Actually, the main idea is to eliminate all early reflections and then enjoy the reflections that are recorded on the record. That way you get the room to sound like the original.
Any early reflections distract from the depth on the recording.
That includes wall reflections and cabinet back wall through-the-cone sounds.
myhrrhleine,
It seemed like Scott was attempting to reinforce those delayed sounds to give artificial depth, at least that was how I was reading his post.
It seemed like Scott was attempting to reinforce those delayed sounds to give artificial depth, at least that was how I was reading his post.
In a small room context reflections created by the speakers are all pretty much "early".
Most of these reflections simply aren't heard as "reflections".
What they are capable of however, is increasing the intensity of direct sound at a given angle relative to the listener.
-particularly left speaker to left wall and right speaker to right wall - tends to help with reproducing a sense of envelopment (provided it's on the recording) simply by increasing intensity of the left channel for the left ear (and right channel for the right ear) at a greater angle than what the loudspeakers alone will provide (again, relative to the listener - approaching that +/- 90 degrees).
"eliminating" early reflections then is almost always a bad idea - it "shrinks" the lateral perspective beyond the loudspeakers positions AND can/often does substantially reduce reproduced envelopment. The perceived angle of intensity in a "dead" room just extends to the speakers - say 30 degrees from the listener (depending on listener position and loudspeaker "spread"), but with the reflections might be more like 45 degrees.
Note: this is not at all *creating* envelopment. That's impossible in a small room context and frankly wouldn't be appropriate anyway because at that point your room would actually be "stamping" an acoustic effect onto every recording (..like a room does in the modal region).
Now what IS bad are contra-lateral reflections - i.e. left speaker reflecting sound to right wall and right speaker reflecting sound to left wall. That also shrinks the sound-field perspective laterally. And while increasing a monophonic emphasis to an extent (and thereby increasing depth of field) - it doesn't do this very well. Instead that cross-correlation of direct sound seems to represent all you need for depth enhancement. This is probably a result of just how the sound diffracts around your head for that cross-correlation - something the contra-lateral reflections aren't really achieving in a similar manner.
I should also note that you have still have to be careful about proximity to the near walls for the loudspeaker - i.e. left speaker to left wall (and the wall directly behind the loudspeaker). While moving the loudspeaker closer to the wall increases apparent intensity it does two negative things (in the context above):
1. it reduces the apparent angle relative to the listener's head (the reflections), and
2. it increases the freq. that the speaker starts "bounding" to the wall - to where the reflection is no longer a reflected wavelength (..and not what I would generally deem as a "reflection").
Again though, about 3 feet from walls should be good enough for most purposes, and 5 feet should be plenty with respect to the near side-wall (...though 7+ feet from the wall behind the loudspeaker is best).
Most of these reflections simply aren't heard as "reflections".
What they are capable of however, is increasing the intensity of direct sound at a given angle relative to the listener.
-particularly left speaker to left wall and right speaker to right wall - tends to help with reproducing a sense of envelopment (provided it's on the recording) simply by increasing intensity of the left channel for the left ear (and right channel for the right ear) at a greater angle than what the loudspeakers alone will provide (again, relative to the listener - approaching that +/- 90 degrees).
"eliminating" early reflections then is almost always a bad idea - it "shrinks" the lateral perspective beyond the loudspeakers positions AND can/often does substantially reduce reproduced envelopment. The perceived angle of intensity in a "dead" room just extends to the speakers - say 30 degrees from the listener (depending on listener position and loudspeaker "spread"), but with the reflections might be more like 45 degrees.
Note: this is not at all *creating* envelopment. That's impossible in a small room context and frankly wouldn't be appropriate anyway because at that point your room would actually be "stamping" an acoustic effect onto every recording (..like a room does in the modal region).
Now what IS bad are contra-lateral reflections - i.e. left speaker reflecting sound to right wall and right speaker reflecting sound to left wall. That also shrinks the sound-field perspective laterally. And while increasing a monophonic emphasis to an extent (and thereby increasing depth of field) - it doesn't do this very well. Instead that cross-correlation of direct sound seems to represent all you need for depth enhancement. This is probably a result of just how the sound diffracts around your head for that cross-correlation - something the contra-lateral reflections aren't really achieving in a similar manner.
I should also note that you have still have to be careful about proximity to the near walls for the loudspeaker - i.e. left speaker to left wall (and the wall directly behind the loudspeaker). While moving the loudspeaker closer to the wall increases apparent intensity it does two negative things (in the context above):
1. it reduces the apparent angle relative to the listener's head (the reflections), and
2. it increases the freq. that the speaker starts "bounding" to the wall - to where the reflection is no longer a reflected wavelength (..and not what I would generally deem as a "reflection").
Again though, about 3 feet from walls should be good enough for most purposes, and 5 feet should be plenty with respect to the near side-wall (...though 7+ feet from the wall behind the loudspeaker is best).
Last edited:
For the most part - yes.
And,
It's not just the contra-lateral reflection.
The problem is that IF the loudspeaker is increasingly directive at higher freq.s (which most are), then what you've also done is
-Lowered the apparent intensity both *(directly from the loudspeaker) and from the near wall reelections - AT AN INCREASED ANGLE.
(..of course if the loudspeaker isn't directive then the toe-in is basically meaningless).
*you can test this outside in a mostly "free-field" environment OR with an absorptive room (like with heavy drawn curtains all around the room walls).. provided you have a sufficiently narrowing directivity loudspeaker (and an appropriate amount of angle of "toe-in" to "activate" this effect). What I've found most surprising is that this effect becomes audible well below the typical psycho-acoustic 1.5 kHz "limit" where intensity is usually specified as being dominate.
Last edited:
I've tested my hearing in the top octave and I simply can not hear any change if the music signal is limited to 14Khz. 13Khz is about the threshold. I can measure the output, but simply can not hear it. No difference in the sound with it there or not. Not sure I could ever hear much above 17K, even as a kid, tho I could hear 15.75KHz TV noise just fine.
I know we like to think that response above 15Khz adds a certain "something", but for many of us, it just doesn't. You can test that easily with filtered music.
I know we like to think that response above 15Khz adds a certain "something", but for many of us, it just doesn't. You can test that easily with filtered music.
Yes, I often think what people are "picking-up" on with super-tweeters is their polar behavior at LOWER freq.s.
-wider dispersion providing more "air".
It' makes me somewhat dubious of the search for a compression driver that extends past 14 kHz without a major linearity problem. Basically, "whats the point?"
They would be far better served "doubling-up" on the waveguide and compression driver that doesn't extend much past 14 kHz without problems - and operating in a dipole or bipole configuration. (..though with the bipole you have to worry about the lower freq. limit and the side-combing.)
-wider dispersion providing more "air".
It' makes me somewhat dubious of the search for a compression driver that extends past 14 kHz without a major linearity problem. Basically, "whats the point?"
They would be far better served "doubling-up" on the waveguide and compression driver that doesn't extend much past 14 kHz without problems - and operating in a dipole or bipole configuration. (..though with the bipole you have to worry about the lower freq. limit and the side-combing.)
I'm right there with you Scott. That sums up my thoughts on the matter.
Kindhorn, what I've found interesting is to test music and pink noise that been FR limited before it gets to the speaker. Can you really hear the difference on that sort of content? Sine waves can be tricky to use as hearing tests.
Kindhorn, what I've found interesting is to test music and pink noise that been FR limited before it gets to the speaker. Can you really hear the difference on that sort of content? Sine waves can be tricky to use as hearing tests.
Yes, they fuse with the direct sound and subjectively colour it, which is undesirable. Further, the higher the SPLs, the more coloured the sound becomes because this is time domain distortion which unlike linear distortion becomes less masked with higher SPLs.
Of course the intensity of the sound is increased because the number of point sources - delayed in time - (the reflections!) is increased which masks the sound coming directly from the speaker to listener. And most importantly, masks the low level detail carried by sound coming directly from the speaker to the listener. This is not desirable.
"A sense of envelopment (provided it's on the recording)," as you call it, is reverberation from hall or provided by recording engineer, or are upper partials of instrumental and vocal sounds, is at low level SPLs compared to fundamental and 1st and 2nd harmonics of most instruments. And their direct signal from speaker to listener will be masked by the cascading early reflections you advocate.
Griesinger notes,
The
perceptions are found to be related most commonly to the lateral (diffuse) energy in halls at least 50ms after
the ends of notes (the background reverberation) and less often but importantly to the properties of the
sound field as the notes are held. Experiments with orchestral music at high reverberant level indicate that
it is the very late >300ms reflected energy which is most responsible for spaciousness.
https://www.google.com/url?sa=t&rct...sg=AFQjCNHJZr1Gl2ti-hEP-wo99UuhF3X_GQ&cad=rja
On this time scale the recorded signals subjectively indicating envelopment and spaciousness are very low level indeed and what you prescribe will eliminate them from the perceptual sound field.
What you are prescribing is replacement of the low level sound that may be recorded or injected by the engineer with a lateral acoustic mush that certainly envelopes, no doubt.
I've not found it to be "colored" at all.
Nor have Olive and Toole's research.
Your Griesinger quote only talks about acoustic envelopment, NOT about anything to do with the reproduction of envelopment - nor any perceived coloration or loss of detail due to room reflections.
Again, if the sound is "colored" to you, or seems to be masking low-level detail - I suggest looking for an alternative reason. (..and there are several, a few of which I've commented on.)
-your brain should be effectively "filtering-out" those early reflections. If it's not, you are an exception to the "rule".
Nor have Olive and Toole's research.
Your Griesinger quote only talks about acoustic envelopment, NOT about anything to do with the reproduction of envelopment - nor any perceived coloration or loss of detail due to room reflections.
Again, if the sound is "colored" to you, or seems to be masking low-level detail - I suggest looking for an alternative reason. (..and there are several, a few of which I've commented on.)
-your brain should be effectively "filtering-out" those early reflections. If it's not, you are an exception to the "rule".
Pano - a true exception here - I AGREE with you!!
When MP3 was being developed there were many blind studies done on what HF limit constitutes a perceptual difference. They concluded that anything above about 8-10 kHz adds very very little to perception even among those who had "normal" hearing.
People will create a 15 kHz tone and blast it out of their speakers and then when they detect this signal they claim that they can hear 15 kHz and so it must make a difference. This test is flawed in so many ways that it is not even worth commenting on.
To me above 10 kHz is just not worth worry about - nice if it is there but hardly worth compromising anything else for.
A speaker with uniformly falling off axis response is hugely advantageous, if toed in, because it affords a large sweet spot.
If one moves off the central listening axis, the nearer speakers sound falls off in intensity while the further gains.
In this day of HT, games, and just listening to music with friends, this strikes us as a Good Thing, right?
Scott - your psychoacoustics is different than mine. I see the first side wall reflection as the worst because it is coherent with the direct sound at each ear and will create a timbre change (summation type filtering) within the ear canal itself. The brain cannot do anything about this and as such, while it does add "spaciousness" to the sound, it seriously degrades imaging.
The contra-lateral reflections - to opposite ears - are uncorrelated with the direct sound (if the two stereo channels are uncorrelated as they should be if this is truly stereo). Hence they will not create a timbre shift but will appear in the ear canal as two separate signals which the ear/brain system can sort out as such.
If you, or anyone else, knows of any studies which contradict this then I'd like to know as this has been my understanding for a very long time.
And it is possible to suppress the very early, say < 10 ms reflections with high directivity speakers toed-in. This creates a very low first reflection, but a larger contra-lateral reflection, but one that significantly later in time than the first wall reflection. Using psychoacoustic metrics this situation should result in good imaging (low image smearing) without any loss in spaciousness. It is important however that the speaker NOT beam (as most do) as this can create a reflection that is greater in magnitude that the direct sound, resulting is a sure failing of imaging (although perhaps greater spaciousness.)
PS. You use the word "envelopment", but as I understand the word it is not what is happening in this discussion. I believe that Griesinger coined that term to describe the effect at LFs of the sound being all around you. The feeling of a larger room at HFs is "spaciousness". Again, if you have references otherwise I'd like to know what they are. It is important to use common terminology or we will misunderstand each other. I am willing to use any terminology that is in common practice in the scientific literature, but using words willy-nilly doesn't cut it with me.
As I said, in the near wall reflection case, the signal changes occur before the sound even impinges on the ear drum. How can the brain "filter" this out? On what basis does it know how that signal aberration was created?
PS. What "rule" is that?
Pano,
I just did that to see if I could hear the 16Khz tone as a simple sine. I was actually doing log chirp testing before I did that as I wanted to see if I was really hearing the chirps that I thought I was. I couldn't detect anything above 16khz, and as Earl was just saying I was only putting less than 1 watt to the speaker. It was only 1V @4 ohms with a 86db efficiency speaker so not so loud as to fool me. I can still hear fairly high by my discrimination of cymbal sounds and such. Not going to hear a dog whistle anymore though.
Frank what you are describing would be a decreasing sweet spot with increasing frequency. So I don't see how toeing in a speaker that has a narrow pattern up high would increase the sweet spot? That would make it more localized in a narrow angle.
I just did that to see if I could hear the 16Khz tone as a simple sine. I was actually doing log chirp testing before I did that as I wanted to see if I was really hearing the chirps that I thought I was. I couldn't detect anything above 16khz, and as Earl was just saying I was only putting less than 1 watt to the speaker. It was only 1V @4 ohms with a 86db efficiency speaker so not so loud as to fool me. I can still hear fairly high by my discrimination of cymbal sounds and such. Not going to hear a dog whistle anymore though.
Frank what you are describing would be a decreasing sweet spot with increasing frequency. So I don't see how toeing in a speaker that has a narrow pattern up high would increase the sweet spot? That would make it more localized in a narrow angle.
What they are *really* capable of is degrading the perception of good sound. Home experimenters beware. The sighted test cycle is prone to delusion and malconclusion.
First, testing for a pure tone in a loudspeaker is a bad idea. Could you hear a 16 kHz tone embedded in a musical source? I highly doubt it.
Next the effect that Frank talks about works very well as long as the speakers are CD. If they are not CD then of course it doesn't work.
WOW! Is this a voice of reason? A light in the darkness? No, probably just a misprint!