I think you'll find if you look very closely at Left versus Left+Right, and Right versus Left+Right you'll see that their summation is less than 6dB, as in the case of my measurements.Yes the thing I find most odd is that the two seem to add pretty much in phase across the entire frequency range (due to the fact that dropping off one speaker results in a curve around 6db lower that follows the two speaker curve fairly closely). However when the phase is reversed on one, it would appear that the speakers are anything but in phase throughout the audio band.
This is bugging me rather a lot as the two things seem to be at odds
When measuring both speakers in phase, only the direct component of the signals which are in phase will add to a 6dB gain - the random phase contribution from the room will only add to a 3dB gain.
Thus if what you're measuring has some contribution from the direct field and some contribution from the room both of comparable amplitudes, you'll get Left and Right adding to more than a 3dB increase over an individual speaker but less than a 6dB increase. You would only see a full 6dB increase if both speakers phase tracked perfectly and there was no room contribution. (Eg anechoic chamber, or reflections windowed out)
Another way to look at it is if you were to measure the relative phase of Left and Right speakers together with their room contribution, (un-windowed, narrow band) you would see that the room contribution causes the phase response of the two channels to semi-randomly deviate from each other, even though the direct field from each speaker may be phase tracking accurately.
The greater the room contribution the more the deviation from them being in phase, and the more you tend towards non-coherant (power) addition.
Another thing to consider is that phase tracking 180 degrees out of phase is much more critical of small phase errors than phase tracking in phase.
If you're in phase to begin with a 20 degree phase error is very small and there will be very little change in amplitude, maybe 1dB at most. (A full 90 degrees is required for a 3dB drop, if they're equal in amplitude)
However if you're trying to track at 180 degrees a 20 degree error is a large error, and depending on how accurately the amplitudes are balanced, a 20 degree deviation could cause a 10dB or more change in amplitude.
This is why phase tracking errors always show up far more obviously when the phase is reversed, and the reason why you check the phase alignment of an in-phase crossover with one of the drivers reversed from normal - the cancellation that forms is a far more accurate indicator of time/phase alignment than looking for maximum output when correctly phased.
Thanks Simon, that makes sense. I've done some more experimenting tonight, played with speaker position a bit (widened slightly and toed in more) and also spent quite a while trying to find the ideal mic position. I've given up with a somewhat better result. The fact that my room is far from symetrical with one wall within about 1.5M of the left speaker and the other wall about 7M from the right speaker I think means I will be hard pressed to get a uniform result. It was quite interesting to see just how far the measurements varied just from moving the mic small amounts.
The first pic is the best result I managed. The second pic is all of the measurements I made tonight with slightly different mic positions (moving left and right probably a maximum of 5cm) as you can see the lower frequencies vary very little but after about 200Hz the differences start to show up.
tonight I've gone with 1/6th octave smoothing rather than 1/3rd
Tony.
The first pic is the best result I managed. The second pic is all of the measurements I made tonight with slightly different mic positions (moving left and right probably a maximum of 5cm) as you can see the lower frequencies vary very little but after about 200Hz the differences start to show up.
tonight I've gone with 1/6th octave smoothing rather than 1/3rd
Tony.
Attachments
Your first graph is quite interesting, above 500Hz or so there is a good fairly uniform ratio between in phase and out of phase, then below 500Hz it all turns to custard, which will be almost certainly your room assymetry with one speaker close to a side wall and the other not. (I see a smaller but similar effect with my slightly asymmetric side-wall spacing)
Of particular interest are the frequencies where the out of phase line actually goes higher than the in phase line - 180Hz, 100Hz, and 60Hz. What that means is at those frequencies the actual phase shift between left speaker and right speaker (inclusive of room reflections) when connected "in phase" is actually mostly out of phase - somewhere between 90 and 270 degrees, due to the room modes and boundary cancellation.
I just thought it looked pretty :
The one I think is most interesting is the 40hZ spike. around 15db difference between in phase and out of phase! But yes those out of phase higher than in phase frequencies are also very interesting.
I guess it is showing me what I knew all along, that my listening room sucks It is quite surprising that I do get an image, though I do have trouble trying to pinpoint locations. Its more like I can hear the trumpet up a bit higher and a bit to the rght, the piano seems lower down with low notes further to the left high notes further to the right. The singer sounds more forward but reasonably central, it's noticable there are differences in location, but they are not that precise.
I'll have to do some polar s using gated measurements, but it might be a couple of weeks before I get an opportunity.
Tony.
The one I think is most interesting is the 40hZ spike. around 15db difference between in phase and out of phase! But yes those out of phase higher than in phase frequencies are also very interesting.
I guess it is showing me what I knew all along, that my listening room sucks
It is quite surprising that I do get an image, though I do have trouble trying to pinpoint locations. Its more like I can hear the trumpet up a bit higher and a bit to the rght, the piano seems lower down with low notes further to the left high notes further to the right. The singer sounds more forward but reasonably central, it's noticable there are differences in location, but they are not that precise. I'll have to do some polar s using gated measurements, but it might be a couple of weeks before I get an opportunity.
Tony.
Last edited:
See raw response below.
Also, not my room so I can't comment on noise. The noise peak is at 46Hz, so no idea what it is. Other than that, very quiet. It's a residential neighborhood, typical US style.
Attachments
indeed WOW!
Sounds like Toole is having some symptoms of non-stereo hearing ! The sound is heard as coming from the speakers. I'm not alone. What a relief
Hear this Radugazon ! We are not alone ! Toole is with us !
Of course this cannot be said out loud in public since if you have some difficulties in hearing phantom images, generally you'll be condemned of having something wrong with your speakers or with your hearing
- Elias
“Originally Posted by a_tewinkel
we are left with what really annoys me about stereo: a relatively spatial set of phantom images created by both loudspeakers, and two "anchor" images created by the left and right loudspeakers playing solo. In some recordings we hear a whole string section emerging from a single loudspeaker. Not realistic, and not even pleasant. “
Much earlier in this thread I wrote about one’s ability clearly hear a single loudspeakers physical location in depth. Also, that this ability is caused by a complex radiation pattern giving “clues” that allow one to localize it’s physical location in depth. Without those clues, by radiating a simple radiation balloon, one obviously hears the direction the sound comes from but you cannot easily hear how far away the speaker is when your eyes are closed, it just comes from somewhere in front of you.
I had suggested several experiments which one can do at home which show this.
Once one has eliminated or reduced those “here I am” clues the speakers produce, the image from the speakers disappears leaving the one in the stereo image. It really is possible to make a very real phantom image or anywhere in between, but you also have to avoid close side reflections etc
Best,
Tom Danley
we are left with what really annoys me about stereo: a relatively spatial set of phantom images created by both loudspeakers, and two "anchor" images created by the left and right loudspeakers playing solo. In some recordings we hear a whole string section emerging from a single loudspeaker. Not realistic, and not even pleasant. “
Much earlier in this thread I wrote about one’s ability clearly hear a single loudspeakers physical location in depth. Also, that this ability is caused by a complex radiation pattern giving “clues” that allow one to localize it’s physical location in depth. Without those clues, by radiating a simple radiation balloon, one obviously hears the direction the sound comes from but you cannot easily hear how far away the speaker is when your eyes are closed, it just comes from somewhere in front of you.
I had suggested several experiments which one can do at home which show this.
Once one has eliminated or reduced those “here I am” clues the speakers produce, the image from the speakers disappears leaving the one in the stereo image. It really is possible to make a very real phantom image or anywhere in between, but you also have to avoid close side reflections etc
Best,
Tom Danley
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.