Have you tried sitting closer with dipoles, so you are off axis and more between them, more reflected energy from the front wall, you might like it
Oh well....Charlie took it to an extreme.... http://www.diyaudio.com/forums/multi-way/317588-baffle-maximum-distance-front-wall.html#post5312189
So, listened to an LX Mini correctly placed in a room and the soundstage depth was huge compared to my existing DIY transmission lines in the same position.
I understand the basics as to why the LX Mini (and other dipoles I assume) and cardioid speakers (such as Kii Three) with controlled directivity will provide a good depth of soundstage. BUT...
I only have a small room, so wouldn't in-wall speakers be technically superior to an LX Mini or Kii three because there would be no reflections from behind the speaker? Assuming all other factors remained the same (as per this article)
Do I need to keep researching freestanding speakers such as the LX Mini / Kii Three, or is in-wall the solution?
Discuss/Flame as appropriate
I rented some Danley SH50s and listened to them in my uber-small room in San Diego. To me, the soundstage sounded deeper than the speakers, which is a tough trick to pull off. I believe these results are due to three things:
1) very good recordings
2) controlled directivity
3) well behaved phase response
My home speakers are Vandersteen, and though they satisfy requirement number 3, they don't satisfy requirement #1, and I think that's why the SH50s image better.
The Vandersteens are no slouch in the imaging department, but they don't make the room "disappear."
I've had corner floor-to-ceiling line arrays for a while and my experience has been different. They were in a relatively large basement (it was a fantastic man cave, which I unfortunately no longer have the pleasure of owning--picture 1), about 14 feet apart, and my listening position was about 2/3 in the back of the room about 16-17 feet away. The whole reason I chose to go down the "corner floor-to-ceiling line array" path is because I wanted to eliminate the front wall reflection and the floor and ceiling reflection.
There was very good soundstage depth. Closing the eyes definitely helped make the depth much more real. With the eyes open, there was depth, but the front wall sort of clouded your judgment. It was very hard to get past what your eyes saw. I've also listened to one of Pano's VOTT systems, and it had even more soundstage depth, but they were placed away from the front wall.
I moved to a new, much smaller house. I've setup the line arrays in corners adjacent to arches that separate two rooms(picture 2), i.e., there is no back wall. Now, I don't have to close my eyes to be completely immersed in the soundstage--it just happens. This space is also much more reverberent and I haven't had time to add treatment yet. But I suspect the main reason for the improvement in imaging is that my eyes are now telling me there is a huge space behind the speakers.
So, the conclusion is that the eyes, they deceive. I agree with krivium that there is no right or wrong way to do this. The key is in the BBC paper, i.e., early reflections should be suppressed but the energy is preserved. There isn't any particular directivity that is preferable, but a smooth power response is important, i.e., no sudden changes in the power response. Of course, a smooth, resonance free on-axis response is THE MOST important of all.
Attachments
Thanks to ra7 for providing an interesting and valuable data point. It is interesting how people looking the same set of facts could come up with similar conclusions based on different reasoning. For instance, when I saw ra7’s two pictures of the two different corner-mounted line source arrangements, I immediately “predicted” that the 2nd arrangement with the wide opening between the two line sources would be “better” (the meaning of which is subject to debate, of course) in terms of soundstage depth as compared to the first arrangement with the line sources separated by a solid wall. Although that prediction is not exactly wrong (it seems to be consist with ra7’s observation), the question is whether the factors I considered in reaching that prediction are valid.
It appears to me that we have at least two models that are based on two respective factors in our attempts to make sense of our observations regarding how a back wall affects the perception of soundstage depth. Those two factors are:
1. (the one ra7 emphasized) the visual image – as ra7 said, the eye deceives, or in my words the visual information affects how we process the auditory input to come up with a soundstage depth perception;
2. (the one I emphasized) The reflections from a far-away back wall, which I think would contribute to (or affect, bias, or distort) our processing of the auditory input to come up with a soundstage depth perception.
To me, those two models/hypotheses do not seem to be in conflict with each other. They could be both correct (or either could be wrong). Also, neither of the two factors has to be the sole factor that affect how we perceive soundstage depth. They both could be at work in helping us generate the illusion of soundstage depth. The verification of the validity of these factors and their significance is complicated however by the existence of other factors, such as the relatively early reflections in a typical system (with two speakers close to the back wall), the difficulty in isolating one factor from the other, and the inherent “subjective” nature of the soundstage depth (because a human brain in involved in the processing).
At this point, I don’t think my observations of my systems and ra7’s observations of his systems are sufficient to provide definitive answers to the questions of if both models are correct, is one factor more important than the other, and if so, under what circumstances. It seems to me that we need to design some experiments to gather more data in order to answer these questions. For instance, perhaps we can build a system with a movable back wall so that the system can be converted from having 0 physical depth (fully “in-wall”) to having various physical depths, and recruit a listener who cannot see (so that there is no deception by the eyes
).I have to admit that I have little knowledge about psychoacoustics, and am just making some speculations out of my ignorance. If the issues regarding soundstage depth perception have been studied and answered, I’ll appreciate it if someone can point to me the right literature.
Kurt
Last edited:
I presume you mean in terms of recording and reproduction, not in the real world. All the information necessary should be in the recording, the important thing is for the reproduction not to distort the spatial information, Linkwitz explains how to achieve this Sound_field_control_for_stereo
Scott don't take it personnaly (you are not a target for me) or don't think i've got something against Linkwitz (in fact each time i think about the Linkwitz Transform or his work with Riley i've got a tear in my eyes and could have a temple made to his glory... ), but please, as good and interesting he is there is many thing he forgot to say...http://microphone-data.com/media/filestore/articles/Stereo zoom-10.pdf
Of interest is the chapter 1.4 (p9 - this is obvious the preceding paragraph are interesting too) and following. *
Now you should understand why i talk constantly about illusion. And please not this is a very short approach to the original book(s -there is two about just the stereo recording more or less summarized in this short paper) of Williams (the author) and i won't talk about other authors...
The process of recording (and it is worse in stereo!) is flawed at the beginning. Spatial cues are modified whichever couple you use!
And this talk just about the theory behind it, the practice reveals a lot of surprise too...
So when Linkwitz (with a bit of condescension) talk about audio engineers and their lack of knowledge it just makes me smile (it was true with the generation which have the same age as him, now you can't survive in a such technical world without solid knowledge! Autodidacts (-selfmade men) in this fields are few since 2000 years). The post from CharlieLaub does explain why a bit: he became obscessive in his search.
In a way this doesn't astound me and this is expected from someone with such a commitment, but as most obscessive persons i met ( and in the proaudio studio field i can guaranty you there is A LOT OF GUYS and GIRLS like that ) the reality is not always said to the end... some parameters (which doesn't enter the pattern they are locked into) are let along the side of the road.
Don't get me wrong here again please. I'm not in anyway saying that the man don't deserve his status or your respect (or mine) or in anyway that his proposition his not valid.
I'm just saying that the reality is much more complex than the way it is presented in his view.
Maybe i'm wrong, maybe it is just oversimplification to make things reachable to the 'normal guys', i don't know, anyway for someone which have a little bit of perspective on the whole process this is not a definitive truth or answer.
Hope i won't be banned and that i wont upset the followers of his proposition but i can't keep that for me. There is a lot of guys which have the same commitment as him in the pro field. In fact there is a S.Linkwitz or two in each fields i've approached in the audio. All does have probably a part of truth with them, but to my knowledge no one have a definitive answer.
* in the article there is a kind of couple of interest which isn't presented: the blumlein couple. It is of interest because it use two fig of 8 microphone each other at 90° and coincident for stereo recording. It is in fact the first one invented and used. It should be interesting to hear what a recording done with it coud render on different type of speakers.
Every one which have used one know how it is interesting in it s rendering of "spatial cues".
Last edited:
Yes Strechneck,
i know this one too, one of my friend buyed this model of behringer after reading that ( a sound engineer and i can tell you Behringer as a bad rep in pro audio) and afaik he still use them to great results to earn his life.
In fact if you want real answer you ll need a double abx ( double blind test) as Kchang purpose (but even harder: listener does not know what he listen to or even see it, the person which must switch the units under test neither). Very difficult and time consumming to perform.
Results are almost always as hard as it is to apply...
i know this one too, one of my friend buyed this model of behringer after reading that ( a sound engineer and i can tell you Behringer as a bad rep in pro audio) and afaik he still use them to great results to earn his life.
In fact if you want real answer you ll need a double abx ( double blind test) as Kchang purpose (but even harder: listener does not know what he listen to or even see it, the person which must switch the units under test neither). Very difficult and time consumming to perform.
Results are almost always as hard as it is to apply...
Other food for thoughts:
http://www.diyaudio.com/forums/multi-way/317122-baffle-diffraction-8.html#post5454722
http://www.diyaudio.com/forums/multi-way/317122-baffle-diffraction-8.html#post5454722
Thanks scottjoplin for the link. In fact, I know that webpage and have read some of Linkwitz's stuff (I've been building dipole speakers for 20 years).
Maybe I should reword my statements to be clear on the particular information I am looking for. I want to know how reflections from the surrounding, especially reflections from a far-away back wall, if delayed long enough, would impact our perception of soundstage depth. If the answer is “not at all,” I definitely want to see some proof.
In this regard, Linkwitz’s philosophy is summarized in these statements from the page you linked to: “Under these conditions the direct sound from the speakers dominates the auditory scene. Spatial properties of source location, spread and distance are derived from cues imbedded in the mix of the recording and cues from the recording venue. Reflections from the listening room are delayed and are copies of the direct sound at a lower level. The room reverberated sound has the same timbre as the direct sound. A listener automatically withdraws attention from the room and falls for the aural illusion, which is delivered by cues in the direct sound streams from the two loudspeakers and is produced in his mind.”
I interpret that view to mean that if the reflections (or “room reverberated sound”) are sufficiently delayed and attenuated (but similar in spectral distribution to the direct sound), they can be “automatically” tuned out by the listener and thus become relatively harmless. I wonder, however, whether that view regarding room reflections being rendered benign is a bit too coarse and too “convenient.” Even if a listener automatically withdraws attention from the room reflections, it does not mean that the room reflections would no longer have any effect at all on the “auditory scene” or “aural illusion.” The withdrawal of attention from the reflected sound cannot be 100%, can it? After all, the energy of the reflected sound is not that negligible compared to the energy of the directed sound. Even if the reflected sound no longer has a first order effect on the “aural illusion,” it might still have a second order effect.
Also, another piece of anecdotal evidence that troubles me is that a well-known member of DIYAudio once said (I forgot who he was and when he said it) that listening to a stereo pair of speakers in an anechoic chamber is like listening to a pair of giant headphones. I don’t have such experience myself as I have never been in an anechoic chamber, but if we assume that statement is correct for arguments’ sake here, then having all the depth cues in the recording does not seem to be a sufficient condition for having great soundstage depth, and perhaps room reflections do play an important role in creating the illusion of soundstage depth. I hope someone here can confirm or debunk this statement regarding speakers in an anechoic room sounding like giant headphones.
By the way, I always listen to music with my eyes open. So, even if a great soundstage depth is just an illusion created by my deceiving eyes, I love that illusion and will always choose, if I have the space, to keep the speakers far away from the back wall.
Kurt
I think you are right, it's hard to imagine the reflections don't impact the illusion at all, and I think the clue is in the word illusion and where psychoacoustics may well play a large part certainly in regards to our expectation and how we have learnt to hear by also using our eyes. You will probably know that Linkwitz believes it's important to have plenty of experience of sound in the real world and to refresh this memory as often as possible and this helps us recreate a believable scene in our minds. I don't believe everything he says, far from it. It seems logical to me that we are all different and have had different experiences, personally, I haven't spent a lot of time in concert halls listening it to orchestras, very little in fact, so there's no way for me to know what this should sound like. But I can use my imagination. As I alluded to earlier, I suspect the reflections do help me create a better depth image in my mind with recordings where the required clues may be lacking, but I have no way of proving this
Hi guys,
Anechoic listening room?
This is true at one point only tho which is te apex of the triangle formed by speaker and listener (iirc i never been in an anechoic room and all this is theorical for me).
Yes, and once you determine a target for attenuation and delay we enter principle of LEDE (-20db, 20ms approximately).
From my experience you are right.
I don't think this would be comfortable.
Here again i think you are on right track/have a good analyze of situation. The way choosen in the LEDE is to use QRD (Quadratic Residue Diffusor) on the backwall (relative to listener). The qrd work in a manner it deal with specular reflection and render them uncorrelated (a multiple of copy of the signal but delayed each one to another so with different time arrival to the listener and so with different phase relationship too. I talked about a cluster of reflections earlier). The attenuation is achieved by distance travelled by sound (-6db each time distance is doubled) and the fact that this clusters are reflected (in parts) around other boundaries too.
Qrd have a bandpass. You can make them efficient for anything you want by size goes quickly out of manageable size. In studio they are usually efficient from 400hz up (around 30cm deep).
Below 400hz absorption is usually easier to manage than diffusion, so this is what is used.
This is the reason why some attempt to treat room with only absorption is always detrimental.
There is a misunderstanding i think between our different views.
I never said the reflections should be killed. They should be controlled. The main difference between the 'lede' (which i don't stick totally with -i think RFZ MUST be added to it- but is easier to understand) and Linkwitz approach is in the time gap specified (around 6ms for linkwitz, 20ms for lede) and the level from the first reflection to the user (-6db for Linkwitz, -20db for LEDE).
The other big difference is in the need for the first reflections to be highly uncorrelated when they first arrive at listener in a LEDE this is not the case with Linkwitz approach for the first reflection (ER). For the reverb (what you call late reflection) it will depend from room furniture, etc,etc,... The more entropy the better. And we enter WAF territory there.
He his absolutely right. But... you do that since your birth but from a subconscient process. You 'learned' all this cues about room dimension, wall materials, etc,etc, each time you experienced something new and you stored all that in a library your brain is constantly refereing too.
About the instruments it i the same and here i agree with Linkwitz if you don't listen to real instruments you can't have a real point of reference of howw they sound.
And you are absolutely right in my view! Where i see (hear) an issue is that the clues which are added to the message doesn't vary with the message played (they are fiwed by the room you play music in). This can be totally adapted to some signals, totally detrimental to others. The main problem being most audiophile listen to classical music, and the venue in which the music is originaly played are HUGE room with very long ER and this is stamped by our domestical room.
This can be enjoyable but this is another layer of illusion added to a less than perfect illusion.
Wesayso give us a hint to some answers. I've spent a large part of previous night reading (quickly) what Griesinger have to offer and this is really interesting (steep understanding slope curve if you are not used to what he is talking about tho) and by far the most interesting read i've had about all that.
http://www.davidgriesinger.com/
Anechoic listening room?
This is true at one point only tho which is te apex of the triangle formed by speaker and listener (iirc i never been in an anechoic room and all this is theorical for me).
Yes, and once you determine a target for attenuation and delay we enter principle of LEDE (-20db, 20ms approximately).
From my experience you are right.
I don't think this would be comfortable.
Here again i think you are on right track/have a good analyze of situation. The way choosen in the LEDE is to use QRD (Quadratic Residue Diffusor) on the backwall (relative to listener). The qrd work in a manner it deal with specular reflection and render them uncorrelated (a multiple of copy of the signal but delayed each one to another so with different time arrival to the listener and so with different phase relationship too. I talked about a cluster of reflections earlier). The attenuation is achieved by distance travelled by sound (-6db each time distance is doubled) and the fact that this clusters are reflected (in parts) around other boundaries too.
Qrd have a bandpass. You can make them efficient for anything you want by size goes quickly out of manageable size. In studio they are usually efficient from 400hz up (around 30cm deep).
Below 400hz absorption is usually easier to manage than diffusion, so this is what is used.
This is the reason why some attempt to treat room with only absorption is always detrimental.
There is a misunderstanding i think between our different views.
I never said the reflections should be killed. They should be controlled. The main difference between the 'lede' (which i don't stick totally with -i think RFZ MUST be added to it- but is easier to understand) and Linkwitz approach is in the time gap specified (around 6ms for linkwitz, 20ms for lede) and the level from the first reflection to the user (-6db for Linkwitz, -20db for LEDE).
The other big difference is in the need for the first reflections to be highly uncorrelated when they first arrive at listener in a LEDE this is not the case with Linkwitz approach for the first reflection (ER). For the reverb (what you call late reflection) it will depend from room furniture, etc,etc,... The more entropy the better. And we enter WAF territory there.
He his absolutely right. But... you do that since your birth but from a subconscient process. You 'learned' all this cues about room dimension, wall materials, etc,etc, each time you experienced something new and you stored all that in a library your brain is constantly refereing too.
About the instruments it i the same and here i agree with Linkwitz if you don't listen to real instruments you can't have a real point of reference of howw they sound.
And you are absolutely right in my view! Where i see (hear) an issue is that the clues which are added to the message doesn't vary with the message played (they are fiwed by the room you play music in). This can be totally adapted to some signals, totally detrimental to others. The main problem being most audiophile listen to classical music, and the venue in which the music is originaly played are HUGE room with very long ER and this is stamped by our domestical room.
This can be enjoyable but this is another layer of illusion added to a less than perfect illusion.
Wesayso give us a hint to some answers. I've spent a large part of previous night reading (quickly) what Griesinger have to offer and this is really interesting (steep understanding slope curve if you are not used to what he is talking about tho) and by far the most interesting read i've had about all that.
http://www.davidgriesinger.com/
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.