Tips on phantom center? UST use case

[tifune]

Recently picked up an UST projector, and now there's no going back (per the kids). I didn't realize, though, how difficult it is to place a center speaker with these. ALR screens aren't audibly transparent and the projector needs to be on the floor or a short table. Ceiling mount is an option, but prefer to avoid it.

The way I see it, my options are:

-Center channel located in center height position, probably pointed directly at LP but I could do something silly like upside-down Dolby if there's a chance it would give the impression dialogue is coming from screen

-Center underneath projector pointed up at LP - not totally sure if I would absorb or ignore floor bounce, probably depends on the speaker but I have hardwood floors so its a consideration

-Improve phantom center - it's pretty good as-is with my R3's, just wondering what I can do to improve it (if anything). a lot of searching suggests it seems to be mostly about placement and toe in/out but maybe there's something I'm missing? This is actually my preference as I generally don't care for the way center channel sounds when upmixing with Auro3D. My only hesitation here is, in some media I don't think my Denon and/or HTPC is downmixing correctly.

Over the weekend I watched a live NIN DVD, during what should have been a pretty slick guitar solo all I could hear was the room reverb of the guitar and almost none of the direct. I am bitstreaming with Center channel set to "none" on the Denon so not sure what other setting(s) I may be missing.

I'm also not opposed to buying something dedicated to the task; the wife would appreciate if it was gloss black but that's about the only externality. Maybe a CBT would help, for example?

 

[Golfx]

just watched an audioholics youtube in which the center speaker was presented as THE dominant speaker (by a good margin) used by studio sound mixes for multichannel movies and streamed tv media. So I would buy a good center channel and experiment on placement. Mine is high and I don’t really notice it’s there once I start watching. I too use auro3d for some music and the center plays a large part in that as well. I don’t know what your budget is but I would spend as much as you can comfortably afford. Polk’s Legend L400 center reviewed very well on audioholics. It’s priced at $1700.00 new. Others by Focal and Revel tested as good or better. Good luck. (I love shopping for speakers.)

 

[Spocko]

Get the Sony HT-A9H - it's designed to be used without a physical center channel and its phantom center was specifically optimized for setups like yours.
Yes, I'll be getting these myself!

 

[Duke]

Improve phantom center - it's pretty good as-is with my R3's, just wondering what I can do to improve it (if anything). a lot of searching suggests it seems to be mostly about placement and toe in/out but maybe there's something I'm missing? This is actually my preference as I generally don't care for the way center channel sounds when upmixing with Auro3D.

Controlled-pattern loudspeakers set up with axes criss-crossing in front of the center listening positions can give a good phantom center image even for listeners well off to either side of the centerline. This is because the ear/brain system’s two localization mechanisms – arrival time and intensity – somewhat balance one another out. The near speaker obviously “wins” arrival time., but the far-side speaker “wins” intensity because the listener is pretty much on-axis of that speaker but well off-axis of the near speaker. The key to this working well is, the near speaker’s output must fall off smoothly and fairly rapidly as we move off-axis. Earl Geddes, who I learned this technique from, has mentioned both 90-degree and 60-degree (-6 dB @ 45 and 30 degrees off-axis, respectively) radiation patterns as being well- suited for this application.

Ime phantom center mode with such speakers seems to give better soundstage depth to center vocalists on music videos, relative to having a dedicated center channel speaker. Also, if a dedicated center channel speaker is not identical to the left and right speakers, the timbral qualiy of center channel sounds can be noticeably different from those delivered by the left or right speaker. On the other hand, a cross-firing setup geometry may not result in a good phantom center image for listeners who have a hearing imbalance (better hearing in one ear than the other).

Here are a couple of imo relevant links:

http://libinst.com/PublicArticles/Setup of WG Speakers.pdf

https://www.audiocircle.com/index.php?topic=100196.msg1010455#msg1010455

Disclaimer: I've been building speakers designed for Geddes-style time/intensity trading since 2005.

 

[alex-z]

Over the weekend I watched a live NIN DVD, during what should have been a pretty slick guitar solo all I could hear was the room reverb of the guitar and almost none of the direct. I am bitstreaming with Center channel set to "none" on the Denon so not sure what other setting(s) I may be missing.

Do you have acoustic treatment for your room? That is the most direct way to improve sound quality. Even just basic stuff like an area rug below the centre channel and a couple ceiling panels can make a noticeable different. Most people with projector screens are just putting their centre channel below and tilting it up as you mentioned.

What centre channel are you using? A third R3 would be ideal. The R2C isn't bad, but doesn't seem to measure as well as the R3.

 

[DVDdoug]

Over the weekend I watched a live NIN DVD, during what should have been a pretty slick guitar solo all I could hear was the room reverb of the guitar and almost none of the direct. I am bitstreaming with Center channel set to "none" on the Denon so not sure what other setting(s) I may be missing.

I assume the DVD has a stereo track? If you get the same distant-reverb sound you know it's just the mix on this particular DVD. Or if you have an extra speaker lying-around you can temporarily try a center speaker.

I'm not 100% sure, but I would expect "none" to "correctly" re-route the center to left and right. I assume you've watched some movies with surround? It should be obvious if the main dialog is missing (or greatly attenuated).

 

[GalZohar]

If your receiver is trying to play on a non-existent center channel, it should be pretty obvious. At least in movies you'll hardly hear anything at all because 80% of the sound comes from the center channel.

 

[DavidMcRoy]

-Center channel located in center height position, probably pointed directly at LP but I could do something silly like upside-down Dolby if there's a chance it would give the impression dialogue is coming from screen

Can you reference a source for this “upside-down Dolby” technique? It sounds like it might be what I stumbled on: my center channel speakers array is similar to the Magnepan-championed “Tri-field Center” arrangement, with 3 dedicated center channel speakers, independent from the main front left and right channel speakers.

There is a speaker as close as possible to the left and right edge of the screen, and a third speaker placed as close as possible to the center of the top of the screen. In the case of a conventional speaker with a tweeter on top, that one “may be” placed upside down if necessary to get good time alignment and optimize dispersion characteristics.

I tried that method with Magnepans, and now use it with JBL 305s. (See photo attachment) I get best results by adding additional delay to the the top-center speaker to tune-out some room effects in the bass and lower midrange created when the three speakers’ output is combined acoustically as observed from the listening position. I then fine-tune global EQ on the center channel for combined flat response at the sweet spot. The results are the most solidly-focused phantom center image I’ve ever encountered anywhere.

Recently picked up an UST projector, and now there's no going back (per the kids). I didn't realize, though, how difficult it is to place a center speaker with these. ALR screens aren't audibly transparent and the projector needs to be on the floor or a short table. Ceiling mount is an option, but prefer to avoid it.

The way I see it, my options are:

-Center channel located in center height position, probably pointed directly at LP but I could do something silly like upside-down Dolby if there's a chance it would give the impression dialogue is coming from screen

-Center underneath projector pointed up at LP - not totally sure if I would absorb or ignore floor bounce, probably depends on the speaker but I have hardwood floors so its a consideration

-Improve phantom center - it's pretty good as-is with my R3's, just wondering what I can do to improve it (if anything). a lot of searching suggests it seems to be mostly about placement and toe in/out but maybe there's something I'm missing? This is actually my preference as I generally don't care for the way center channel sounds when upmixing with Auro3D. My only hesitation here is, in some media I don't think my Denon and/or HTPC is downmixing correctly.

Over the weekend I watched a live NIN DVD, during what should have been a pretty slick guitar solo all I could hear was the room reverb of the guitar and almost none of the direct. I am bitstreaming with Center channel set to "none" on the Denon so not sure what other setting(s) I may be missing.

I'm also not opposed to buying something dedicated to the task; the wife would appreciate if it was gloss black but that's about the only externality. Maybe a CBT would help, for example?

 

[NTK]

I am totally speechless after reading about the "Tri-Center" at the Magnepan site.
https://www.magnepan.com/tri-center

Here at ASR we talk about how compromised it is using a horizontal MTM as center channel speaker. Magenpan wants people to use 3 separate speakers, one on each side of the screen, and one on top (or, I guess, bottom is OK too)?

I'll ignore them saying home theatre is "inferior" to 2-channel. But they aren't aware of the concept of comb filtering? Here is a little diagram on the radiation pattern of three speakers spanning 1.5 m (0.75 m spacing between each) at 1 kHz.

 

[DavidMcRoy]

I am totally speechless after reading about the "Tri-Center" at the Magnepan site.
https://www.magnepan.com/tri-center

Here at ASR we talk about how compromised it is using a horizontal MTM as center channel speaker. Magenpan wants people to use 3 separate speakers, one on each side of the screen, and one on top (or, I guess, bottom is OK too)?

I'll ignore them saying home theatre is "inferior" to 2-channel. But they aren't aware of the concept of comb filtering? Here is a little diagram on the radiation pattern of three speakers spanning 1.5 m (0.75 m spacing between each) at 1 kHz.

View attachment 153801

You think comb filtering doesn’t happen with 2 channels? A computer simulation of that would look like something from a Busby Berkeley musical, too.

Also, note (in my post) that I tweak delay in a way that deviates from the norm.

We can get all theoretical and “purist” all we want, but stereo is an illusion, no matter how many channels are involved. And if something works, it works.

 

[NTK]

You think comb filtering doesn’t happen with 2 channels?

You have 3 speakers outputting the exact same signal. Totally different scenario.

 

[DavidMcRoy]

You have 3 speakers outputting the exact same signal. Totally different scenario.

Again, with my technique, the timing is different and the nature of the comb filtering is different because of that. And the speakers are not lined up in a row, so your perception of the sound from the top center speaker is further altered.

And for phantom images, 2-channels can have 2 speakers outputting the exact same signal. That’s how all this works. What is your point?

 

[NTK]

Again, with my technique, the timing is different and the nature of the comb filtering is different because of that. And the speakers are not lined up in a row, so your perception of the sound from the top center speaker is further altered.

And for phantom images, 2-channels can have 2 speakers outputting the exact same signal. That’s how all this works. What is your point?

Those in the know are fully aware of the 2-channel 2 kHz phantom center dip. That's why we want one single center channel speaker.
https://www.audioholics.com/room-ac...ons-human-adaptation/what-do-listeners-prefer

 

[DavidMcRoy]

Those in the know are fully aware of the 2-channel 2 kHz phantom center dip. That's why we want one single center channel speaker.
https://www.audioholics.com/room-ac...ons-human-adaptation/what-do-listeners-prefer

View attachment 153806

I know. I was a broadcast audio engineer for 30 years.

Did you not read the part where I said that I apply corrective global EQ for the 3 center channel speakers?

 

[Duke]

Again, with my technique, the timing is different and the nature of the comb filtering is different because of that. And the speakers are not lined up in a row, so your perception of the sound from the top center speaker is further altered.

Your three-speaker center channel with the delayed high-center speaker is very creative and intriguing.

One of the issues I think I've heard with a (single) center channel speaker is, that depth in the center of the soundstage seems to extend no deeper than the location of the center-channel speaker (with music videos), and I find that distracting. I close my eyes and yup the sound is obviously coming from right there. But in phantom center mode, it seems to me that the center of the soundstage extends much deeper.

Of course I could be imagining it.

Anyway, have you noticed that your three-speaker center channel with the delayed high-center speaker has better soundstage depth than a single center channel speaker?

And if you don't mind me asking, how much delay do you use?

Thanks!

 

[DavidMcRoy]

Your three-speaker center channel with the delayed high-center speaker is very creative and intriguing.

One of the issues I think I've heard with a (single) center channel speaker is, that depth in the center of the soundstage seems to extend no deeper than the location of the center-channel speaker (with music videos), and I find that distracting. I close my eyes and yup the sound is obviously coming from right there. But in phantom center mode, it seems to me that the center of the soundstage extends much deeper.

Of course I could be imagining it.

Anyway, have you noticed that your three-speaker center channel with the delayed high-center speaker has better soundstage depth than a single center channel speaker?

And if you don't mind me asking, how much delay do you use?

Thanks!

l happen to be using distance rather than time for the delay, however the time is calculable.

In my particular room, the top center speaker is 256 cm from the listening position and the flanking center channel speakers on either side of the screen are positioned roughly a meter lower and are 225 cm away. Setting the “distance” of the top speaker at 185 cm instead of 256 cm puts things at the sweet spot. That’s 207 m/s, I think? [EDIT: 2.07 m/s]

This ratio isn’t set in stone, because of room effects, so I adjust it by ear, listening for minimal coloration from reinforcements and cancelations in the lower midrange and upper bass. Once I settle on a delay, I EQ the center channel so that the combined output of the 3 speakers at the listening position is as flat as desired.

I’m sure this has been done before, but I’ve never come across it anywhere. It’s possible that Magnepan does this in their installations without my having read anywhere how it’s done. I stumbled on it by accident while mucking with delays and listening to pink noise and looking at it on a spectrum analyzer. Note that personal preference is going to dictate the delay. I happen to like a clearly focused and uncolored voice sound, but altering the EQ by creating a 3 or 4 dB dip centered at around 3500Hz with a Q around 2.5 or so enhances depth for me personally. I go for neutral response instead, since that’s more important to me than an illusion of greater apparent depth.

 

[Duke]

In my particular room, the top center speaker is 256 cm from the listening position and the flanking center channel speakers on either side of the screen are positioned roughly a meter lower and are 225 cm away. Setting the “distance” of the top speaker at 185 cm instead of 256 cm puts things at the sweet spot.

Thank you!!

If I'm understanding correctly: The top center speaker is 256 - 225 = 31 cm further way than the flanking center speakers.

Presumably the flanking center speakers have their "distance" set at 225 cm, and the top center speaker has its "distance" set at 185 cm, resulting in another 225 - 185 = 40 cm worth of electronic delay, for a total of 31 + 40 = 71 cm worth of delay.

Does this sound right to you?

If so, the arrival time delay for the top center speaker, relative to the flanking center speakers, should be 71 centimeters divided by 34,300 cm/second = 2.07 milliseconds. Unless I made a mistake somewhere.

How is the depth of image in the center of the soundstage? Is it consistent from one recording to the next, or does it change from one recording to another?

 

[DavidMcRoy]

Thank you!!

If I'm understanding correctly: The top center speaker is 256 - 225 = 31 cm further way than the flanking center speakers.

Presumably the flanking center speakers have their "distance" set at 225 cm, and the top center speaker has its "distance" set at 185 cm, resulting in another 225 - 185 = 40 cm worth of electronic delay, for a total of 31 + 40 = 71 cm worth of delay.

Does this sound right to you?

If so, the arrival time delay for the top center speaker, relative to the flanking center speakers, should be 71 centimeters divided by 34,300 cm/second = 2.07 milliseconds. Unless I made a mistake somewhere.

How is the depth of image in the center of the soundstage? Is it consistent from one recording to the next, or does it change from one recording to another?

That’s correct. I think I misplaced a decimal. LOL

Center image depth does vary with the recording. It’s dictated by the spectral frequency balance of the phantom center and/or any delay to it that may have already been applied in production. That’s something mixing engineers manipulate all the time. Our brains are conditioned to perceive a familiar sound as being more distant with increasingly rolled-off high frequency energy as compared to other sounds in the mix which may be intended to sound closer. (It’s all relative.) Lipshitz also wrote about an EQ dip like the one I mentioned above for 2-channel stereo to “restore” depth. I leave that to the mixing engineer rather than altering everything coming into my system during playback. Lipshitz himself debated internally just “where” along the signal path a correction ought to be made. But if you prefer lots of depth on everything you hear, go for it. Home sound reproduction is part of the entertainment industry and need not necessarily be approached as a science project.

 

[Duke]

Center image depth does vary with the recording...

This is what (to my ears and/or imagination) seems to be lacking with systems I've heard that have a single center channel speaker: The depth in the middle of the soundstage seems to be limited to the distance to the center channel speaker.

I think the implication of the center image depth varying from one recording to the next with your three-speaker center channel configuration is that the spatial cues on the recording are dominating your perception. My guess is that your distributed trio of sources and your 2 milliseconds of delay for the top-center speaker are disrupting the undesirable "this is where the sound is coming from" cues of a conventional single center channel speaker, but not disrupting the soundstage depth cues on the recording, such that the latter set of cues becomes more plausible to the ears.

Also, thanks for the education about the tools mixing engineers use to manipulate image depth on a recording!

 

[DavidMcRoy]

This is what (to my ears and/or imagination) seems to be lacking with systems I've heard that have a single center channel speaker: The depth in the middle of the soundstage seems to be limited to the distance to the center channel speaker.

I think the implication of the center image depth varying from one recording to the next with your three-speaker center channel configuration is that the spatial cues on the recording are dominating your perception. My guess is that your distributed trio of sources and your 2 milliseconds of delay for the top-center speaker are disrupting the undesirable "this is where the sound is coming from" cues of a conventional single center channel speaker, but not disrupting the soundstage depth cues on the recording, such that the latter set of cues becomes more plausible to the ears.

Also, thanks for the education about the tools mixing engineers use to manipulate image depth on a recording!

That’s a very interesting theory, yet there is something else that may be at play here that I failed to draw attention to: my 3 center channel speakers are identical and therefore have identical timbres, so there isn’t anything anomalous about the top center speaker to make it sonically stand out against its lower left and right center channel mates, once its unique room effects colorations have been sufficiently EQed away. Indeed, the entire system consists of JBL LSR305s and the very similar JBL 305P MkIIs, all EQed to a consistent target curve.

Timbre matching is often cited as critical, yet seldom adequately achieved in multichannel systems, what with dedicated center channel speakers being most commonly of the horizontally arranged MTM configuration with that setup’s own peculiar radiation geometry, coupled with so many insisting on larger, often more elaborate front left and rights channel speakers.

Since I was going to use 80Hz bass management on all channels anyway, I established a budget and set out to do the best I thought I could accomplish within that framework. I came up with using 4 matched subs, all DSP matched for EQ, level and delay, and 13 (nearly) identical speakers for the other 11 channels, these also being closely matched via DSP, and dedicating three of these to the center channel based on my experiments with actual Magnepans in a prior system in the same room. Importantly, every speaker’s radiation geometry is identical to all the others. And let me tell you, after decades of using Magnepans and Quad ESLs at home, getting a uniformly neutral sound and a relatively non-phasy soundfield from a constant directivity design was a comparative cinch to achieve.