Meanwhile.. Back at the Stereo Phantom Center thingie.. 
Has this been tested with BMR's or ESL's?
I built a pair of BMR's and haven't noticed it much. With my standard cone type builds, it's really noticeable.. (2 spaced speaker units adjusted to a sweet spot)
Could the "beaming" of the cones have something to do with it? I've experienced it in mono also..
Has this been tested with BMR's or ESL's?
I built a pair of BMR's and haven't noticed it much. With my standard cone type builds, it's really noticeable.. (2 spaced speaker units adjusted to a sweet spot)
Could the "beaming" of the cones have something to do with it? I've experienced it in mono also..
A phase coherent loudspeaker with multiple drivers can be engineered but they are still both individual units separate from each other. Even high end speaker units have this condition (The center channel thingie) and I'm sure the phase coherent factor was considered in the design, or should have been..
Would a single full range cone driver be coherent on its own? (Beyond the baffle step compensation issue) I have many single driver units (Full range) and can notice the condition..
On the other hand, my BMR's didn't have this. The sweet spot was very wide with no focus of sound anywhere between them. The stereo factor was still there, being different instruments could be heard in the different channels. (L-R)
I made an opportunity to go to a high end speaker shop to experience some ESL's. That was the first thing I noticed by walking around the front of the staging area. Huge sweet spot with no focus of sound or frequency anywhere..
The conclusion that this noob came up with is the difference of the drivers. One had cones and the other didn't. The beaming was gone in the units without cones. The SPL was less but the sound stage was big with no major mix in between. A compromise, I guess..
Just my thoughts..
Would a single full range cone driver be coherent on its own? (Beyond the baffle step compensation issue) I have many single driver units (Full range) and can notice the condition..
On the other hand, my BMR's didn't have this. The sweet spot was very wide with no focus of sound anywhere between them. The stereo factor was still there, being different instruments could be heard in the different channels. (L-R)
I made an opportunity to go to a high end speaker shop to experience some ESL's. That was the first thing I noticed by walking around the front of the staging area. Huge sweet spot with no focus of sound or frequency anywhere..
The conclusion that this noob came up with is the difference of the drivers. One had cones and the other didn't. The beaming was gone in the units without cones. The SPL was less but the sound stage was big with no major mix in between. A compromise, I guess..
Just my thoughts..
With my cone speakers, I can adjust to a distance of ~10-12 feet away to a point of where both projected sound(s) merge to create a hot spot, per se.. (Phantom Image?) Slight adjustments of both speaker unit (L&R) changes the hot spot from one point to two at the same distance away, then the focal point is set further away. (Moving left or right to find the focal point of each individual speaker unit at the original distance) All cone type speakers, IMO, have a focal point or "Beam" pattern, being the nature of a cone..
"Did the BMR or ESL for that matter produce a believable phantom image? The "no focus of sound anywhere" part makes me confused."
My BMR's are great and they don't have the beaming like a cone driver. Flat against the wall, toed in towards a sweet spot.. It doesn't matter. You need to be very far off axis to notice any drop in the sound stage between the two. Of course, the stereo effect is best centered between the two, being the distance change from one driver to the other moving left or right. Running them in mono and walking across the sound stage from left to right is good. No hot spots and no focus/increase of sound. This is the same effect that I observed in my experience with ESL's..
In short.. No. The sound pattern projected from a BMR or an ESL is totally different compared to a cone driver and the sound doesn't add or distort as they mix in the center.. (In my observations)
Sorta like comparing an isotropic radiator to a 3 element yagi. (The BMR being isotropic and the cone being more directional/focused) If sound waves are more dispersed, the less they collide.. (In intensity)
Just some more noob blather..
"Did the BMR or ESL for that matter produce a believable phantom image? The "no focus of sound anywhere" part makes me confused."
My BMR's are great and they don't have the beaming like a cone driver. Flat against the wall, toed in towards a sweet spot.. It doesn't matter. You need to be very far off axis to notice any drop in the sound stage between the two. Of course, the stereo effect is best centered between the two, being the distance change from one driver to the other moving left or right. Running them in mono and walking across the sound stage from left to right is good. No hot spots and no focus/increase of sound. This is the same effect that I observed in my experience with ESL's..
In short.. No. The sound pattern projected from a BMR or an ESL is totally different compared to a cone driver and the sound doesn't add or distort as they mix in the center.. (In my observations)
Sorta like comparing an isotropic radiator to a 3 element yagi. (The BMR being isotropic and the cone being more directional/focused) If sound waves are more dispersed, the less they collide.. (In intensity)
Just some more noob blather..
This can/does happen with a panned monaural signal, but does it happen with an ORTF pair (17cm spacing) or a spaced pair 3-6 meters apart (where there is no coherent phasing from the source)? If it does I've never noticed it. Perhaps this is simply an artifact of studio mixes?
Some folks here (Elias, CLS, and others) have played extensively with mixing L&R in all sorts of analog matrices and put up a third center channel derived from L & R, and even altered the L and R channels based on the matrix of the day (Nagaoka, etc). Does this help the situation for a more stable and coherent center image?
It might in some cases, but matrix extraction of a center channel works far better with studio mixes that place a phase coherent mono signal in both channels and not so much with "stereo pair" mic placements. So it's very dependent on the recording . . .
Pano, I remembered reading a passage related to this in Toole's book. Here it is. Looks like I was on to something.
“Listeners appeared to prefer the sound from wide-dispersion loudspeakers with somewhat colored off-axis behaviour to the sound from a narrow-dispersion loudspeaker with less colored off-axis behavior. In the years since then, it has been shown that improving the smoothness of the off-axis radiated sound pushes the subjective ratings even further up, so it is something not to be neglected. Perhaps related to this is the acoustical crosstalk associated with the phantom center image. This coloration cannot be ignored in a situation where the direct sound is strong. Early reflections from different directions tend to fill the interference dip, making the spectrum more pleasantly neutral......Not to be ignored in any situation in which reflected sounds have been removed is the fact that the acoustical crosstalk that plagues stereo phantom images is present in its naked ugliness, without any compensation from reflected sounds"
These findings also indicated users had lower preference ratings for Earl's preferred tight beamwidth approach.
It also indicates that damping the first side wall reflection makes the phantom center crosstalk worse.
Dave
DDF, doesn't that sum up what was written in the paper linked in the first post?
It stated something along the lines of what you're saying here. Lack of early reflections making the phantom centre more dull sounding. The proposed shuffler looks a lot like introducing fake reflections at higher frequencies to fix that.
From what I have read this was already addressed in that original paper. One of the reasons I tried it as I reduced the level of early reflections with damping panels in my room and recognised what Pano said about a duller sound in the centre.
That paper also stated that the centre is possibly pushed further away after applying the fix. I can confirm that it did exactly that. If you have more early reflections than this fix is probably not needed as much.
It stated something along the lines of what you're saying here. Lack of early reflections making the phantom centre more dull sounding. The proposed shuffler looks a lot like introducing fake reflections at higher frequencies to fix that.
From what I have read this was already addressed in that original paper. One of the reasons I tried it as I reduced the level of early reflections with damping panels in my room and recognised what Pano said about a duller sound in the centre.
That paper also stated that the centre is possibly pushed further away after applying the fix. I can confirm that it did exactly that. If you have more early reflections than this fix is probably not needed as much.
Thanks Dave for finding that. It does seem to agree with the paper by Earl Vickers. Vickers cites David Clark for much of this. As Wesayso has pointed out, this is what the paper says.
From the paper "Fixing the Phantom Center: Diffusing Acoustical Crosstalk"
Earl Geddes asks (and rightly so) if it necessary to correct for this in two channel recordings, as it should have already been fixed in the mix. I don't think there is a strict yes or no answer. It depends. It's going to depend a lot on the amount of early reflections your room/speaker combo has, and some on how the recording was done. I would guess that recordings mastered in mono, or those made for a center channel would benefit the most. I can say that most of the mono and stereo radio shows and dramas I have certainly benefit from this.
On the system and room I had in North America the dull center effect was quite pronounced. Big horns, room treatments, heavy carpet, high ceiling. All of that led to a tight, controlled sound with a high direct/reflected ratio.
My new system is basically dipole and in a very dry acoustic. The dull center is pronounced there, too. But the phase shuffler does not seem to fill in the hole as well.
From the paper "Fixing the Phantom Center: Diffusing Acoustical Crosstalk"
Indeed the impulse they propose is a short, multi-tap echo that shuffles phase and fills in where early reflections are absent. I produced a different version that is less colored, I think.
Earl Geddes asks (and rightly so) if it necessary to correct for this in two channel recordings, as it should have already been fixed in the mix. I don't think there is a strict yes or no answer. It depends. It's going to depend a lot on the amount of early reflections your room/speaker combo has, and some on how the recording was done. I would guess that recordings mastered in mono, or those made for a center channel would benefit the most. I can say that most of the mono and stereo radio shows and dramas I have certainly benefit from this.
On the system and room I had in North America the dull center effect was quite pronounced. Big horns, room treatments, heavy carpet, high ceiling. All of that led to a tight, controlled sound with a high direct/reflected ratio.
My new system is basically dipole and in a very dry acoustic. The dull center is pronounced there, too. But the phase shuffler does not seem to fill in the hole as well.
Pano, it looks like the 2 ways to ameliorate this and identified in this thread are to add side wall reflections or the shuffler. Have you seen any common application of the shuffler in recording? I haven't. I think pragmatically, if that method is chosen, playback is by far the best option.
I don't think it's possible to fix in the mix adding artificial side wall cues: too artificial. 7-channel presents some interesting possibilities though!
Tangentially related, early reflections increase intelligibility but it's through adding more "near direct" sound level, above the room din. In these cases, the trade off is tonal accuracy. Way back when I spec'ed and had built teleconferencing rooms, I used to spec a hard ceiling. This is perhaps the worst reflection point (vertical) for timbre distortion, but it significantly increases intelligibility back in the room. So I'm not sure it's legitimate to apply this same principle to stereo reproduction under the premise of enhanced intelligibility.
OTOH as a another data point, it's well known that increased vocal intelligibility is achieved through boosting response ~ 3 kHz (see Yamaha sound reinforcement handbook, others). Maybe reducing the crosstalk notch (which apparently lands ~ 3 kHz) via sidewall reflection is a similar effect?
I don't think it's possible to fix in the mix adding artificial side wall cues: too artificial. 7-channel presents some interesting possibilities though!
Tangentially related, early reflections increase intelligibility but it's through adding more "near direct" sound level, above the room din. In these cases, the trade off is tonal accuracy. Way back when I spec'ed and had built teleconferencing rooms, I used to spec a hard ceiling. This is perhaps the worst reflection point (vertical) for timbre distortion, but it significantly increases intelligibility back in the room. So I'm not sure it's legitimate to apply this same principle to stereo reproduction under the premise of enhanced intelligibility.
OTOH as a another data point, it's well known that increased vocal intelligibility is achieved through boosting response ~ 3 kHz (see Yamaha sound reinforcement handbook, others). Maybe reducing the crosstalk notch (which apparently lands ~ 3 kHz) via sidewall reflection is a similar effect?
Last edited:
I've never seen it used or talked about in recording. I was mostly a live sound guy, anyway. I realized this morning that the effect has caused me problems in live sound, and I didn't know it. I fixed it by other means, tho.
Seems to me that some clever fellow at the major manufacturers could easily implement a phase shuffler (mine, for example) into a home theater receiver DSP for Phantom Center Mode. It could do a lot to clear up dialog when running without a real center speaker.
Seems to me that some clever fellow at the major manufacturers could easily implement a phase shuffler (mine, for example) into a home theater receiver DSP for Phantom Center Mode. It could do a lot to clear up dialog when running without a real center speaker.
I think its important to understand a few things about the room characteristic that Toole use to do his testing. When you look at the testing room of harman using jbl lsr6332, it may explain why he prefer a room without first reflection treated.
the standard is to have early reflections within 15 ms be at least 10dB down and ideally 20 db down. The AES/EBU/ITU recommendations for early reflections, defined as the first 15 ms after the initila signal, is at least 10 dB down, from 500 Hz up.
in a room like the one from harman which is maybe 30 feet wide, no absorption on the side walls may work because the side wall are so far away from the listening position that the side walls reflection may be 10db down at LP. but in any normal rooms and especially small rooms, letting the first reflection points untreated will badly compromise the sound.
Also, ''Floyd spoke of his studies where the lateral reflections fell into the subjective category. It was speaker, listener and content dependent. His studies also showed that speakers with smooth off-axis response were 100% selected as preferred. ''
I think one should test this himself and try in their room first reflection treated and untreated. For me, I found that first reflection of side wall increased spaciousness and intelligibility and clarity very evidently.
Last edited:
https://www.gearslutz.com/board/stu...erstanding-precedence-effect-etc-shift-2.html
"Toole has said himself that it varied with types of music whether sidereflections were preferred or not. The results were not very conclusive"
"Toole has said himself that it varied with types of music whether sidereflections were preferred or not. The results were not very conclusive"
?
Last edited:
do you mean what is the distance of the side walls first reflections points from my listening position?
imo, pano. if you dont have at least a free reflective zone at the listening position which means to at least absorb early reflections, theres no way you can have good imaging and phantom center.
the difference when treating early reflection is exactly clarity of image, intelligibility, the room sounds much bigger, and inner details are greatly improved. the stage grows way outside the speakers in a very evident matter. treating the early reflections made a world of difference in my room.
I now can hear so much how important treating a room is. I recently added treatment in my room and removed temporarily the side walls absorption, and the sound becomes for me almost unintelligible. everything gets mushed, clarity is non existent as the room is adding way too much reflections.
Last edited:
youknowyou, look at Pano's thread about his cave. I'm jealous about the behaviour in that space. I do agree about bringing the first reflection down, I tried to get a clean ~ 20 ms after the main pulse as much as possible myself. Not that easy in a living room but not completely impossible either. As long as you don't look too closely at 1/3rd octave filtered ETC plots .
But I can subscribe to what you mention, opening up of the stage. Without turning our living room into a complete treated studio I could get quite reasonable results (with a little processing help). Your wave guided speakers probably help immensely with that.
Could you show an ETC?
.But I can subscribe to what you mention, opening up of the stage. Without turning our living room into a complete treated studio I could get quite reasonable results (with a little processing help). Your wave guided speakers probably help immensely with that.
Could you show an ETC?
Im not sure if I understand pano cave LOL! does he have that kind of space in his house??? if he do, I personally think that this must be a very reflective space.youknowyou, look at Pano's thread about his cave. I'm jealous about the behaviour in that space. I do agree about bringing the first reflection down, I tried to get a clean ~ 20 ms after the main pulse as much as possible myself. Not that easy in a living room but not completely impossible either. As long as you don't look too closely at 1/3rd octave filtered ETC plots
But I can subscribe to what you mention, opening up of the stage. Without turning our living room into a complete treated studio I could get quite reasonable results (with a little processing help). Your wave guided speakers probably help immensely with that.
Could you show an ETC?
Isnt pano room is this http://www.diyaudio.com/forums/gallery/showphoto.php/photo/9975
Wesayo, I wonder how created a free reflective zone without absorption.
Unless your talking about LEDE design which still would be very WAF unfriendly...
once I created a free reflective zone, it totally transformed my system and I realized how much flaws a untreated room creates and how simply treating my room removed all my unsatisfaction about my system. Id say that the room is the most important and the most detrimental to good sound along badly designed speaker. the amount of inner details, imaging and clarity that gets lost due to early reflection is very obvious and honestly dramatic once you treat those reflections.
Ive became quite passioned about room acoustics and looking to build RPG bad panels
Last edited: