Jump to content

Sound Staging Of Klipsch Speakers


Recommended Posts

In the positive reviews of Klipsch speakers that I have read,most say that the left to right sound staging of them is quite good,expanding past the speakers themselves.However,I have also read in Klipsch reviews that being horn speakers,they lack the depth/front to back imaging of other models.

What say you all?

Link to comment
Share on other sites

I suppose you're talking about certain Klipsch models, but you didn't specify which ones.  That gives me the license to respond how I see it.

 

I've found that dialed-in Klipsch Jubilees exhibit the depth of the recording itself.  The Chesky "Ultimate Demonstration Disc" has a track to demonstrate depth of stereo image capability (track #5--"If I Could Sing Your Blues" by Sara K).  No problems there.  The Jubilees however don't add artificial stereo depth however--as I would expect that "reviews" as you put it would expect, i.e., depth that's not actually in the recording. 

 

The issue of depth of stereo imaging actually has a couple of measurements that can tie the imaging performance of the loudspeakers to that subjective capability:

  1. Flat phase and SPL response.  This implies time alignment of drivers.  Only the Jubilees and other DSP crossover-aligned loudspeakers have this capability.
  2. Wide polar coverage above 200 Hz.

There is one other effect that can give the listener an impression of greater depth of stereo soundstage image: dipole or bipole radiation into the room.  This means that a lot of sound is reflected off the front and side walls of the room--like Bose advocates.  The problem with this is that bouncing a lot of sound off the walls, ceiling and floor actually destroys the fidelity of the stereo imaging but gives the listener an impression of greater depth or presence.  This isn't a desirable trade off, in my experience.  True dipole loudspeakers (like Magnepan, MartinLogan, etc.) have the issue of "head-in-a-vise" imaging and very low efficiency such that the loudspeakers can't reproduce the full dynamics of recordings--the type of recordings that haven't been compressed in their dynamics.  I don't think that's a very good tradeoff, either.

 

To get both soundstage focus (i.e., being able to point at individual sources of sound within the horizontal stereo image), wall-to-wall imaging and greater front-to-back depth requires loudspeakers having great phase and SPL fidelity, as well as full-range directivity control.  Jubilees do that...in spades.  But I can think of no other Klipsch models that do all that well (except perhaps an MCM Grand or some other Klipsch Professional [cinema] loudspeakers having full-range directivity). 

 

I'll stop there.

 

Chris

  • Like 3
Link to comment
Share on other sites

Since it seems to be just you and me thus far, perhaps I can add the following explanation.  I just found this at diyAudio--a quote by Tom Danley of Danley Sound Labs (the inventor of the Synergy™ and Unity™ horns...multiple entry horns--MEHs).  It's a bit long--but nevertheless quite interesting in discussing where this apparent depth of soundstage comes from, first starting with the ESL-63s:

 

Quote

Hi Guys


...If/when our ears are each presented with an identical signal from a forward position, that signal is affected by the shape of our ears and head BUT both L and R are affected equally and so we hear the height location due to those acoustic alterations (our ears and head impose depending on the sounds direction) but it sounds like it’s in front of us.

If you take away the angle-dependent effects of your outer ear by using headphones (that present sound perpendicular to the eye/ear view), you have removed what tells you height etc.  It is the effect your ears and head have (that we are all 100% unaware of).  We can’t hear any of that consciously BUT those changes ARE part of what our brain uses to tell the spatial stuff.  It is because of all those direction related head / ear changes, that one can put tiny microphones inside a persons ears and record (in the ear recordings) which when played back properly (minimizing the ear effects), one can make a hair-raising real-sounding recording (for one listener in one location).

I heard some recordings Don Davis made (old days [of] SynAudCon).  One, a recording made of a fellow walking around the pits at the Indy 500 really gave me goose bumps, and I guess it's partly why I am so interested in imaging and capture of a stereo image.

The first time I ever heard a speaker with little identity was in the 90s at Intersonics when I was “fixing” (removing the protection) from my bosses ESL-63s. I built the levitation transducers and electronics and made ESS speakers for fun back then and my boss (to my horror) asked me to “fix” his (very expensive) speakers.  I explained the annoying spark gaps were there for a reason and while they were a bother, they did tell you something.  Anyway, out they came and a few months later of louder sound, they needed to be re-built.  A hi-fi writer, Dan Sweeney said to me once back then and it stuck like a tack in my forehead “there are speakers that sound good, there are speakers that go loud, but none do both”.

Anyway, when I fired the first of Roy's ESLs back up in my listening room, that is when I was struck by the way some sound / recordings seemed to come from behind the speaker. It made me wonder why, and I thought it had to be the way the speaker radiated, not how it worked.  In that case, the difference between the ones I was making and this was that the speaker was driven in concentric rings, starting at the center. The light went on.  We made acoustic levitation systems at work and at 20KHz, the wavelength [of sound] is about 5/8 inch.  A surface as large as the ESS speaker would be many wavelengths across but by driving the rings sequentially, a segment of a spherical wave front was produced. You can see the ring constriction here:

http://www.audiodesignguide.com/esl/esl63ph.jpg

 

esl63ph.jpg

The signal for each ring is progressively delayed by using an all ladder style [all-]pass filter, each filter provides a small time delay. A model of that is here;

http://mark.rehorst.com/Quad_ESL-63/...0Schematic.png

The result was that even though the source was large compared to the wavelength, it radiates as a partial sphere “as if” the source really was behind the film and the sound wave simply traveled through it’s flat plane...


Fast forward about 15 years.

When I was really struck by this effect again was working on the early Unity Horns as the TEF measurements [phase, group delay, impulse, and energy time curves--ETC] showed they were getting closer and closer to one source in time and space.  I kept hearing an increasing disconnect with the physical depth location. There wasn’t anything in particular I could see in the measurements but the effect was quite audible and puzzling.

Later, Synergy horns eliminated the crossover phase shift and the effect [even] more pronounced. A friend, Doug Jones has been an acoustics researcher and professor and was department chair at Columbia College in Chicago and had worked with Gary Kendall in the old days on stereo imaging. Doug made the old LEDR stereo test recordings if any of you are that old.  He was the curator of the Heyser archive/library at Columbia and he allowed me to look around at Dick’s stuff and read some of his unpublished work...

...The idea I proposed was that to localize the depth location, one has to be getting sounds to the R and L ears that have “enough” difference for your brain to detect/triangulate the source location.  So Doug’s condensation was that a true simple spherical, plane and cylindrical wave would all have this effect because in each case, what reaches the two ears is identical.

The inverse of this is that the more irregular the radiation pattern is, the more clues the radiation pattern carries about the source itself.  We can’t see this domain and while we can “hear” aspects of it, a visual example is better.  In one case if you consider “ loudness” on a 3d polar plot, one has an omni-directional simple point source which looks more like a cue ball and then, one can have a complex radiation pattern looking more like this:  http://www.worldoceans.com/pix/c_spa07.jpg

 

c_spa07.jpg

Understand, right now, this isn’t about the room or other possible speaker flaws, only how uniformly it radiates and how similar the R and L ears pressures are from the source alone.

In the past here, I have tried to describe this effect here and if one wanted to “hear” how the radiation pattern is audible, to put a small good full-range driver in the middle of a large baffle. Yes, the source has other serious limitations as a faithful source, but because it radiates as a simple source over much of the band (radiate a hemisphere up to upper mids), they can have phenomenal imaging...

What causes the more “puffer fish” radiation [pattern]? Any time you have more than one source of sound, that is more than about 1/3--1/2 wavelength spacing from the other source. This might be a tweeter with a step in the baffle which re-radiates or edge diffraction, a driver in breakup or used too high for its size, etc., etc.

We have to remember, too, that while an acoustically small driver on a flat baffle looks like just that, it radiates as if it were a 180 degree [coverage] horn, up to the point the source has directivity and subject to problems [that] impedance discontinuities cause and reflections, re-radiations, and so on. The things in horns Earl [Geddes] has identified as higher order modes can have their origin in the same kinds of things as edge diffraction, discontinuities, etc. on a flat baffle, except the angle isn't closed enough for the re-radiation to bounce off a second surface. Any or all of these and more can produce multiple paths which add differently in a space as small as the distance between your R and L ears, and THOSE differences let you hear/localize the physical depth of the source.

Without those origin clues, all one has are the remaining “clues” encoded in the recording which then (ideally) dominates your perception. Without those clues, your brain hears a compelling mono & stereo phantom image and not the R and L speakers where the sound originates physically.  Again, none of this has to do with the room or other speaker flaws.  It is only an observation about how a loudspeaker radiates so far as its spatial identity.  Also, playing a recording, all [of] that also depends on the recording microphone, its placement, use, and every part of the chain up to the speaker.

What ever this part of the radiation effect is, it is audible in a generation loss recording [i.e., re-recording an original recording through loudspeakers several times] or one-sided capture. A dealer of ours in Europe made a camcorder video that captures what you hear firsthand with this effect.  Understand, these are not hi-fi speakers and our SM-80 is more a loud/lower cost cabinet than a refined one.  Yet with headphones you can still hear some of what you hear in person with the source identity thing:  Danley Sound Labs SM80 vs. d&b Q7 - YouTube

So, what would one need to do this [i.e., get a unified sound image]?  Exactly what many have observed themselves, mostly speakers that radiate as a simple source over a broad band.  These would be some kinds of Horns that cover a broad enough band as one source, [that are] acoustically small (point) sources on simple [smooth flat] surfaces like a large baffle, horns which extend out to the pattern loss frequency [i.e., lower frequency at which the horn loses directivity control] for that [horn coverage] angle and dimension, and so on.


A floor-to-ceiling line source using a ribbon or continuous full range driver could also do it, a planar ESS speaker can do it if you’re up close and it really is a planar source.  The problems here are that the line and planar sources MUST be large compared to the lowest F, longest wavelength of concern and for an array to work like a continuous source. They can’t be more than about ¼ wavelength apart at the highest frequency of concern.  A point source is automatically what results from a source too small to have directivity.  The pressure freely radiating away is the 3d equivalent of a pebble tossed into a pond governed by the pressure gradient and speed of sound. This is why a small full range driver has this effect, a 4-inch driver only starts to get directivity as a simple piston source when the radiator is more than about 1 wavelength in circumference (K=1).

In the room?


Pretend all your walls, floor and ceiling are mirrors and your loudspeakers are now bare light bulbs.  From your listening position, how many light bulb images do you see?  Hard to count them all.  Each image is an equal-angle reflection (like sound can produce) and ALL the reflections compete with the direct image even if they "look nice".

Corner-dipole-2s.jpg.64c01de6b2ab7110a800571b86bb2eec.jpg

Mirror images formed by a dipole loudspeaker near a room corner

 

Now pretend one has a flashlight instead of bare bulb.  Because of [that added] directivity, there are very many fewer images/[reflections] of the light bulb [from the listener's position], [perhaps] only two [images/reflections] if it’s narrow enough.  Now with that in mind, a typical loudspeaker radiates like a bare bulb up [in frequency] to where it begins to have directivity.  Many [loudspeakers] reach [a] narrow flashlight pattern very high up [in frequency].  In the room, the higher the frequency, the more like light a sound reflection behaves.  If you take an impulse or ETC [Energy/Time Curve] measurement at the listening position, you can usually see the first few reflections as they arrive with progressively longer delays because of the longer and longer path lengths.


Directivity.

Directivity is inconvenient.  It has the “what frequency?“ question attached [to it].  It is wavelength- and geometry-dependent and that makes having it at a lower frequency a non-starter for the home.  So [that's what makes] it "not important” to most [loudspeaker] manufacturers.

 

That assertion aside, many [or most people] can apparently still hear a difference when they use [loudspeaker driver array] self cancellations to produce a figure-8 dipole pattern and suppress/eliminate most or much of the side radiation [coming out of the loudspeakers] (which would normally reflect off the walls and harm the stereo image in the recording) in exchange for stronger but later and possibly scattered rear[-radiated] energy.

 

DPOB3a.gif.3fa5a7d662358b4c83c867aae068090a.gif

Dipole radiation patterns as a function of the ratio of effective diameter of the radiating source (d) to the wavelength of sound (w).

 

reflections-3s.jpg.6566f3ae74b53c9ed106a817c5b51532.jpg

Comparing complex in-room reflections from acoustic drivers (infinite baffle - L, dipole-R) .

 

It is really cool people are talking and thinking about this stuff.  Our hobby is in a sad state, it needs some breakthroughs and for the most part, the “business” looks to me like it is selling what you can make and like most marketing, a dollar [spent on] marketing the image of technology brings in more sales than a dollar spent on developing technology.


Best,
Tom Danley

 

  • Like 2
Link to comment
Share on other sites

On 7/15/2020 at 4:25 PM, milwaukeebill said:

Thanks Chris,feel free to continue.To be more specific,I was more interested in the overall sound staging capabilities of  the Reference Premier line in both book shelf and floor standing speakers.

Looking at Danley's post above, it can be seen what improves the perception of depth of the loudspeakers: loudspeaker drivers are closer together at their crossover frequencies.  In the case of the home theater line of loudspeakers that Klipsch markets, these multi-way (usually two-way) loudspeakers cross somewhere between 1 and 2 kHz.  At 1.5 kHz crossing frequency, a 1/4 wavelength (center-to-center) distance is ~2.6 inches (6.5 cm).  So if the woofers are mounted more than that distance from the higher frequency driver centerline (i.e., the horn-loaded compression driver in the top of the loudspeaker cabinet), the degrading effects on subjective perception of soundstage depth that Danley discusses above become audible. 

 

Tower loudspeakers have a much greater separation distance of the woofers from the high frequency compression driver horn centerline than the bookshelf loudspeakers, and so their apparent soundstage depth will suffer to the degree that the woofers are separated from the HF driver. 

 

Also note that the crossover filter order introduces at least a 90 degree phase lag on the lower frequency drivers.  I'm not sure what order crossover filters are used in the Reference Premieres, but I'd guess that they're at least second order filters, implying that they are adding another 180 degrees of phase lag on the woofers, which equates to another 5 inches of time misalignment and effective separation of the drivers.

 

Chris

  • Like 2
Link to comment
Share on other sites

On 7/14/2020 at 1:13 PM, milwaukeebill said:

In the positive reviews of Klipsch speakers that I have read, most say that the left to right sound staging of them is quite good, expanding past the speakers themselves. However, I have also read in Klipsch reviews that being horn speakers, they lack the depth/front to back imaging of other models.

What say you all?

 

I like to loan some of my speakers to some of my friends so that I can compare the sound of different listening rooms. The room acoustics have a big influence on the overall sound quality.

Link to comment
Share on other sites

3 hours ago, Chris A said:

I just found this at diyAudio--a quote by Tom Danley of Danley Sound Labs (the inventor of the Synergy™ and Unity™ horns...multiple entry horns--MEHs).  It's a bit long--but nevertheless quite interesting in discussing where this apparent depth of soundstage comes from, first starting with the ESL-63s:

 

That is a good article.

Link to comment
Share on other sites

When using a digital electronic crossover, I can alter the perceived width or depth of the soundstage from a pair of loudspeakers by changing the time delay in micro-seconds. So it seems that any given pair of speakers, combined with any stereo/room combination might have its own un-intended audible spatial effects because of multiple reasons.

 

My comments are not meant to minimize the importance of the logic of what Chris A and Tom Danley say about the point source designs.  

Link to comment
Share on other sites

4 hours ago, Khornukopia said:

When using a digital electronic crossover, I can alter the perceived width or depth of the soundstage from a pair of loudspeakers by changing the time delay in micro-seconds. So it seems that any given pair of speakers, combined with any stereo/room combination might have its own un-intended audible spatial effects because of multiple reasons.

 

My comments are not meant to minimize the importance of the logic of what Chris A and Tom Danley say about the point source designs.  

 Exactly.

 

They go hand in hand.

 

We need to realize that even in "large" dedicated residential rooms, the rooms are "acoustically" small. And in turn causes a lot of reflection, short-term delay problems. This is very unlike outside or in large auditoriums were the delay between the source (instrument or voice) and any reflections is much longer (time).

Link to comment
Share on other sites

Hey Chris, you had me going there for a while. The ESL-63 picture caught my eye, I saw the "quote", and assumed it was you referring to one of your usual long posts (not saying that's a bad thing).

 

And then I'm reading, thinking, wow, I didn't know Chris worked at Intersonics and worked on acoustic levitation devices. He must have worked with Tom Danley. Chris knows Doug Jones? I've spoken with Jones. Also, my niece was studying opera at DePaul and switched to Columbia to study audio recording instead and had Jones as a teacher for a couple courses.

 

Read on a little further and I'm thinking wow, this sounds more and more like Tom Danley and the Synergy horns like my SH50.

 

THEN, I go back and read your first couple of lines.

 

ROTFLMAO.

 

Duh, Duh and DUH.

It so much fun to feel stupid sometimes. 🤩

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...