Are you guys assuming a monpole at ALL frequencies. If so then that is of no use at all since no one here is recommending that. And if you are NOT assuming monopole response at all frequencies then what are you assuming? How are you seperating out the fact that things can be monopole at LF and still be highly directional at HFs?
Sounds like an evasive response. What is the directivity of the sources as a function of frequency then? And If the directivity does change with frequency then every reflection will have a different frequency response. Your data as to arrival times and level thus becomes somewhat meaningless since they do NOT show the frequency response of these reflections. Many will be limited to LFs while others will be full range. None of this is evident in your results so not much can be learned from them.
Me I assume monopole up to 1k the most. Spatial clues are dense in the power range 200-800Hz in my experience but strongest around 1kHz. Don't know if there is some paper confirming that or I am perceiving wrongly.
Earl, already stated what you said (Post #396) - Elias seems to ignore it.
Maybe I shouldn't post this but it shows pretty good what psychoacoustic studies about room reflections have found so far:
Maybe I shouldn't post this but it shows pretty good what psychoacoustic studies about room reflections have found so far:
An externally hosted image should be here but it was not working when we last tested it.
Hello,
You see it wrong. I'm not ignoring.
There is the cardioide, which is good approximation of a monopole approaching high freqs where cone starts to have some directivity.
Regarding your picture, envelopment is right there in 100Hz-1kHz range as expected. This is what interests me most, and that is one reason I'm simulating what different directivity sources do in a typical reverberant living room.
To have envelopment, IACC must be low.
- Elias
markus76 said:
You see it wrong. I'm not ignoring.
There is the cardioide, which is good approximation of a monopole approaching high freqs where cone starts to have some directivity.
Regarding your picture, envelopment is right there in 100Hz-1kHz range as expected. This is what interests me most, and that is one reason I'm simulating what different directivity sources do in a typical reverberant living room.
To have envelopment, IACC must be low.
- Elias
Hello,
Actually, human hearing around 1kHz have the poorest ability to locate sounds. ITD has become ambiguous due to too short wave length and ILD is not strong enough because head shadow is too less still.
- Elias
salas said:
Actually, human hearing around 1kHz have the poorest ability to locate sounds. ITD has become ambiguous due to too short wave length and ILD is not strong enough because head shadow is too less still.
- Elias
Yes but the diagram that Markus posted has a spread for direction from 500Hz-5kHz. Image shift is shown to start at 1kHz.
Elias said:
yes that is correct BUT in case of REAL sound sources
what about PHANTOM sound sources where the ears are presented simultaneously with significant ILDs and ITDs across the full spectrum, including the 1 kHz area?
I insist that we should discuss spatial hearing/localization of PHANTOM sound sources because this is what we have when we are reproducing stereo
and I believe that localization of PHANTOM sound sources mechanism is special and differs significantly from localization of REAL sound sources
also we should take into account that we listen to stereo and localise those PHANTOM sound sources in reverberant space
and spatial hearing mechanism seems to follow different spatial cues in case of free field/anechoic spaces and in case of reverberant spaces
what do You think?
best!
graaf
Hello,
There is some interesting things going on
http://www.tutschku.com/content/festival-18.de.php
- Elias
There is some interesting things going on
http://www.tutschku.com/content/festival-18.de.php
- Elias
Elias said:
There is a lot more to sound system perception than "envelopment". Imaging is much higher on my list. And Toole himself admits that there is likely atradeoff in a small room between imaging and spatiousness or envelopment. The IACC is more associated with spaciousness than envelopment.
The cardiod is a poor model of a good loudspeaker. Its directive when it shouldn't be and its too wide for a good CD waveguide at HF. Your drawing a lot of conclusions from some very limited and questionable data.
Elias said:
That's WFS (wavefield synthesis). The only way (besides binaural recording) to rebuilt an original wavefield.
Available from http://www.iosono-sound.com/
Hello,
It always sound very unrealistic to me if I can 'image' a miniatyre orchestra playing between my speakers without transfering me to the recording space. For the transfer you need envelopment.
On the other hand very enveloping sound can sound very realistic even you cannot localise much. Think yourself at the back row of a concert hall below the balcony and a lady sitting in front of you blocking the view with a big flower hat 😉 You surely can hear the envelopment because you can hear you actually are in the concert hall.
Obviously one must thing in a creative way and interpolate the behaviour of a speaker to be a monopole and cardioid depending on the freq range in question.
- Elias
gedlee said:
It always sound very unrealistic to me if I can 'image' a miniatyre orchestra playing between my speakers without transfering me to the recording space. For the transfer you need envelopment.
On the other hand very enveloping sound can sound very realistic even you cannot localise much. Think yourself at the back row of a concert hall below the balcony and a lady sitting in front of you blocking the view with a big flower hat 😉 You surely can hear the envelopment because you can hear you actually are in the concert hall.
Obviously one must thing in a creative way and interpolate the behaviour of a speaker to be a monopole and cardioid depending on the freq range in question.
- Elias
Why's that? Can't you imagine the space defined by the aperture angle of your stereo triangle stretching out far behind your speakers? This is mainly done by adding reverberation to a recording and not so much by "adding" spaciousness or envelopment (which is by the way with stereo and a small room not possible at all because you would need cues arriving later than 80 ms). Or do you want to hear a singer with a 2 m big mouth between your sepakers because the mixing engineer forgot to add reverberation?
Elias said:
that is correct
I would say that in real life there is no "imaging" in the sense used in the phrase "stereophonic imaging" AT ALL
"stereophonic imaging" is an artifact
best regards!
graaf
markus76 said:
precisely speaking not space - it is purely 2D between the speakers
markus76 said:
funny thing that things like that actually happen 😉
best regards!
graaf
Live non amplified orchestral perception from over 20m far has velvety flow, open tone, and a general grouping of left - center - right. If we go to the conductor's position then everybody has a set position around. Like in a Jazz club.
No. It's a three dimensional plane. Elevation isn't possible (or an error in the setup) with stereophony. But it's still a three dimensional perception.
Imaging "performance" after Theile:
An externally hosted image should be here but it was not working when we last tested it.
markus76 said:
"+-30 deg." "simulated distance" - correct
with SLS much more than "+-30 deg." is it accurate or pure artifact? - I don't know but it sounds certainly more realistic - more like real life
"simulated spatial impression" - in what sense "spatial impression"?
where is 3D space?
what is the third dimension?
with SLS there is a kind of "spatial impression" of a sound source
in comparison in stereo triangle there is certainly "something" but I wouldn't call it "spatial impression" of a sound source
phantom sources are somewhat perceived in 3D space that is before us but they are not spatial THEMSELVES
best!
graaf
Elias said:
I'm not into "classical" music and what you are saying is really only true in that exact situation. With small "groups" in "typical" spaces the instuments are quite localized.
I'm not sure that this "interpolation is correct, but at any rate there is no evidence that you have done that, because each reflection would be frequency dependent and you don't show that.
- Status
- Not open for further replies.