFi Trade?
  1. No
As I said.. you definitely are knowledgeable sir :)

But you don't have the art of masking it :D

Your posts, though being well intended, from your perspective. Always reek of superiority to me :D

I would rather go by the advise of someone who has tried stuff, and will speak from their experience in a humble manner. Than your incessant posting of a plethora of graphs to try and prove a point sir.
So you feel that your comments are less agreesive and ad hominem by adding a dumb :D

Good for you.

Yes I use graphs to prove and also to illustrate a point. I also find it strange that people would rather read reviews instead of learning about audio. Strange world...

 

newlash09

newlash09
Wammer
Aug 10, 2018
2,856
2,399
148
India
AKA
Y.Manohar
HiFi Trade?
  1. No
So you feel that your comments are less agreesive and ad hominem by adding a dumb :D

Good for you.

Yes I use graphs to prove and also to illustrate a point. I also find it strange that people would rather read reviews instead of learning about audio. Strange world...
I wouldn't want to give the notion that my posts are less aggressive in anyway. Though I would admit, that I think you are someone, you hasn't tried or couldn't try enough stuff. So has finally sought solace in measurements.