• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Stereo vs mono - what is the ideal frequency response?

hyperplanar

Active Member
Joined
Jan 31, 2020
Messages
177
Likes
304
Location
Los Angeles
#61
If so, why is then a perfectly flat speaker preferred in mono?
That’s a good point, I suppose that takes us into the psychoacoustical field a bit I guess. Considering the combination of room reflections and the brain’s processing the crosstalk cancellations might not be as pronounced as the plain steady state frequency response graphs would suggest. A third speaker in the center with a filter simulating the effects of crosstalk cancellation is likely to sound much worse than the left and right speakers playing a mono signal simultaneously, in my opinion.

So maybe the effect is not significant enough to manifest a pronounced difference in the mixing of songs and so a roughly flat speaker is preferred even when these originally stereo songs are played in mono. The preference scores have a good correlation but more work needs to be done IMO as to the correlation of preferences between multiple very competent loudspeakers (e.g. approaching +-1 dB flatness and good directivity) and their measurements. So many speakers do not approach this level in the first place so I feel the existing work on preference scores is more of a separating the wheat from the chaff type of thing, but it seems like a wash when all the speakers under test meet a basic level of competence. In other words the perceived FR deviations caused by the crosstalk and/or mixing engineers compensating for it may be dwarfed by the sins of the speakers themselves the vast majority of the time.

I’m not an expert in this area by any means so these are just a couple possibilities that came to mind. I’m also drunk and posting this in bed so sorry if I missed something obvious. It’s an interesting question!
 
Last edited:

March Audio

Major Contributor
Manufacturer
Joined
Mar 1, 2016
Messages
6,035
Likes
8,125
Location
Albany Western Australia
#62
That's what I also wrote in #3
This is the unfortunate reality of the circle of confusion :)

No standards. The studio monitors could be anything but neutral, their positioning, the room etc will all affect what the engineer mixes and tweaks. We then play it back on disperate systems with the same issues.
 
OP
T

Thomas_A

Addicted to Fun and Learning
Forum Donor
Joined
Jun 20, 2019
Messages
647
Likes
351
Thread Starter #63
Also, what about sounds that are panned left or right? Fully left or right wouldn't require any correction.
That is why you need a compromise; full compensation +/- 2 dB or so in a standard reflective room would only apply for the central phantom image (where quite much of the information is anyway).
 

March Audio

Major Contributor
Manufacturer
Joined
Mar 1, 2016
Messages
6,035
Likes
8,125
Location
Albany Western Australia
#65
That is why you need a compromise; full compensation +/- 2 dB or so in a standard reflective room would only apply for the central phantom image (where quite much of the information is anyway).
There is no appropriate compromise, for all the reasons/variables explained.
 

March Audio

Major Contributor
Manufacturer
Joined
Mar 1, 2016
Messages
6,035
Likes
8,125
Location
Albany Western Australia
#68
Again, the experiment is lacking.
You don't need an experiment to know that any correction is only applicable to a dead centre sound, when the listener is dead centre and facing straight ahead, for speakers and listener in a fixed geometry, in a specific room.

You simply can't, and shouldn't try to build this into the fundamental frequency response of a speaker.
 
OP
T

Thomas_A

Addicted to Fun and Learning
Forum Donor
Joined
Jun 20, 2019
Messages
647
Likes
351
Thread Starter #69
You don't need an experiment to know that any correction is only applicable to a dead centre sound, when the listener is dead centre and facing straight ahead, for speakers and listener in a fixed geometry, in a specific room.
I hear what you say, but I don't agree. Only a DBT and a preference scoring can solve the issue.
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
1,949
Likes
2,596
Location
NYC
#70
Now to the dilemma. If ideal speaker should be flat as judged from mono speaker playing in front of you, you will hear different timbral response if you place the same speaker in a stereo triangle. This is evident from previous research by Shirley, Toole and others. The conclusion from Toole that "stereo is flawed" (which is true) and one should go for multichannel instead. True multichannel recordings are however few, and you are still stuck with tons of ordinary stereo recordings. The real question is however avoided: Given that a flat response is preferred for a mono speaker - stereo speaker cannot show the same flat response if the tonality of the centre phantom image should be the same. So what is the ideal frequency response for speakers playing in stereo?

Binaural recordings beg to differ with this conclusion (even though I am with Toole on his statements in other parts of his paper about no such thing as "normal room responses" existing).

But I understand your whole post was talking about speakers, so obviously this little tidbit possibly doesn't apply.
 

March Audio

Major Contributor
Manufacturer
Joined
Mar 1, 2016
Messages
6,035
Likes
8,125
Location
Albany Western Australia
#71
I hear what you say, but I don't agree. Only a DBT and a preference scoring can solve the issue.
Then I'm not sure that you are understanding what I'm saying.

OK, let's say you need to make a cut of 2dB at 4kHz (arbitrary example and numbers) to correct the centre image timbre due to stereo issues. This means that any sound panned left or right that doesn't need a cut now has one.

Say you have made an eq correction for a specific speaker/listener placement, angle and geometry.

Now take the speakers move them to a different sized room with different acoustics and place them in a different geometry and distance from each other and the listener. Is the EQ still correct?

The problem is not an individual speakers frequency response, the problem is the use of two speaker stereo.
 
Last edited:
OP
T

Thomas_A

Addicted to Fun and Learning
Forum Donor
Joined
Jun 20, 2019
Messages
647
Likes
351
Thread Starter #72
Then I'm not sure that you are understanding what I'm saying.

OK, let's say you need to make a cut of 2dB at 4kHz (arbitrary example and numbers) to correct the centre image timbre due to stereo issues. This means that any sound panned left or right that doesn't need a cut now has one.
So here you can make a compromise; cut 1 dB at 4 kHz, the error is now 1 dB both for the central phantom image and the left-right panned sounds.

Say you have made an eq correction for a specific speaker/listener placement, angle and geometry.

Now take the speakers move them to a different sized room with different acoustics and place them in a different geometry and distance from each other and the listener. Is the EQ still correct?
As long as you put the speakers in the stereo triangle, you will have the effect. I am still talking about the optimal stereo setup here. Room problems are dealt with using other means. And if you cannot put yourself in the center of the stereo setup, as intended, whats the purpose of stereo? Tilting of the image in stereo when you move sideways is another problem and can only be reduced by other compromises, such as heavy toe in.

The problem is not an individual speakers frequency response, the problem is the use of two speaker stereo.
Yes, the dilemma is stereo reproduction, which I have been clear with from start of the thread. And it cannot be solved, only compromises can be made. The issue is what is the best compromise.
 

March Audio

Major Contributor
Manufacturer
Joined
Mar 1, 2016
Messages
6,035
Likes
8,125
Location
Albany Western Australia
#75
As long as you put the speakers in the stereo triangle, you will have the effect. I am still talking about the optimal stereo setup here. Room problems are dealt with using other means. And if you cannot put yourself in the center of the stereo setup, as intended, whats the purpose of stereo? Tilting of the image in stereo when you move sideways is another problem and can only be reduced by other compromises, such as heavy toe in.
compromise.
It will be a variable effect. Also affected by the room acoustics. So what is the correct EQ for the speaker manufacturer to apply?

Whilst I'm sure there are some people that put their head in a vice and listen in a perfect stereo triangle in a perfect room, the real world means this simply doesn't happen in many if not most circumstances.

Therefore the only sensible "compromise" is not to mess with the speakers fundamental on axis frequency response (assuming it's good - flat on axis anechoic with smooth of axis)
 
OP
T

Thomas_A

Addicted to Fun and Learning
Forum Donor
Joined
Jun 20, 2019
Messages
647
Likes
351
Thread Starter #76
It will be a variable effect. Also affected by the room acoustics. Therefore what is the correct EQ for the manufacturer to apply?

Whilst I'm sure there are some people that put their head in a vice and listen in a perfect stereo triangle in a perfect room, the real world means this simply doesn't happen in many if not most circumstances.

Therefore the only sensible "compromise" is not to mess with the speakers fundamental on axis frequency response (assuming it's good)
We are going in circles in the discussion without serious experiments made. We are also talking about on-axis deviations in the range of +/-1,5 dB or +/-1 dB. There are rather few speakers that match those numbers; a few of them vary with dips in the 2-4 kHz range, others with peaks in that range. The question then, is there a preference for one or the other, all other things equal? It is a simple experiment in theory, compare the same speaker with the Shirley and inverted Shirley frequency EQ curve, both conditions within +/- 1,5 dB, in stereo. Which one is preferred?
 
Joined
Apr 21, 2019
Messages
84
Likes
118
#77
Yes, been trough that thread a few years ago...
Oh, great. The included paper in the second post suggests that it may not be a simple matter of imposing the inverted Shirley curve: "2.3. Equalization Another way to fix the non-flat magnitude response caused by acoustic crosstalk is to apply inverse filters to the left and right signals [22]. However, the frequencies of the comb filter notches vary greatly depending on the relative positions of the speakers and listener. For example, the cancellation frequencies increase as the angle subtended by the speakers becomes narrower (such as when the listener moves further back) [9]. In addition, as the listener moves to the side and is no longer equidistant from the speakers, the notches move closer together and become different for each ear [12]. Without a good estimate of the relative positions, it would be impossible to accurately equalize the effects of the crosstalk." Here's the reference for 22: https://www.pearl-hifi.com/06_Lit_Archive/02_PEARL_Arch/Vol_16/Sec_53/AES 109/00042.pdf
 

Pluto

Addicted to Fun and Learning
Forum Donor
Joined
Sep 2, 2018
Messages
759
Likes
1,113
Location
Harrow, UK
#78
The first post in the thread referenced above was interesting because it singularly failed to provide the most obvious answer to the question it posed while getting very close indeed!
If a voice is panned from left to center to right, you'll hear a dip in tonality was the voices crosses the center
The question of “pan-pot law” has long been debated amongst studio people. The backroom techs say that a 6dB centre dip is correct because, when summed to mono, this keeps the level the same across the entire journey from L to R (assuming the overall law of the pot to be correct). However, most people who sit in front of the speakers will tell you that this, typically, results in a subjective dip in level at the centre. So, let's change the pan-pots to a 3dB law. In a decent room, this results in a slight bump in perceived loudness at the centre. Personally, I believe the less common 4½dB law to be about right although there's no doubt that this does vary from room to room.

So do bear in mind the solution to this issue is possibly a lot simpler than the route in which some of you appear to be heading – you need to decide what amount of simple reduction in volume is necessary to achieve the same apparent loudness when both speakers are driven equally compared to the level sent to one channel alone.
 
OP
T

Thomas_A

Addicted to Fun and Learning
Forum Donor
Joined
Jun 20, 2019
Messages
647
Likes
351
Thread Starter #79
Oh, great. The included paper in the second post suggests that it may not be a simple matter of imposing the inverted Shirley curve: "2.3. Equalization Another way to fix the non-flat magnitude response caused by acoustic crosstalk is to apply inverse filters to the left and right signals [22]. However, the frequencies of the comb filter notches vary greatly depending on the relative positions of the speakers and listener. For example, the cancellation frequencies increase as the angle subtended by the speakers becomes narrower (such as when the listener moves further back) [9]. In addition, as the listener moves to the side and is no longer equidistant from the speakers, the notches move closer together and become different for each ear [12]. Without a good estimate of the relative positions, it would be impossible to accurately equalize the effects of the crosstalk." Here's the reference for 22: https://www.pearl-hifi.com/06_Lit_Archive/02_PEARL_Arch/Vol_16/Sec_53/AES 109/00042.pdf
All agreed; it depends on the angle. Ideally a speaker or EQ is constructed with a given speaker- listener position.
 

KSTR

Addicted to Fun and Learning
Joined
Sep 6, 2018
Messages
532
Likes
1,231
Location
Berlin, Germany
#80
Simple (static) re-matrixing 2-ch content to 3 channels with a "trinaural" or "trifield" matrix solves the timbre problem very elegantly and fixes some other issues as well. It is not fully compatible with all recording techniques, notably HRTF-based phantom-source projection doesn't work anymore as the wave-field situation has changed.
Image sources are created in a better or more benign way, the center content is more discrete and the (usually less important) side content gets more virtualized, sort of.
 
Last edited:
Top Bottom