Center speaker advice: Is timbre matching real?

[MattHooper]

Doesn't Kal have identical speakers all around? That's always nice if they're capable.

I seem to remember he’s using LS60s for surround speakers. I could be wrong.
But I think a full blade surround system is extreme decadence, even for Kal

 

[Flaesh]

Document from 2001)

 

[Chrispy]

I seem to remember he’s using LS60s for surround speakers. I could be wrong.
But I think a full blade surround system is extreme decadence, even for Kal

Identical, or close to it, all around works well. Don't remember what he uses, thought it was something else but not something I'm prone to memorizing.

 

[Chrispy]

This is his system? https://www.stereophile.com/content...402-silverback-wireless-streaming-loudspeaker

 

[kyuu]

This is his system? https://www.stereophile.com/content...402-silverback-wireless-streaming-loudspeaker

What? No, but he offers a clue in the article:

Throughout the Silverback DS review process, I switched back and forth from the review speakers to the KEF Blade 2 Metas, using the latter as a reference for the setup and tweaking of the Silverbacks. Considering the differences in price and scale—the KEFs are much larger and more expensive—a comparison would be unfair. But lurking at the back of the room was a more appropriate option, my pair of KEF LS60 Wireless speakers.

Kal, AFAIK, uses Kef Blade 2 Metas up front and KEF LS60s for surrounds.

 

[JSmith]

and, if it is affordable, all the same.” - Floyd Toole

JSmith

 

[Chrispy]

Actually was thinking he had B&Ws....but interesting he has this combo

 

[Heinrich]

OK, if you simply meant that, it’s not wise to blindly assume that a speaker manufacturer has done a good job with their centre channel, in terms of design designing to blend with the left and right speakers, I would certainly agree.

I thought you were making a stronger statement than that.

Such a statement would be silly ;-)

I‘m coming from the idea, that there is one right but many wrongs. At least as long as we do not have a definition on what deviation from identical speakers is tolerable in terms of right—center—left panning of virtual sound sources in the surround panorama.

Maybe it‘s defined in one or the other standard of the many certifiable surround formats, I don‘t know. Not the least, in the cinema the use of center channel has to somehow match the very fact, that the majority of listeners don‘t pile-up in the ideal listening position straight in the middle. A strange business, indeed.

 

[Dal1as]

A speakers sound can have a different timbre for many reasons. In fact the same speaker will sound different due to location in a room.

The ideal would be using the same speaker in every location, in a large room with room boundary treatment, with needed eq.

Unfortunately most don't have the funds, size room, and gumption to do this. I actually got to listen to such a room but it was a large dolby atmos setup with over 60 speakers or something.

Just like this hobby the best we can do is lessen the amount of compromises and variables in the equation and go from there.

 

[KSTR]

Right, first rule in the book is the same speakers all around. Mostly not possible (flat panel = horizontal center) or what most would not consider as wise allocation of resources.

But isn‘t the same type for center and surrounds the only logical path? If one allows differences, how to mix the recording so that it would be as nice with another mismatch of arbitrary type? If tolerable differences are not yet defined, and so it seems, the only tolerable decision would be to avoid them altogether.

Anybody to define tolerable deviances between surround speakers—some reasoning attached, that is verififyable with all recordings of the past and future?

As explained earlier in this thread, it is paramount that the phase response for all speakers is about the same (within a like +-45 degrees tolerance band), otherwise simple panning between L-C-R speaker will not work and phantom sources will fall apart. For example, if L/R are 3-way and C is 2-way, all with, say, 4th-order XOs, then the midrange will be pretty much out of-phase.

Next step is then comparable directivity and the icing on the cake is using identical drivers and number of ways for maximized similarity between speakers.

All three aspects contribute to what might be defined as timbre, but phase matching is the dominant and mandatory requirement.

 

[Oddball]

As explained earlier in this thread, it is paramount that the phase response for all speakers is about the same (within a like +-45 degrees tolerance band), otherwise simple panning between L-C-R speaker will not work and phantom sources will fall apart. For example, if L/R are 3-way and C is 2-way, all with, say, 4th-order XOs, then the midrange will be pretty much out of-phase.

Next step is then comparable directivity and the icing on the cake is using identical drivers and number of ways for maximized similarity between speakers.

All three aspects contribute to what might be defined as timbre, but phase matching is the dominant and mandatory requirement.

Phase is what EQ should do. If not there is REW. Phase is not that difficult to adjust for a speaker.

Could be more challenging for the multi-way speaker that is off. I that case you might want to bi-amp and have a DSP that can handle that. But then, it will depend on a speaker.

 

[KSTR]

Phase is what EQ should do. If not there is REW. Phase is not that difficult to adjust for a speaker.

You cannot adjust phase with EQ unless there are specific options for FIR-based "phase EQ" somewhere in the system that can counteract excess phase.
Ideally, one would use proper DRC for all speakers anyway which can cater for systematically different phase responses of speakers. Multichannel is not any different than Stereo in this regard.

 

[Putter]

After having identical LCR speakers I don’t think I can go back to mixing. I then got identical rears to have the bed layer the same.

Now if I do upgrade, they will all be coaxial at least if not all identical again.

AFAICT coaxial speakers are the best way to have identical speakers with the same dispersions notwithstanding the different placements. The fly in the ointment is that the center is expected to handle a heavier load by carrying all the vocals and much of the action in movies and to a lesser degree with music esp. most popular music which still makes it a compromise albeit for different reasons. Single coaxial speakers tend to have doppler distortion especially at higher volume. Adding an an extra woofer as KEF does changes the dispersion which admittedly may not be noticeable.

Personally I've never noticed or been bothered by a change in timbre of a moving object across the screen possibly because our ear are attuned to the doppler effect, i.e. that the sound changes as it moves nearer and then farther away.

 

[MattHooper]

JSmith

I would add, and sometimes even more important,

“And if aesthetically/ergonomically acceptable.”

(which would be pretty rare, I think)

 

[drodgers]

I managed to find a pretty great local deal on the C426Be (A$5500 new). Unfortunately in black (my 228Be's are Walnut), but I'll live

If I end up with a Genelec for another system, I still try that blinded A/B panning test.

 

[scotchrocks]

I managed to find a pretty great local deal on the C426Be (A$5500 new). Unfortunately in black (my 228Be's are Walnut), but I'll live

If I end up with a Genelec for another system, I still try that blinded A/B panning test.

That's a great price; glad you were able to find one. Looking forward to hearing your impressions of the setup.

 

[antcollinet]

Well I've watched 2 films in the last few days - specifically listening for the effects of any timbre mismatch in my totally mismatched 9.1 set up. Couldn't hear anything that I could specifically identify as timbre mismatch - the sounds sent to the various speakers just changes too much in any case.

Or perhaps I just have tin ears.

 

[dlaloum]

They don’t have to match. A center speaker is going to emphasize vocals and dialogue. Having a smaller speaker that has less bass may not be a bad thing, and when full size L/R speakers are used more of the stereo image is retained. I am using a stereo source. The center speaker plays mono. I have a center subwoofer though, and calibrate it together with my center speakers. I use two center speakers. A horizontal one on top of the desk and a smaller speaker on top of the subwoofer. Because the speaker underneath the desk is occluded, I boost the treble trim and decrease the bass all the way. I also boost the treble and decrease the bass of the horizontal center. I use 100 Hz 12 dB slope inline HP filters on both. When I calibrate I aim for a balanced response and then my center channel plays clear vocals/dialogue and complements the L/R well.

IF they have few or no resonances/identifying distortions then yes.... the undistorted signal, can be EQ'd to match voicing on most of todays setup, making the matching more or less moot.

But the key here is "RESONANCES" - most speakers have them... and even if the voicing is matched using DSP, as a sound, voice, or effect pans across from main to center to main, if the resonances of the speakers differ, then it will be noticeable, and will interfere with immersion.

Even with many speakers that are purportedly "matched" - the fact that the "box" is horizontal rather than vertical, often involves differing construction and (as a result) differing resonances.... which is why there are recommendations to match the speakers fully. The closer the match the better the effect.

How much better? how much matching is "enough"? all a matter of taste... one person may not notice what another considers egregious mismatching!
The better ALL the speakers are, the fewer resonances there will be (resonances being one form of distortion amongst many)... so the better the speakers are the less mismatching will be an issue....

 

[dlaloum]

If you choose only to use neutral loudspeakers, then by definition they are timbre matched. If not, then timbre mismatch will just be an additional compromise.

Neutrality has to do with F/R voicing, and not resonances...

Neutrality can be adjusted for by using most DSP's that tailor to a target curve.

Resonances cannot be adjusted for.

Neutrality is no longer a contributing issue for mid market and up processors (some of the budget ones may not have RoomEQ)

 

[dlaloum]

Right, first rule in the book is the same speakers all around. Mostly not possible (flat panel = horizontal center) or what most would not consider as wise allocation of resources.

Center speaker was for ages the ugly bastard in most lines of products for many manufacturers. But as noted, they are exceptions. Have not looked into reasonably priced lines, but for the more expensive lines Revel BE, KEF R and Arendal 1728 are where they are trying to get a good match and size for horizontal centers. There are obviously other examples of good matching, but these seem to be most common.

They are all exceptional centers, with Revel BE, when on sale, being reasonable value. But then if you buy by kilo or pound, Arendal 1728 big center is still a bargain for its 80kg.

Also the Gallo Reference AV center (or Strada2 center) are good timbral matches for their floor standing cousins...

The ones that do good matching, tend to be the ones where resonances have been very well controlled or eliminated.