Early reflections/diffraction <1ms from speaker are primary source of image distortion. This may be thought of as fine polar behavior.
When small changes in toe in/out of one speaker cause image shift in dead room/outdoors, it is the speaker's fault. This leads to shrinking of the sweet spot and loss of image detail.
Reflections >3ms have nothing to do with imaging. These are integrated into loudness. Any spaciousness derived from such reflections are the spaciousness of the listening room, not the recording.
Funny you should mention that, I've been waiting for it to come up on another thread
Ok, If a speaker has a rising level toward the opposite side that leads to the first opposite wall reflection mirroring the loudest axis of the speaker, is this worthy of changing the triangle, or is it a reasonable compromise that could be dealt with later?I still don't understand the question.
If time/intensity trading is used out to an un-occupied region, would the effort be wasted on the reverberant field.
I agree with you in principle, but not necessarily the numbers. One problem is that the numbers are not the same across frequency, but, in general, I think that yours are too small. I think that your last sentence is kind of self evident. "The sky is blue because its blue."
This is why I eventually want to look at the reflection pattern.
Again, the question here (if it is one) is not clear. And if its just a statement thenh I don't understand.
Just in case anyone wants more subjective reports than two, here's mine:
With any speaker reasonably close to constant directivity behavior, it seems pretty easy to get a phantom center-panned image like my little sketch here, where the lines point to listener location and the ellipses indicate the perceived locations of various centered sounds.
It gets more fussy when it comes to sounds that are panned pretty hard to one side and the listener is closer to that side. Those tend to want to collapse onto the speaker-to-listener axis. The more edge diffraction and early reflections are eliminated, the less this seems to happen, so I'm not sure if those things might be the only cause of the problem, or if there is more to it.
In any case, it's an understatement to say that I find the overall effect of an image that stays relatively stable while moving around very pleasing, even when it's far from perfect.
With any speaker reasonably close to constant directivity behavior, it seems pretty easy to get a phantom center-panned image like my little sketch here, where the lines point to listener location and the ellipses indicate the perceived locations of various centered sounds.
It gets more fussy when it comes to sounds that are panned pretty hard to one side and the listener is closer to that side. Those tend to want to collapse onto the speaker-to-listener axis. The more edge diffraction and early reflections are eliminated, the less this seems to happen, so I'm not sure if those things might be the only cause of the problem, or if there is more to it.
In any case, it's an understatement to say that I find the overall effect of an image that stays relatively stable while moving around very pleasing, even when it's far from perfect.
Attachments
Last edited:
Okey-dokey. Did you catch that game on the box last night? Great, wasn't it.
I sometimes like to ask questions when the opportunity arises, even if I'm ill prepared and really ought to be glueing the side on that tapped horn I'm keen to listen to.
Out of curiosity, with this study would you leave the reflections as pressure when adding them?
You know I had kind of taken for granted that this image tradeoff worked and created a much more stable image as one shifted seats laterally But there is this one guy for which nothing seems to work.
I don't think that this would be of much use. What I will probably do is show the impulse response for the reflection cases. The delay and the level of the reflection are, to me, the important aspects and I am interested in how this maps with different speaker directivities and placements. Adding the pressures would yield a real mess, although I will probably look at it.
You know I had kind of taken for granted that this image tradeoff worked and created a much more stable image as one shifted seats laterally But there is this one guy for which nothing seems to work.
There's always this one guy that ruins it for everybody. I'm close to tears
I find it interesting that nobody tried the test I had proposed.
I think your "sky is blue because its blue" response is dismissive hand waving.
All impulsive sounds attracting attention contain higher frequency components. The localization processing time proportional to wavelength. The brain figures out source location from higher frequency components long before lower frequency components can be processed into location. Impulsive sounds devoid of high frequencies have no detailed localization information.
My Pluto Clone are extremely omnidirectional <3kHz. As angle of listening triangle gets wider, by speaker placement or getting head closer to speaker plane, image gets wider and shallower. When I set the speakers up midway in my 25' x 17' living room ( 50% of wall surfaces are brick and glass with no treatment, rest of wall surfaces are wood), placed 3' from long walls with tweeters pointing directly at each other, the image is identical listening from either end of the room.
With speakers placed closer together the image rolls down into floor with head in vertical plane of speakers, and above horizontal plane of speakers. Laying on floor between speakers, image is up in the ceiling.
My room is more of a reflective rectangular prism than an anechoic chamber or outdoors. My expectation is that if all furnishings were removed, the imaging specificity of the speakers would remain, and that the sound would be louder due to room's increased reverberation time.
So, on what basis are my numbers too small?
Great diagram- it's exactly what I experience with my CD systems, and my experiences track yours with the rest of it.
I am not dismissing it, but it just seems like the comment "Any spaciousness derived from such reflections are the spaciousness of the listening room, not the recording." is obvious since any room reflections are from the listening room and not the recording.
Maybe I missed what you are trying to say, but I read it as "Any audible room reflections are from the room not the recording." Doesn't that go without saying? - unless "room refections" is ambiguos here meaning both the recording venue reflections and the listening room reflections, but that is not how I read it.
Last edited:
There's always this one guy that ruins it for everybody. I'm close to tears
I find it interesting that nobody tried the test I had proposed.
I don't have the time to do tests like that, I am pretty sure you don't have a family and that makes a big difference. But I always appreciate your opinion as you have a knowledge and objectivity that few here have. But you do always seem to be on the extreme edge of what others believe and perceive.
I agree with you completely although there are many here who would not. I think summing localization, when taken as a single number average across the audio bandwidth, would be higher than 1 ms. and that imaging effects would likely still be significant above 3 ms. But, as I said, I agree with you in principle and arguing about 1.5 ms versus 1 ms does not seem fruitful.
It's hard to pull any single number for summing localization from Blauert, but in section 3.1 he quotes an "average" time interval for a displaced judgement in a test to be 2 ms. I'm not saying this IS the number nor does he say that, but it is significantly longer (double) what you were stating.
There is a lot of talk about phantom center and phantom image here. I'd like to know, what happens to more than one image between your speakers, if you move to one side - let's say the left.
This would be the scene if you sit in the middle between the speakers:
Now you move to the left. Do the musicians stay in their place? Or do they move as an entity towards the left speaker:
Are they compressed:
Are they blurred in any way:
Are they compressed and moved towards the left speaker:
Or are they even progressively compressed:
Or does it happen in a completely different way?
Rudolf
Rudolf,
The answer to any and all of your questions depend on the dispersion pattern of your speakers, their location in the room, toe in angle, room size, shape and reflectivity.
Art
Marcus,
I tried listening to mono pink noise at a level of about 78 dBC and shuffling my feet with my eyes closed across my room just in front of my couch listening location which puts one's head in the center of the room.
The speakers are corner wall mounted, toed in to point at the opposing room corner, room dimensions are 11 foot wide x 20 long x 8 & 7.5 high, 3.38 meters x 6.15 x 2.46 x 2.3.
No subjective off center "snapping" of the phantom image with this set up.
That said, what "snaps" to your ears might be a brush to mine..
When the speaker levels are balanced, the general "phantom location" of the noise appears to be directly in front of the listening location from left to right across the room.
The sense of location is not very identifiable other than being centered in front over the middle 1/3 of the left to right listening location.
When going left or right from the middle 1/3 the image appears to be more a vector sum, L/R locatable points "averaging" as one in between when outside the center 1/3 listening location.
This effect changes to a direct in front location when shouldered up to the side wall.
Reducing level of the right speaker 10 dB the only "balance point" was about 2 feet (.6 meter) in front of that speaker, any other position the left speaker was the obvious location of the noise.
Reducing level of the right speaker 5 dB the "balance point" moved to in front of that speaker at the original listening distance, about 8.5 feet (2.6 meters) back from the speakers, all the way to the wall.
Subjectively, listening to a mono spoken voice, same general perceptions.
Listening to stereo classical music, the general locations of recorded placements seem to stay relatively placed while moving across the center of the room in front of or sitting on the couch.
Towards the back of the room the sound is too diffuse to specifically locate recorded sources that are not hard panned left or right.
Art
Attachments
Last edited:
Thanks Art -now Markus can't complain anymore that we are ignoring him
As with all things, I am motivated to the theory and the mathematics and look to that as a handle on what will happen. I am not much for "listening tests" because they generally haven't worked out too well for me I guess.
Rudolf - my impression is that there is some compression of the stage width as one move laterally and that the center of this image moves slightly with the location. As above, I don't know how to represent this mathematically so I used the best that I could find which was the Blauert curves.
As with all things, I am motivated to the theory and the mathematics and look to that as a handle on what will happen. I am not much for "listening tests" because they generally haven't worked out too well for me I guess.
Rudolf - my impression is that there is some compression of the stage width as one move laterally and that the center of this image moves slightly with the location. As above, I don't know how to represent this mathematically so I used the best that I could find which was the Blauert curves.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.