Klipsch Roy Delgado explaining why Klipsch measures so bad.

[Chris A]

Why not just switch brands?

Believe it or not, I'm not really a "brands" kind of guy. Brand means little to me except in the general type of loudspeakers offered by the company, and their acoustic performance.

In the case of Klipsch, there basically are no competitors in fully horn loaded loudspeakers that are relevant/readily available. (This specifically excludes boutique loudspeaker companies that publish astronomically high MSRPs for their products.)

In any case, I actually modify the designs to fit the needs...as you can see in my setup.

 

[PristineSound]

Cornwalls are not Jubilees (not even close...even if the Cornwalls are tri-amped and dialed in). They are very far apart in almost all acoustic performance areas.

Cornwalls come in varieties. I suspect your assessment would change if you listened to CW IVs instead of CW Is-IIIs.

That model/version is no good.

I instead suspect your recordings listened to during auditioning and/or the room acoustics and setup....or simply your listening preferences/hearing acuity.

The source material is bad or you don't have golden ear.

but "spinorama" I think isn't one of the best moments of his illustrious career.

The measurements are flawed.

Once you get a stock Khorn from years passed dialed in and tri-amped with the old drivers/horns, the difference wasn't quite as great as one might like to hear. However, having come from 1st-gen Jubilee performance, it's easy to hear what you get by tri-amping an old pair of Khorns--and what you don't.

It's very particular with how you EQ and amplify them. You need three amps btw, not one, not two but three. Oh, that EQ and triamp only works with an older model, that's not in production anymore.

My personal experience over the past 17 years is that any loudspeaker of the performance level of the Jubilee (both versions) is highly susceptible to setup and room acoustics issues.

You don't have it set up right and your room sucks.


Maybe the Klipsch just doesn't sound good? Anyways, going to Capital Audio Fest later this year, if any of the higher end Klipsch are there, I'll check them out.

 

[Keith_W]

Clearly, I missed this earlier. I think the word hyperbole likely applies to your use of the word "unlistenable" in this case. I've heard a lot of loudspeakers in a lot of environments, but virtually none of them could I apply that word to their performance. I instead suspect your recordings listened to during auditioning and/or the room acoustics and setup....or simply your listening preferences/hearing acuity.

But unfortunately, this is not the first time I've run into this sort of thing. I question the circumstances under which you listened to them. If you describe the room size and placement in-room in which you heard them, it might help to dissuade my assessment of the likelihood of this condition. Also, you would need to describe the exact music selections you used for your audition, since these loudspeakers do ruthlessly expose any defects in mastering, etc.

It turns out that I was in error, I thought I was listening to Jubilees but I wasn't. I don't know what Klipsch model is depicted here but it was certainly pretty horrible. It was in a hi-fi show, in a fairly large room (I am guessing 4m x 5m), and it was virtually empty because nobody could stand the sound. It was playing typical hi-fi show shlock: female vocals, jazz, some rock.

What I have empirically found, however, is that all-pass phase growth from the crossover filters = harshness. If you eliminate the all-pass phase growth due to the 4th order crossover filters used in the Klipsch DSP box, you eliminate the induced harshness. Any other harshness you might encounter would mostly come from the recordings themselves.

That is an interesting observation indeed. I can't say I have ever made that observation, mostly because the EP spikes like the one you demonstrated come from my room, not my XO's or my speakers. Maybe i'll go put a nasty AP into my speakers and deliberately create that EP spike and see if I can hear any harshness.

 

[Chris A]

That model/version is no good.

Which version? The one designed in 1965, or the one designed in 2020?

The source material is bad or you don't have golden ear.

There are a lot of bad recordings out there, however. But, there are things that you can do to repair some of the damage done. This includes real-time corrections to mastering EQ issues.

I've never said anything about "golden ears". Perhaps this is important to you? It's not to me. However, like PWK said, the more you listen to the real thing (live acoustic music), the better your hearing typically gets.

The measurements are flawed.

Not flawed, but rather typically incomplete, and often missing critical information. The generated early reflections curve from spinorama is a guess only--not measured.

It's very particular with how you EQ and amplify them.

This is basically true with the old Heritage models (Khorn, La Scala, Belle, Cornwall, Heresy) due to issues that were discovered well after they were first designed in the 1940s-1970s. The newer models (especially those time-aligned via DSP)--not so much.

You need three amps btw, not one, not two but three.

Only if you select a three-way version that you choose to triamp/DSP. If you choose a Jubilee, then you can think of it as using a good HF amplifier and a good subwoofer/bass bin amplifier--low power. These are actually fairly inexpensive, as hi-fi amplifiers go.

Oh, that EQ and triamp only works with an older model, that's not in production anymore.

I don't know where you got this, but it's wrong. Anyone can multi-amp/DSP correct their loudspeakers.

You don't have it set up right..

You do have to set it up right.

...and your room sucks.

Just plopping your loudspeakers into a room without any thought of what the result will be--will get you pretty much what you put into it.



Chris

 

[Chris A]

I don't know what Klipsch model is depicted here but it was certainly pretty horrible.

That doesn't surprise me at all. This is not a serious contender (to say the least). I would agree wholeheartedly (not even listened to them--just based on acoustics fundamentals).

That is an interesting observation indeed. I can't say I have ever made that observation, mostly because the EP spikes like the one you demonstrated come from my room, not my XO's or my speakers. Maybe i'll go put a nasty AP into my speakers and deliberately create that EP spike and see if I can hear any harshness.

Here is an interesting (and non-calibrated) resource: https://www.linkwitzlab.com/Attributes_Of_Linear_Phase_Loudspeakers.pdf

Here are my experiences with finally getting all the all-pass phase growth removed from the crossovers. The differences I heard once I discovered the "zeroth order" approach to crossover filtering start right here.

It's not subtle if you've got full-range directivity control (and this excludes most hybrid horn/direct radiating woofer designs, or anything not fully horn loaded or a full-range dipole radiator), and you've taken a few pains to reduce early reflections from around your loudspeakers and at your listening position (including removing the coffee table in front of the MLP). It takes all three at the same time:

1) full-range directivity control
2) suppression of early reflections (D'Appolito calls this the "reflection-free zone"), and
3) flat phase response

The first two factors increase the direct/reflected acoustic energy ratio in-room. Another way to say it--you're well inside the critical distance for your loudspeakers and room acoustics.

Flat group delay response goes with the flat phase response, since group delay is nothing but the first derivative of phase WRT frequency.

Chris

 

[PristineSound]

Which version? The one designed in 1965, or the one designed in 2020?


There are a lot of bad recordings out there, however. But, there are things that you can do to repair some of the damage done. This includes real-time corrections to mastering EQ issues.

I've never said anything about "golden ears". Perhaps this is important to you? It's not to me. However, like PWK said, the more you listen to the real thing (live acoustic music), the better your hearing typically gets.


Not flawed, but rather typically incomplete, and often missing critical information. The generated early reflections curve from spinorama is a guess only--not measured.


This is basically true with the old Heritage models (Khorn, La Scala, Belle, Cornwall, Heresy) due to issues that were discovered well after they were first designed in the 1940s-1970s. The newer models (especially those time-aligned via DSP)--not so much.


Only if you select a three-way version that you choose to triamp/DSP. If you choose a Jubilee, then you can think of it as using a good HF amplifier and a good subwoofer/bass bin amplifier--low power. These are actually fairly inexpensive, as hi-fi amplifiers go.


I don't know where you got this, but it's wrong. Anyone can multi-amp/DSP correct their loudspeakers.


You do have to set it up right.


Just plopping your loudspeakers into a room without any thought of what the result will be--will get you pretty much what you put into it.



Chris

"or simply your listening preferences/hearing acuity." <- golden ear-ish, maybe?

Anyway, my point is there is a reason or excuse for almost everything as to why Klipsch doesn't sound good. . .to some (or a lot of) people at least.

Maybe Klipsch should start designing speakers not just for corners as Roy have stated as a reason why they don't measure well. Maybe they should leverage all the research that has been conducted in acoustic and psychoacoustic science over the years (in another interview Roy said Paul told him to not get too deep into established research papers because it may take away creativity and/or innovations).


I have my theories: I don't think Roy and Klipsch are incapable of designing an excellent measuring and sounding speakers at all. In fact, I think they are more capable then many companies because they generate so much revenue and can afford the R&D resources. I think what it comes down to is:

(a) The current formula and business model works, enough people are buying, they are selling and they are making money, so don't fix it if it's not broken.

(b) Many of their customers buy based on nostalgia, it's kind of a like a muscle car, sure it's fast off the line, but there is no way in hell will you ever get a good lap time, but who cares? People who wants to buy a muscle car, because they want a muscle car. So instead of designing from the ground up for a very expensive pair of speakers, retrofitting fixes to a 80 year old model that was design with fairly primitive knowledge of modern science 80 years ago, is what this specific group of customers want.

I also think Roy was blindsided by Thomas's question about why Klipsch don't measure well, so he may have stumbled on his replied because the above answers (if my theories A and B is right) would not be good for PR. You see, Thomas is a hardcore subjectivist, that may be a question that he is asking to validate that measurements carries very little weight.

There is a reason why (1) in all the years Klipsch only sent one of their newer model speakers to an objective based review outlet, Audioholics (2) there is essentially no interaction between Klipsch and the objective based speaker community that I can observed, you will never see Roy giving interviews with people like Erin, James Larson or Amir. Or if he did, I missed it.

 

[Arindal]

Maybe Klipsch should start designing speakers not just for corners as Roy have stated as a reason why they don't measure well.

If I am not mistaken, Klipsch has produced only a single model, the Klipschorn, for which corner placement is mandatory. I understand it as an example what leads to flawed measurements if the actual design is not taken into account.

Maybe they should leverage all the research that has been conducted in acoustic and psychoacoustic science over the years

Maybe they did, but did disagree with what some people regard to be the optimum and design speakers according to their philosophy? You may like them or not like them, but at least the more modern conceptions seemingly are designed intentionally the way they are. Take the latest active monitors as an example, nothing in them looks like a random decision.

Thomas is a hardcore subjectivist, that may be a question that he is asking to validate that measurements carries very little weight.

Cannot say anything about the interviewer, but I understand the answers so that they absolutely care for measurements and fact-based parameters, but set different priorities what they prioritize to have optimized.

I might not agree with that and might not like this or that model, but the whole conception sounds to me just as objective and legitimate, if not more, as the idea of delivering a flat anechoic response in the first instance at the expense of other parameters.

 

[GXAlan]

There is a reason why (1) in all the years Klipsch only sent one of their newer model speakers to an objective based review outlet

Klipsch Heresy IV Loudspeaker Lab Report

Lab Report With two horns on show there's a hope of high sensitivity and while Klipsch's 99dB spec. is a trifle optimistic, a mean figure of 95.4dB/1m/2.83V (500Hz-8kHz) is still very generous. The figure at 1kHz is higher still at 96.7dB but this coincides with a peak in output [Graph 1, below]...

www.hifinews.com

Klipsch Klipschorn AK6 Loudspeaker Lab Report

Lab Report Such is the difference in acoustic centres/path lengths of the drivers – between the treble and mid/bass horns in particular – that the wavefront from the former arrives significantly ahead of the latter, complicating the time-windowed response measurement. The step function at the...

www.hifinews.com

Klipsch La Scala AL5 Loudspeaker Lab Report

Lab Report Measuring a horn speaker such as the La Scala AL5 using a quasi-anechoic, time-windowed technique is complicated by the time delays associated with the path lengths through the midrange and bass horns. If the measured impulse response is windowed to remove the first (floor) reflection...

www.hifinews.com

Klipsch The Nines integrated loudspeaker system Measurements

Sidebar 3: Measurements I performed the measurements with the Klipsch The Nines Primary speaker (serial number 107128623160250). Before I started the testing, I installed the Klipsch Connect app on my iPad mini and paired it with the speaker with Bluetooth. I used the app to set the equalizer...

www.stereophile.com

KLIPSCH Jubilee

Klipsch przygotował Jubilee z okazji swojego 75-lecia. To największe i najlepsze dzieło w historii firmy, nad którym pracował już sam założyciel. ...

audio.com.pl

 

[PristineSound]

Klipsch Heresy IV Loudspeaker Lab Report

Lab Report With two horns on show there's a hope of high sensitivity and while Klipsch's 99dB spec. is a trifle optimistic, a mean figure of 95.4dB/1m/2.83V (500Hz-8kHz) is still very generous. The figure at 1kHz is higher still at 96.7dB but this coincides with a peak in output [Graph 1, below]...

www.hifinews.com

Klipsch Klipschorn AK6 Loudspeaker Lab Report

Lab Report Such is the difference in acoustic centres/path lengths of the drivers – between the treble and mid/bass horns in particular – that the wavefront from the former arrives significantly ahead of the latter, complicating the time-windowed response measurement. The step function at the...

www.hifinews.com

Klipsch La Scala AL5 Loudspeaker Lab Report

Lab Report Measuring a horn speaker such as the La Scala AL5 using a quasi-anechoic, time-windowed technique is complicated by the time delays associated with the path lengths through the midrange and bass horns. If the measured impulse response is windowed to remove the first (floor) reflection...

www.hifinews.com

Klipsch The Nines integrated loudspeaker system Measurements

Sidebar 3: Measurements I performed the measurements with the Klipsch The Nines Primary speaker (serial number 107128623160250). Before I started the testing, I installed the Klipsch Connect app on my iPad mini and paired it with the speaker with Bluetooth. I used the app to set the equalizer...

www.stereophile.com

KLIPSCH Jubilee

Klipsch przygotował Jubilee z okazji swojego 75-lecia. To największe i najlepsze dzieło w historii firmy, nad którym pracował już sam założyciel. ...

audio.com.pl

This is the same publication house that John Atkinson works for, the guy who never said anything bad about any products, regardless how they measured. That is the reason why I didn't mentioned Stereophile nor Soundstage.

Also, it would have to be Klipsch who actually sent the speakers knowing that it will be measured and reviewed by someone who will tell the good, the bad and the ugly. Do you know if Klipsch, themselves actually sent the Jubilee to audio.com.pl? Cause all the Klipsch that Erin and Amir ever reviewed were not sent by Klipsch.

 

[kemmler3D]

Power is cheap now, efficiency is overrated

I agree with the first half.

An 87dB sensitivity speaker that can handle 400w peaks (healthy values IMO) can only do 113dB @ 1m. If you want to talk about reference levels at normal seating distances like 3-4m then you probably have to find more efficiency. I think there is a reason the people who really care about dynamics and the people who like horns are often the same people.

Personally I'm not in the market for super loud, peaks or otherwise, but if you are I don't think you can overlook efficiency because it's easier to find an amp that can put out 800w than it is to find a speaker to handle that.

 

[PristineSound]

Maybe they did, but did disagree with what some people regard to be the optimum and design speakers according to their philosophy? You may like them or not like them, but at least the more modern conceptions seemingly are designed intentionally the way they are. Take the latest active monitors as an example, nothing in them looks like a random decision.

I don't think anyone of us would know if we didn't work at Klipsch, so we can only go what what we read and see that is publicly available.

Here is an interview that is publically available where Roy said, he was told by Paul to not get too deep with published AES papers so that he doesn't go down the "beaten path," in which Roy obliged.

These papers are peer reviewed and they are scientific studies with data to back. It doesn't make any sense to me to not read them for the knowledge.

At 3 mins mark

 

[PristineSound]

I agree with the first half.

An 87dB sensitivity speaker that can handle 400w peaks (healthy values IMO) can only do 113dB @ 1m. If you want to talk about reference levels at normal seating distances like 3-4m then you probably have to find more efficiency. I think there is a reason the people who really care about dynamics and the people who like horns are often the same people.

Personally I'm not in the market for super loud, peaks or otherwise, but if you are I don't think you can overlook efficiency because it's easier to find an amp that can put out 800w than it is to find a speaker to handle that.

If anyone listens to 113dB, forget about music, you probably won't be able to listen to anything for long.

If a driver is designed to have low compression and low distortion at high excursion, efficiency is not a bad thing to sacrifice in the design of a speaker.

 

[kyuu]

If anyone listens to 113dB, forget about music, you probably won't be able to listen to anything for long.

We're talking about brief peaks, not average listening levels. Although even then, I generally consider 100dB peaks to be the max you should be concerned about for a home setting (at 3 meters, for far-field listening), so 80dB average plus 20dB for peak, not 113dB which implies average levels over 90dB which would still be deafening and not healthy for long-term exposure.

 

[PristineSound]

We're talking about brief peaks, not average listening levels. Although even then, I generally consider 100dB peaks to be the max you should be concerned about for a home setting (at 3 meters, for far-field listening), so 80dB average plus 20dB for peak, not 113dB which implies average levels over 90dB which would still be deafening and not healthy for long-term exposure.

So efficiency is over rated then?

 

[kemmler3D]

If anyone listens to 113dB, forget about music, you probably won't be able to listen to anything for long.

If a driver is designed to have low compression and low distortion at high excursion, efficiency is not a bad thing to sacrifice in the design of a speaker.

Totally, but as @kyuu points out we're just talking peaks here, (not as damaging to hearing) and if you are listening at 4m instead of 1m we're down to 101dB at that level.

That's definitely loud, but it's also not unreasonable at all if someone wants to listen to some dynamic classical music in a larger room. And that's a fairly popular use case among capital-A audiophiles.

I'm someone who is happy to sacrifice efficiency in certain contexts, but more power can only get you as much SPL as the system is inherently capable of, so efficiency is only overrated to the extent you don't care about high SPL.

 

[USER]

Wanted to hate my friend's Klipsch RB-5s because of my prejudices but they were fine. Ignore everything below 100Hz as that is a sub. They are proper Klipsch bookshelves that retailed for around $800 in the late 1990s, early 2000s. Certainly doesn't cut it today despite internet nostalgia (more so because of the lack of low end starting too high up for proper sub integration). Seems like they went for showroom appeal tuning shortly after these. This doesn't strike me as a speaker from people that don't follow engineering.

Had some fun and brought over my Dennis Murphy Affordable Accuracy Monitors to show him there was no correlation between cost and performance. Took some measurements and had a good time going over the data. Again, ignore below 100Hz. NOW THAT IS HOW YOU CONVINCE SOMEONE OF PROPER AUDIO SCIENCE.

He now has Revel M106s.

 

[kyuu]

So efficiency is over rated then?

Nah. Doesn't make up for sins as basic as a chewed up FR though.

Edit: Also, while power may be "cheap" nowadays, I think a relatively small population even around here have 1kW of power on tap, or even half that. Even nowadays, a couple hundred watts is probably more than average, and for a system that's expected to play peaks without substantial clipping you'll need fairly efficient speakers.

If you do have a 1kW amp, then sure, don't worry about efficiency.

 

[GXAlan]

Do you know if Klipsch, themselves actually sent the Jubilee to audio.com.pl?

Yes.

"Klipsch jest jednak aktywnym partnerem EISA, a więc stowarzyszenia ekspertów, a nie influencerów, i bez oporów dostarcza nam do testów wszelkie produkty"

Translation:
However, Klipsch is an active partner of EISA, i.e. an association of experts, not influencers, and they provide us with all their products for testing without hesitation.

 

[Cbdb2]

listening at 4m instead of 1m we're down to 101dB

Not in a room.

 

[PristineSound]

Not in a room.

I guess it depends on the room. And if you are getting significant room gain than your tonal balance is likely off.