Well I suppose the shallow vs. steep argument will just go on and on

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
when that spice is spaciousness then for me it's better to have spaciousness that is samey because it comes from the room reflections - which happens when the recording itself is dry - than to have no spaciousness

because spaciousness is a requirement of realism of sound reproduction
So to get realism of sound, you want to add artificial spaciousness ? I see a contradiction here somehow... ;)
 
Member
Joined 2008
Paid Member
After some recent experiences I'm beginning to wonder if a good part of "You are there" vs "They are here" is as simple as specific differences in the frequency response, particularly in the presence region...

We know that a large part of our ability to localise sound comes from the subtle HRTF induced changes in frequency response.

This means that frequency response aberrations of the speaker that happen to coincide closely with actual valid HRTF localisations can have a strong influence on where the sound is perceived to emanate in relation to the actual speakers.

For example an emphasis at 10Khz can shift the apparent source location of the image vertically upwards so that it floats in the air considerably above the speakers actual location, because an actual sound source at that higher position would have an HRTF that emphasises those frequencies. This is an effect I've observed many times.

Another example is that an emphasis at 1Khz can cause confusion about whether a sound is directly in front of you or behind you, as an emphasis at that frequency corresponds to a sound directly behind you in a HRTF. Conversely a deliberate dip around 1Khz can more strongly (artificially) focus the source of a sound to the front. (Funnily enough, many drivers have a surround dip in their response at about 1Khz...)

Recently I've been rebuilding and actually finishing some perpetually never finished speakers whose drivers I've been using in various configurations for years, previously the tweeter was sitting on the top separately so I could adjust the acoustic offset, and they were using a fairly basic 3rd order crossover.

These speakers have had outstanding "They are here" imaging performance in past configurations where the image literally hovers in front of the speakers, so I know the drivers are capable of it. In the new configuration the tweeters are now flush mounted on the baffle and as a result the acoustic centre of the tweeter is significantly in front of the other driver, by approx 35mm.

Predictably the performance through the crossover region is poor now with the old simplistic crossover resulting in a significant response depression from about 3-5Khz and poor phase tracking through the crossover.

I've been designing a new crossover from the ground up for the new driver positioning which will resolve this and other issues and I'm currently waiting for parts to arrive so I can build it, meanwhile I've been making do with the old crossover with a bit of EQ, which has been quite informative.

In particular I notice that the "They are here" imaging of recordings I'm familiar with is largely absent, even with a bit of EQ to boost the presence region. So it's pretty clear that the "They are here" illusion is critically dependent on the response through presence region from about 3-5Khz.

It doesn't sound terrible in fact it sounds quite pleasant, clean and laid back on most recordings but just doesn't have that presence and image floating in the air quality that I miss and used to get, and you can definitely tell the drivers aren't integrating well.

On the other hand I have many other recordings which have a strong sense of "You are there" imaging in them (some acoustic, but many with "artificial" studio generated soundscapes) and they sound just fine despite the problems in the presence region.

It sounds like the source of the sound field is beyond the wall behind the speakers, and doesn't appear to originate from the speakers or within the room at all. There is a strong sense of being in a large environment with the sources of sound being quite some distance from you.

From prior experimentation I know that this sense of space and the sound coming from behind the speakers wall is partly due to ambience in the recording, but a good part of that perception of depth comes from lower frequencies in the 150-400 Hz range, approximately.

Holes or lack of response in that region can destroy the illusion of depth in the recording while a speaker that reproduces that region well, (and that includes dealing with problems like floor bounce cancellations) can produce a convincing sense of depth on these kind of recordings.

Both the 3-5Khz and 150-400Hz regions have their own challenges in speaker design, 3-5Khz is often at or near a crossover frequency so good driver integration (including good phase tracking) and crossover design plays a major part in whether the speaker can do "they are here" imaging well or not.

But 150-400Hz is also quite challenging, especially for 2 ways where floor bounce cancellation and room vertical mode cancellation can be major issues that put holes in that region, especially around 150Hz.

A well designed 3 way system with a low mounted woofer and carefully chosen crossover frequency can do better here, and a wide baffle system will usually handle this region a lot better than a small baffle mini monitor style design as well. Getting the baffle step compensation wrong can easily ruin that depth illusion by bringing this lower midrange region out of balance.

What my long rambling post is suggesting is that "They are here" and "You are there" are two different possible illusions in a recording, and that they are both largely frequency response related but in different parts of the spectrum - one at low frequencies and the other at high frequencies.

And it is certainly possible to design a speaker that excels at one of them but doesn't do so well with the other and vica versa.

For example a small 2 way mini-monitor design with small baffle, very closely spaced drivers, carefully optimised crossover with good phase tracking etc, or a small full range driver, may excel at the "they are here" image floating in front of the speakers, but perform poorly in the depth illusion due to the small baffle and high midbass driver location causing large response holes and aberrations in the 150-400Hz range, primarily from room boundary effects.

Conversely an old fashioned wide baffle 3 way system with a floor mounted woofer could (assuming the crossover is competent) perform very well in the depth illusion and sound good on "you are there" recordings, but lack in the pinpoint imaging stakes on "they are here" recordings. I've certainly heard speakers like these before.

Is it possible to design a speaker that does both well ? I think it is yes, but I don't think it's common. The modern trend is to tall slender speakers and I don't think those will ever perform optimally in the 150Hz-400Hz region due to the baffle being too narrow, and not many 3 ways these days will deliberately place the woofer close to the floor and far from the midrange driver - but I think this is necessary to get this 150-400Hz frequency range right and therefore get the optimal reproduction of depth illusions.

I just wish to add to everything that you have stated; the ability to accurately track the dynamic swings that are inherent in a live acoustic performance, lends itself to the illusion of, "they are here".
 
I just wish to add to everything that you have stated; the ability to accurately track the dynamic swings that are inherent in a live acoustic performance, lends itself to the illusion of, "they are here".
I'm not sure that that's specific to achieving a "they are here" illusion vs "you are there" or any other kind of illusion.

I think good reproduction of dynamic range at all frequencies is important to any sort of reproduction. If you turn the volume up and it falls apart then it won't sound convincing on any kind of music IMHO.
 
After some recent experiences I'm beginning to wonder if a good part of "You are there" vs "They are here" is as simple as specific differences in the frequency response, particularly in the presence region...

I don't see how it can be as simple as just frequency response - reflections are known to be significant aspects of localization and without good localization you can't have a good "they are here" quality. I can't point to any specifics in your dialog that bother me, but your assumption above does. That frequency response is a major aspect, I don't doubt, but very early wall reflections are also a major factor as well.

There is also diffraction, which may or may not be part of what you are calling "frequency response".
 
I don't see how it can be as simple as just frequency response - reflections are known to be significant aspects of localization and without good localization you can't have a good "they are here" quality. I can't point to any specifics in your dialog that bother me, but your assumption above does. That frequency response is a major aspect, I don't doubt, but very early wall reflections are also a major factor as well.

There is also diffraction, which may or may not be part of what you are calling "frequency response".
I'm not saying that early reflections and/or diffraction don't upset "they are here" localisation, nor am I saying those can be ignored when trying to achieve it.

What I'm saying is that the ability to produce a convincing "they are here" image where the sound appears to float in the air in front of the speakers seems to strongly correlate with the presence region between about 2-5Khz, (as heard at the listeners ears...) and not surprisingly, since that's how it probably got its name.

Screw up that frequency region in any way (especially by suppressing it or having poor phase tracking between drivers crossing over in the region causing partial cancellation between drivers) and you just won't get a good solid "they are here" illusion, and in extreme cases, none at all.

And artificially boost that region, (as many speakers do - just look at the large 4Khz peak in the otherwise very good B&W Nautilus 802's midrange driver which shows through in the final response, which could have easily been dealt with with a single RLC notch if they so chose) and you can end up with an artificially vivid and present image that many like, but one that is ultimately a bit fatiguing and unnatural. (Personally I don't like any peaks in the presence region at all, I find them quite fatiguing after a while, so do what I can to eliminate them even if it means a more complex crossover)

Diffraction can easily affect the presence region detrimentally, I suspect primarily by causing large changes in the frequency response for small off axis deviations from the speaker.

Part of getting a convincing and stable image is for the frequency response of the direct signal from both speakers at the listener to be identical and not vary much when you move your head out of the "vice"...

Anything like diffraction that caused for example the presence region to be suppressed or greatly modified just because you moved your head a foot to one side, and perhaps caused a different change in response for the left and right ears, (speakers not perfectly symmetrically placed in regards to the listener) would pull apart the image.

So in this sense diffraction is detrimental to "they are here" imaging because it makes achieving a smooth, balanced presence region over a range of spatial listening positions impossible - your head is literally in a vice to get the one correct frequency response and as soon as you move your head the response through this region is wrong due to diffraction effects and the illusion is spoiled.

I've certainly noticed this head in a vice imaging problem with speakers with diffraction issues or odd (old fashioned, asymmetrical) layouts...if you listen carefully as you move your head sideways you can hear the image pull alternatively left and right as you move through the "ripples" in the polar response caused by diffraction. Audible both on music and pink noise. So I think minimising diffraction at high frequencies (say 2Khz and above) is crucial for good "they are here" imaging.

That's what I mean when I say its the frequency response that is important - but at the listeners ears, not at one specific speaker axis. Having a low diffraction speaker with well controlled near off axis response so that you minimise the frequency response changes at the listener when they move around slightly (and if the speakers are not ideally positioned) is half of the battle, but once you have that you still need to get the actual frequency response through the region correct as well.

As for VER, I suspect that's a different issue at play more related to changes in ASW.

I don't think any of the above is controversial, but the additional observation I was bringing is that while both are critically affected by frequency response, I believe that "you are there" imaging focuses on a different frequency range than "they are here".

With my speakers which are at the moment not working correctly through the presence region, I find that they are not providing a good "they are here" image (almost absent on some recordings that I know should have that effect) but they are still able to provide a strong and convincing "you are there" "image" on other recordings which I know have that kind of effect - my conclusion being that the problems in the presence region are not detrimental to localisation of depth that is required to make a sound source sound like it is behind the speaker wall, as those frequencies are not involved.

On those kind of "you are there" recordings there is not necessarily a distinct pinpoint image, often there is not and it is a very diffuse sound field, but there is a sense of you being in a large space with the sources of sound being "outside" the listening room behind the speakers.

Past experimentation leads me to believe that this sense of depth perception comes primarily from lower frequencies especially in the upper bass lower midrange region about 150-400Hz.

My point being that if you are able to get both frequency regions right, at the listener you could build a speaker that does well with both "you are there" and "they are here" imaging. One of the major problems with getting that low frequency region right is room boundary effects, and I think a lot of otherwise technically very good speakers just don't get this right as they aren't willing to take some of the steps that are necessary - such as a wide baffle or a woofer that is close to the floor. (The latter only feasible with a 3 way, really, as you don't want your midrange driver producing frequencies >300Hz close to the floor)
 
Last edited:
:snowman2:We've only had a little bit here and it didn't settle, I'm only half a mile from the sea, a mile or so in land and in the mountains it's a different story. My brother and sister-in-law live near Brecon, and they've had a foot of snow, which is a lot for here!:snowman:
 
Last edited:
There is also diffraction, which may or may not be part of what you are calling "frequency response".

^this


So I think minimising diffraction at high frequencies (say 2Khz and above) is crucial for good "they are here" imaging.

That's what I mean when I say its the frequency response that is important - but at the listeners ears, not at one specific speaker axis. Having a low diffraction speaker with well controlled near off axis response so that you minimise the frequency response changes at the listener when they move around slightly (and if the speakers are not ideally positioned) is half of the battle, but once you have that you still need to get the actual frequency response through the region correct as well.

a short up-firing coaxial meet those requirements


I don't think any of the above is controversial, but the additional observation I was bringing is that while both are critically affected by frequency response, I believe that "you are there" imaging focuses on a different frequency range than "they are here".

...

Past experimentation leads me to believe that this sense of depth perception comes primarily from lower frequencies especially in the upper bass lower midrange region about 150-400Hz.

My point being that if you are able to get both frequency regions right, at the listener you could build a speaker that does well with both "you are there" and "they are here" imaging. One of the major problems with getting that low frequency region right is room boundary effects, and I think a lot of otherwise technically very good speakers just don't get this right as they aren't willing to take some of the steps that are necessary - such as a wide baffle or a woofer that is close to the floor. (The latter only feasible with a 3 way, really, as you don't want your midrange driver producing frequencies >300Hz close to the floor)

again, a short up-firing coaxial meets the requirement
 
Thought i share this homemade tilting EQ tweak here because it also influence a lot on "you are there" and "they are here" imaging in room balance, and think it works pretty good after house-curve is dialed in as a last stage subjective tweaking tool where one can step small 0,1dB numbers at a time and evaluate acoustic domain performance. Its not a strait line completely ripple free when zoomed into but if it was it will take too many PEQs to change then every time one would step filter, so ended up use only two shelving filters to get it done.

attachment.php
 

Attachments

  • 3001.png
    3001.png
    41.1 KB · Views: 415
Thought i share this homemade tilting EQ tweak here because it also influence a lot on "you are there" and "they are here" imaging in room balance, and think it works pretty good after house-curve is dialed in as a last stage subjective tweaking tool where one can step small 0,1dB numbers at a time and evaluate acoustic domain performance. Its not a strait line completely ripple free when zoomed into but if it was it will take too many PEQs to change then every time one would step filter, so ended up use only two shelving filters to get it done.

650664d1513004688-suppose-shallow-vs-steep-argument-3001-png

Thx BYRTT, I really agree that such simple shelving EQ can have a major effect on how we perceive the issues being discussed in this thread.

In fact, I think the boost or roll-off on the ends of the spectrum, inherent in most all recorded material, has a super influence in how our systems responds in general.

My first attempt to address this issue was with sets of shelving filters, that allow me to basically create a seesaw frequency response, where I can control the fulcrum point and the tilt of the frequency response.
Though a lot of trial and error I've settled in on 640Hz as the fulcrum point, with 3 or 4 low shelf filters below, and 3 or 4 high shelf filters above. Filters are generally spaced an octave apart.
Works really well for making almost any recording sound natural. (My systems all start tuned flat without any EQ...really flat , freq and phase:)).

My latest attempt, to make the above less cumbersome, is I now have a volume control on each passband on my 4-way system. Each passband is about 2 to 3 octaves wide.
I use a mixer where I have 4 adjacent sliders for relative passband adjustments, all controlled jointly by a DCA for overall volume.
Each passband is linear phase thru x-over and summation to at least -20db, so I have a lot of ability to maintain linear phase throughout the spectrum, even as I move acoustic x-over points up or down.

I'm loving this setup. I can make almost anything sound good.
What I've learned in adjusting the passbands is I'm really compensating for whatever spectral tilt is in the recording.

My theory is: let's assume our hearing is reasonably the same...whoever did the mastering did their best to make the recording sound good in their mastering room.
Which means they dialed in levels and tonality in response to what they were hearing from their speakers with whatever spectral bias the speakers and room imparted..
I don't know a lot about making recordings, but I assume all must be mastered, whether recorded in studio or in live ensemble. If this is true, then all recordings reflect what the mastering room/speakers sound like.

If recorded where speaker/room was flat...I don't need to touch the sliders. If recorded where bright, the mastering will turn down the highs and I will hear they need to be turned up.

IOW, if I hear a spectral imbalance on any of the passbands relative to each other, I figure that reflects the opposite imbalance in where the recording was mastered.

All FWIW:D
 
I'm not saying that early reflections and/or diffraction don't upset "they are here" localisation, nor am I saying those can be ignored when trying to achieve it.

...

That's what I mean when I say its the frequency response that is important - but at the listeners ears, not at one specific speaker axis.

I would agree to that with one caveat and that is that the window time has to be adjusted as per our hearing, i.e. very much like the sliding window that is currently popular, which is a good approximation to hearing.

Greisinger says that 700 - 7k Hz is the critical range for imaging. Kind of agrees with what you are saying.
 
Last edited:
What do you mean by short? How about a wideband in a tube similar the LX Mini?

by short I mean 20-30 cm above the floor - if DBMandrake is right that "this sense of depth perception comes primarily from lower frequencies especially in the upper bass lower midrange region about 150-400Hz.
...and I think a lot of otherwise technically very good speakers just don't get this right as they aren't willing to take some of the steps that are necessary - such as a wide baffle or a woofer that is close to the floor."

however my personal experience accords with Griesinger's "700 - 7k Hz is the critical range for imaging"

it just have to be heavily tilted back, 45 degrees or more, and no need for it to be short
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.