The Advantages of Floor Coupled Up-Firing Speakers

is the directivity of the 10'' high enough to prevent the influence of the floor reflection in such a relatively upright position?


EDGE simulation of the floor reflection :)

An externally hosted image should be here but it was not working when we last tested it.


Note that EDGE cannot model the tilting of the baffle !

In reality the notch at 1.5kHz will not be so severe. At 1.5kHz 10" driver diameter is about one wavelength and it starts to beam already. Also the center distance to floor is less than a wave length below 1.5kHz. And at cross over of 1.5kHz the tweeter reproduces half of the signal energy.


- Elias
 
if the baffle would be tilted the response around 1kHz would be smoother due to on axis direct sound from the elements.

I was thinkig about the floor reflection's influence in time domain, that in such a configuration the floor reflection would be audible as a separate sound event, incoherent with the recorded sound event.

At 1.5kHz 10" driver diameter is about one wavelength and it starts to beam already.

I think that we need to control the directivity also below 1 kHz, preferably from around 500 Hz
 
Last edited:
I was thinkig about the floor reflection's influence in time domain, that in such a configuration the floor reflection would be audible as a separate sound event, incoherent with the recorded sound event.

Audible ? really ? Not sure about that. From the EDGE plot can be seen the notch at 1.5kHz. This corresponds distance difference between direct sound and floor reflection of about 45cm. In time domain it is 1.3ms.

Now the question is can two same sounds having time difference of 1.3ms below 1kHz be regarded as a separate sound events ?

Note again the situation is improved over that of the simulated when the baffle is tilted backwards. If only EDGE could do that..

I may have to run FDTD simulation on this case :D

- Elias
 
"fusion effect" ?

Got any reference ?

Usually psychoacoustic models employ time domain integrator after frequency domain filtering. I think time differences less than a wavelength below 1kHz result in summation response and they would not be separable perceptually.

Timbre may have shifted if there is frequency domain variation though.


- Elias
 
No, I mean "fusion effect". You imply it is "analogical to Haas effect", but not explicitly the same.

I imply? Not at all. Quite the contrary - I doubt. Therefore I ask :) because You seem to imply when You say: "I think time differences less than a wavelength below 1kHz result in summation response and they would not be separable perceptually."
 
Last edited:
The classical precedence effect / law of the first wavefront / Haas effect is based on interaural signals. So it operates in horisontal spatial domain.

We also know that HRTF is responsible for vertical spatial hearing. However the frequencies that allow vertical directional hearing is much above 1kHz.

Thus it could be safely assumed below 1kHz there is negligible available contributions to allow height perception. And we also know that a direct sound and its floor reflection are having same incidence angle relative to median plane i.e. interaural difference is the same for both sounds (this follows the Blauertian cone of confusion).

Based on all of this, I conclude there is no vertical precedence effect below 1kHz. Thus the floor reflection at 1.3ms below 1kHz cannot be perceived, other than changes in timbre perhaps but to what extend is unknown.


- Elias
 
conclude there is no vertical precedence effect below 1kHz. Thus the floor reflection at 1.3ms below 1kHz cannot be perceived, other than changes in timbre perhaps but to what extend is unknown.
- Elias

surely I can be wrong but isn't it just the other way round?

there IS a precedence effect -> reflection CANNOT be perceived as separate sound event (like in the case of horizontal lateral reflections covered by the Haas effect)

consequently:

there IS NOT a precedence effect -> a reflection CAN be perceived as separate sound event
 
Last edited:
surely I can be wrong but isn't it just the other way round?

there IS a precedence effect -> reflection CANNOT be perceived as separate sound event (like in the case of horizontal lateral reflections covered by the Haas effect)

consequently:

there IS NOT a precedence effect -> a reflection CAN be perceived as separate sound event


Well, the precedence is related in spatial direction. But below 1kHz it is not possible to detect the elevation, because as said the ITD will remain the same and HRTF does not give clues. So since it is not possible to determine the perceived direction for such sounds, there cannot be a precedence of any of the reflections. The sound direction remains ambiguous regardles of the order of the reflected sounds.


- Elias
 
"Research into
localization in the median plane, [1] [2] [10] [11],
concluded that three requisites were necessary for an
auditory stimulus to be accurately placed in a vertical
space: (a) the sound must be complex, (b) the complex
sound must include frequencies above 7000 Hz, and (c)
the pinna must be present."

"5 Conclusions
Our ability to localize sound sources on the median
plane relies primarily on pinna effects. According to
Blauert [1][2] and others, the curves within the pinna
reflect different frequencies depending on the source
elevation. It was clear from the experiments conducted
that localization on the median plane is poor for
phantom images generated by loudspeakers at any of the
three angles tested, with inaccurate median locations
and significant localization blur. Further, for sources
behind the subject on the median plane the results were
particularly poor. This suggests that a single
loudspeaker placed on the median plane and used for
height enhancement would be of little benefit for
vertical localization."

ELEVATION PERCEPTION: PHANTOM IMAGES IN THE VERTICAL HEMI-SPHERE , JAMES L. BARBOUR, AES 24th International Conference on Multichannel Audio

[1] Blauert, J. 1997, Spatial Hearing, MIT Press,
ISBN 0-262-02413-6
[2] Blauert, J. 1969-70, Sound Localization in the
Median Plane, Acustica, Volume 22, pp. 205-
213
[10] Roffler, S. and Butler, R. 1968 Factors that
Influence the Localization of Sound in the
Vertical Plane, Journal of the Acoustical
Society of America, Volume 43, No. 6, pp.
1255-1259
[11] Roffler, S. and Butler, R. 1968 Localization of
Tonal Stimuli in the Vertical Plane, Journal of
the Acoustical Society of America, Volume 43,
No. 6, pp. 1260-1266
 
Well, the precedence is related in spatial direction. But below 1kHz it is not possible to detect the elevation, because as said the ITD will remain the same and HRTF does not give clues. So since it is not possible to determine the perceived direction for such sounds, there cannot be a precedence of any of the reflections. The sound direction remains ambiguous regardles of the order of the reflected sounds.


- Elias

I am not sure if I understand You well.

But what I said before can be explained by the fact that lateral localization relies completely on binaural triangulation of a sound source, exactly therefore a lateral reflection carries no useful information about localization and it can only be a distraction from that perspective - hence the effect of suppression of perception.

on the contrary, in case of the vertical localization we have no "built-in" triangulation mechanism so why would our auditory system suppress a ground reflection?

Some sources indeed suggest that floor reflection may carry useful information about a sound source localisation, about its elevation, apparently also providing a sort of vertical triangulation:

preliminary observations suggest that (1) the upward bias in elevation seen for anechoic sounds is reduced when a floor reflection is added
see: Human Systems Integration Division @ NASA Ames - Projects

It seems that vertical localisation is not just HRTF but more complicated mechanism and floor reflection may be important.

It may also be a useful cue in the distance perception:

The dominant factors in determining distance perception are to some extent dependent on the reverberation time of the room. For two similar sounds close to and far away from the listener the one further away will have the following differences: (...) Less difference between time of direct sound and first floor reflection
see: http://www2.ph.ed.ac.uk/teaching/course-notes/documents/67/1569-SoundLocalisation1.pdf
 
ELEVATION PERCEPTION: PHANTOM IMAGES IN THE VERTICAL HEMI-SPHERE , JAMES L. BARBOUR, AES 24th International Conference on Multichannel Audio


We can place the bass-midrange speaker at any height and the perception does not change due to the elevation, IF the frequency band if below 1kHz.

However there might be perceptual changes on speakers at different heights because of room induced complex reflection patterns which are height dependent. This can change the low frequency ITD which results in perceived difference on speaker height. But the reason is not the elevation perception perse but the room.


- Elias
 
"5 Conclusions
Our ability to localize sound sources on the median
plane relies primarily on pinna effects. According to
Blauert [1][2] and others, the curves within the pinna
reflect different frequencies depending on the source
elevation. It was clear from the experiments conducted
that localization on the median plane is poor for
phantom images generated by loudspeakers at any of the
three angles tested, with inaccurate median locations
and significant localization blur. Further, for sources
behind the subject on the median plane the results were
particularly poor. This suggests that a single
loudspeaker placed on the median plane and used for
height enhancement would be of little benefit for
vertical localization."

the operating term above is primarily

interestingly Markus usually talks rather about " summing localization in the median plane":

page 100 of this thread

and also about "breakdown of precedence" in the median plane:

page 172 of this thread

apparently He believes that this is possible in the median plane but only in case of the ceiling reflection, not in case of the floor reflection, interesting, isn't it? :rolleyes:
 
the perception does not change due to the elevation, IF the frequency band if below 1kHz.

it follows from Your personal experience or from any scientific study?

because generally it appears to be a consensus in literature that frequency range critical for localisation perception is >0.5 kHz

I suggested above (and with some good reasons I believe) that the important range extends roughly 0.4<4.0 kHz

perhaps Allison would concur and exactly therefore the floor coupled woofers in His designs are low-passed at 350 Hz?

and perhaps this was main cause of Snell Type 1 problematic soundstaging - not the floor position of the tweeter but floor position of the midwoofer crossed to high (at 2 kHz)?
 
Last edited: