Fixing the Stereo Phantom Center

Administrator
Joined 2004
Paid Member
Very recently I added a sub to the center speaker, so I now have three of the same speaker stacks.
I'm loving it. It is a big improvement for both the central image and the overall sound field.
Thanks for the info, that seems like good news. :up: What you found agrees with some research I read years ago that found the importance of matching LCR speakers. The matching seemed more important than the quality of the speakers themselves. Surprisingly I have never been able test this myself either at home or with a large PA. In all these years I've never had a center speaker that was an exact match for the left and right. :(

Well I have had many times in cinema rigs, but that's different because the mix wasn't 2 channel stereo with a derived center, but an actual LCR mix.
 
Hi Simon

I have read about the trinaural setup on the AH forum also. I guess the shown system profits a lot from it because the huge horn setup is probably not able to generate the same sharp phantom center as smaller speakers can do (but it can do other things better than small speakers of course). When one of the visitors mentioned that with trinaural switched off he was not able to hear Lou Reed coming from an unambiguous phantom center and that he was hearing him from left center and right at the same time I must say that I experience it differently at home (i.e. I have just tried it with the same track). It is more like the contrary: It is veeeeeery difficult trying to not perceive him form between the middle of both loudspeakers when you sit in the sweetspot. I really tried to hear him coming from either loudspeaker but as long as I stay in the sweetspot it is almost impossible.
The poster assumed that his perception was due to the fact that Lou was originally recored mono and then added to both channels equally and that it was therefore impossible to get a sharp center image. But this is exaclty how the panning of a center phantom source is usually done in intensity stereophony ...........

Regards

Charles
 
Administrator
Joined 2004
Paid Member
because the huge horn setup is probably not able to generate the same sharp phantom center as smaller speakers can do
Interesting. I don't know the horn system on question, but I'm rather surprised because IME there has been no problem at all with a high quality phantom image coming from large horn systems. In fact quite the opposite, Like you I have a hard time hearing anything more than the center image, it's difficult to perceive it coming from either left or right - unless you are way off axis. It's a cool trick that two speakers can do that.

That's held true for me in rooms ranging from small domestic to medium size to a 3000 seat amphitheater.
 
Maybe because the person in question was a visitor and therefore his hearing not trained to the system as much as the owner's hearing the whole situation might change after some time.

P.S.: Here is a picture of the impressive horn system. The quality seems to be very high with excellent measurements.

https://abload.de/image.php?img=img_2969tneog.jpg
 
Last edited:
  • Like
Reactions: 1 user
Administrator
Joined 2004
Paid Member
Thanks for the photo, it looks like a very nice system. This leaves me rather puzzled, as just looking at the photo of the system I can't see why it should not present a strong and stable phantom image. FWIW, I have never, ever needed training on any system to hear a phantom image. If it's there, I hear it immediately.
But there is a recent thread by someone who simply does not hear it at all. Maybe this (or some degree of it) is more common than we think?
 
Member
Joined 2008
Paid Member
An interesting thing is, that the best phantom image I have heard so far was from a pair of Lowther DX2 backloaded horns in an equilateral triangle, slightly toed in, running with a cheap class A tube amp. In theory, there are much better speakers and amps, but these just work so well in my small room. Maybe it is the point source nature, since my synergy attempts do it very similar, but I do not have an identical pair yet. The 3D printed multicell horn was also awesome, but again, I have only one, so the other side used another much smaller horn. So maybe it is the lack of crossover in the critical band and defined directivity? I am now trying to build some 300 Hz sized horns with DCX464 and hope these will be fine, too, in this regard.
 
Administrator
Joined 2004
Paid Member
Using the trick described by pano, but in a very moderate way, an approach might be found to blur the stereo phantom to smooth this switching behavior a bit. So blur this phantom, but blur it only a little bit, and blur it a contolled bit. Say, blur it by an asymmetry of 1dB, which would broaden a stereo phantom to a width of some 20cm at a 2m listening range.
That would certainly do something, but there is a better way. Leave the amplitude alone and shuffle the phase. It works very well, but not perfect of course. @wesayso didn't like some of the phase shuffle artifacts he was hearing on his system, so is now using mid/side EQ instead. For me and my system the phase shuffle wasn't bad, so I used it.

FWIW, I don't listen with my head in a vice, and did find the tonal difference in the phantom center to be evident across a reasonable listening area. That might be room and system dependent, I don't know.
 
The matching seemed more important than the quality of the speakers themselves.
Thanks for the thumbs up, Pano.

I too think that matching probably makes the most difference.

I've had rock solid center images on every last pair of PA and synergy builds made for a half dozen years, and one of the steps in their builds was confirming equal transfers. (By rock solid, I mean both speakers running summed LR mono.)
 
  • Like
Reactions: 1 user
Administrator
Joined 2004
Paid Member
@mark100 I once heard Tom Danley speaking on mic at a presentation - pure mono. The PA was a pair of SH50s. I have never, ever heard a PA hang a phantom image in the air like that. Amazing - and hard to describe. They must have been very well matched. They did very well on recorded music, too.

Never thought about it, but maybe I'm someone for whom the phantom center easily clicks. It's an illusion that I enjoy and notice whenever it's particularly strong. My problem is hearing depth, but that's another topic.
 
Administrator
Joined 2004
Paid Member
Yes they do! I think the Valencias had the 811 horns, right?
At one time I had a pair of these goofy speakers in my living room with a 31" CRT between them, then an LCD. Great for TV sound, with one reservation. The sound was so much bigger, richer and more detailed than the image that there was something of a disconnect.

lounge_lizard.jpg
 
  • Like
Reactions: 1 user
@mark100 I once heard Tom Danley speaking on mic at a presentation - pure mono. The PA was a pair of SH50s. I have never, ever heard a PA hang a phantom image in the air like that. Amazing - and hard to describe. They must have been very well matched. They did very well on recorded music, too.

Never thought about it, but maybe I'm someone for whom the phantom center easily clicks. It's an illusion that I enjoy and notice whenever it's particularly strong. My problem is hearing depth, but that's another topic.
I can totally relate what you're saying about the Danley presentation. He so surely builds a better speakers than mine, and even mine can go spooky center.

I've been playing a trick on my kids when they've visited for first time since setting up my current LCR. I set them in a chair in the middle, and turn on just the left and right playing mono. So far, every one has asked 'why is the center speaker the only one playing?' !


I enjoy the center illusion too...perhaps a bit more than a soundstage.
What I really like is a broad sound stage that mimics a live stage with either left-right mains or/and a backline. And when lead vocals come from a center speaker.
I think i have the closest to that as of yet, with the 3 same stacks all having balls.. I'm really pretty happy listening almost any where in the room.
 
  • Like
Reactions: 1 user
Administrator
Joined 2004
Paid Member
What I really like is a broad sound stage that mimics a live stage with either left-right mains or/and a backline.
I was hoping the conversation would come around to this. The phantom image should not be just a pinpoint center, but a nice spread of sounds between the speakers. Some sound should be pinpoint dead center, but not all. Of course this is talked about endlessly in the Hi-Fi press and on forums. Different speakers certainly do this differently.

For whatever reason, I much prefer two speaker mono over single speaker for serious listening. In my lava cave listening room old mono recordings of orchestras and operas truly came to life. There was real depth and some separation of sections of the orchestra. That surprised me and made some recordings that I originally did care for a real delight to hear. Was that information in the recording, or an artifact of the speakers and room? I don't know. But lately I've been working with headphone virtualization, using crossfeed, delay, EQ room reflections, etc. The crosstalk, delay and EQ difference from the HRTF makes a big difference in perception of not only the space, but details in the music as well.
 
  • Like
Reactions: 1 users
Maybe because the person in question was a visitor and therefore his hearing not trained to the system as much as the owner's hearing the whole situation might change after some time.

P.S.: Here is a picture of the impressive horn system. The quality seems to be very high with excellent measurements.

https://abload.de/image.php?img=img_2969tneog.jpg
You`re completely correct. This system has excellent imaging as well in depth as in width even in stereo mode only and Lou Reed is IMHO perfectly localized. With the trinaural setup the imaging in all dimensions gets even better. We have tried different angles (factors) for the relation of output level of the outer stereo speakers/inner "trinaural" speakers and got very interesting results. All the measurements were taken with the software "Acourate" of Dr. Ulrich Brüggemann. I am the owner of the system.

Regards

Holger
 

Attachments

  • Uncorrected (pure) frequency response and target curve of 6-way-hornsystem 15 - 44100 Hz with ...jpg
    Uncorrected (pure) frequency response and target curve of 6-way-hornsystem 15 - 44100 Hz with ...jpg
    201.9 KB · Views: 70
  • Step response trinaural setup 6-way-hornsystem.jpg
    Step response trinaural setup 6-way-hornsystem.jpg
    215.2 KB · Views: 59
  • ETC first 150 ms (marked first 20,99 ms) 6-way-hornsystem with supplementing trinaural 2-way-h...jpg
    ETC first 150 ms (marked first 20,99 ms) 6-way-hornsystem with supplementing trinaural 2-way-h...jpg
    272 KB · Views: 64
  • IMG_2976.JPG
    IMG_2976.JPG
    401.1 KB · Views: 70
  • Like
Reactions: 1 users
I was hoping the conversation would come around to this. The phantom image should not be just a pinpoint center, but a nice spread of sounds between the speakers. Some sound should be pinpoint dead center, but not all. Of course this is talked about endlessly in the Hi-Fi press and on forums. Different speakers certainly do this differently.

I think so too. Maybe it's from liking the sound of small live bands, where such a sound field and center image rather naturally occur.
For whatever reason, I much prefer two speaker mono over single speaker for serious listening. In my lava cave listening room old mono recordings of orchestras and operas truly came to life. There was real depth and some separation of sections of the orchestra. That surprised me and made some recordings that I originally did care for a real delight to hear. Was that information in the recording, or an artifact of the speakers and room? I don't know. But lately I've been working with headphone virtualization, using crossfeed, delay, EQ room reflections, etc. The crosstalk, delay and EQ difference from the HRTF makes a big difference in perception of not only the space, but details in the music as well.
Yep, I've always felt serious evaluation is best mono single speaker, then mono two speaker, and last stereo. By the time stereo kicks in, all bets are off as far imo.
Not to say stereo can't be extremely pleasing, it's just stereo seems so hit or miss compared to more reliably pleasing two speaker mono.
Interestingly, until I adding the sub to the center channel, 3 channel mono was often winning on hard hitting stuff like AC/DC Thunderstruck.
But now, with the center sub, Thunderstruck hits so hard with a simplified Gerzon matrix in place, it makes anyone jump and laugh....or flat run haha!

Cool what you're doing with headphone virtualization. After i've exhausted playing with various LCR techniques, I'll need to study what guys like you and wesayso have been doing...
 
  • Like
Reactions: 1 users
You`re completely correct. This system has excellent imaging as well in depth as in width even in stereo mode only and Lou Reed is IMHO perfectly localized. With the trinaural setup the imaging in all dimensions gets even better. We have tried different angles (factors) for the relation of output level of the outer stereo speakers/inner "trinaural" speakers and got very interesting results. All the measurements were taken with the software "Acourate" of Dr. Ulrich Brüggemann. I am the owner of the system.

Regards

Holger
Awesome Holger !!!

Would love to know what angles and the settings for them you've gravitated to.
Do you use any frequency dependence in your settings?
That's the next step i have planned for playing with summing/subtracting matrices on my LCR.

edit: so i just went to see why Lou Reed has been mentioned several times, and listened to Dirty Blvd on the New York album.
What a piece of dung, as far as a stereo recording goes...incredible vocal wandering. Yikes!

My most aggressive matrix, as far as moving content to the center speaker, was best at localizing Reed to center. Very nice really.
Surprisingly, all three speaker running summed left & right mono, did an equally well job of localizing Lou to center,
but of course that lost any sense of instrumental stereo separation which was still present in the LCR matrix.
 
Last edited:
Administrator
Joined 2004
Paid Member
@mefistofelez That was a really nice Selenium 15" driver that isn't made any more. :( It was the least expensive 15 from Selenium at the time, but was such a good OB woofer. Fairly low FS, Qts near 0.7 and it worked so well, sounded so good. Of course I don't remember the model number. Maybe 15W500
The 15W700 PRV woofer seems close to the same, but I have not seen it in person. One of my Seleniums got crushed when we were moving house, it was a sad day.
loungelizback.jpg
 
  • Like
Reactions: 1 user