Klipsch....buy or avoid?

[Knight]

I sold Klipsch and I don't like them at all. But what I can tell you is that I noticed while selling them there were a small group of people that liked them alot. The horn tweeters drive me crazy but some people love them. You should listen to them and no one else. Just keep in mind that they tend to be power hungry speakers but if you have a good amp you'll be ok. After you listen to them next to an speaker that has say a bipolar design you will figure out which style you like, at least that's how I decided.

 

[UFObuster]

Check out -

http://www.theaudiocritic.com/back_issues/The_Audio_Critic_28_r.pdf

and on the usefulness of listening tests -

http://www.aes.org/technical/documentDownloads.cfm?docID=168

OK, I checked it out...along with visiting ABX testing sites (which I have been aware of)...these do not have bearing on this discussion.
As far as listener preference in ABX testing is concerned...it still remains completely under the control of the tester. It may be valid or invalid depending on how it's set up. I'm not an electrical engineer nor a constructor of enclosers, but I am an advanced degree holder in a medical field with more than enough...might I say expert....knowledge of double-blind testing procedures, study design, statistical evaluation of results.
As I see it, and in the very simplest challenge, these ABX speaker comparisons don't even account for the differences in efficiency of the speakers tested...How in the world can you head-to-head test different speakers used in the same system without introducing bias based on the fact that components USED may favor one or the other speaker?....Using efficiency alone...ABX tested....one will always be LOUDER than the other...and that will likely get the "best" rating by blinded listeners.
...After only a quick survey...it seems ABX testing may be useful where practically every point in the system chain has no chance of affecting the part being tested. This is wildly inappropriate for speaker testing because, using impedance for example, speakers have highly varying impedance at different frequencies and will NOT sound their best unless the right system components are in place. The testing I've read about did not take any of this into consideration.

 

[WmAx]

.
As I see it, and in the very simplest challenge, these ABX speaker comparisons don't even account for the differences in efficiency of the speakers tested...How in the world can you head-to-head test different speakers used in the same system without introducing bias based on the fact that components USED may favor one or the other speaker?....Using efficiency alone...ABX tested....one will always be LOUDER than the other...and that will likely get the "best" rating by blinded listeners.
...After only a quick survey...it seems ABX testing may be useful where practically every point in the system chain has no chance of affecting the part being tested. This is wildly inappropriate for speaker testing because, using impedance for example, speakers have highly varying impedance at different frequencies and will NOT sound their best unless the right system components are in place. The testing I've read about did not take any of this into consideration.

We are really wandering off of the original thread topic now. But hey, it's interesting so far.

Double-blind testing speakers is laborious, but is possible, and done routinely. At NRC and at Harman, this is done by placing speakers on large turntables, behind an acoustically transparent curtain. The different speakers are rotated into place into the same position real-time via these devices. These studies are of course, level matched. These studies I reference, are peer reviewed by the world's leading acoustic/loudspeaker design experts. The JAES is the standard publication for these works. Like a medical journal, but for audio engineering.

The loudspeaker studies use a very large number of subjects and speakers, in order to minimize isolated error(s). The largest of such radiation/response tests conducted at the NRC, by Ian Paisely of Mirage a number of years ago, used almost a thousand test subjects, and many different speaker systems, in order to systematically deduct what was favored vs. what was not, under blinded conditions.

Some studies now use bin-aural recordings made in real room acoustics, so that tests can be conducted over headphones via software.

Personally, I do not [yet] have a turntable system. When I compare modification(s) or new ideas, I use a control and a variable. Identical except for the desired variable(s). Using a calibrated measurement microphone, I record one unit in a small anechoic chamber(or in a real room, if the test is supposed to take room effects into account), and then I make a recording of the 2nd unit. Or, I may use the same unit, before and after modification. I then level match and sync the recordings in software, then load the recordings into ABX software in order to judge for differences under blinded conditions. I use an extremely linear headphone to compare the recordings.

As for interaction with components, what interaction? A good power amplifier with low source impedance and with adequate output devices to supply sufficient current capacity under the particular loads, makes this a non-issue. A good amplifier has no audible differences with differing loads(within reason). Only poor quality, or flawed designs exhibit such behaviour(s). Of course, I refer to the perceptual research once again, in reference to audible measured artifacts in relation to amplifiers. No one has yet shown credible data that suggests otherwise.

-Chris

 

[UFObuster]

We are really wandering off of the original thread topic now. But hey, it's interesting so far.

Double-blind testing speakers is laborious, but is possible, and done routinely. At NRC and at Harman, this is done by placing speakers on large turntables, behind an acoustically transparent curtain. The different speakers are rotated into place into the same position real-time via these devices. These studies are of course, level matched. These studies I reference, are peer reviewed by the world's leading acoustic/loudspeaker design experts. The JAES is the standard publication for these works. Like a medical journal, but for audio engineering.

The loudspeaker studies use a very large number of subjects and speakers, in order to minimize isolated error(s). The largest of such radiation/response tests conducted at the NRC, by Ian Paisely of Mirage a number of years ago, used almost a thousand test subjects, and many different speaker systems, in order to systematically deduct what was favored vs. what was not, under blinded conditions.

Some studies now use bin-aural recordings made in real room acoustics, so that tests can be conducted over headphones via software.

Personally, I do not [yet] have a turntable system. When I compare modification(s) or new ideas, I use a control and a variable. Identical except for the desired variable(s). Using a calibrated measurement microphone, I record one unit in a small anechoic chamber(or in a real room, if the test is supposed to take room effects into account), and then I make a recording of the 2nd unit. Or, I may use the same unit, before and after modification. I then level match and sync the recordings in software, then load the recordings into ABX software in order to judge for differences under blinded conditions. I use an extremely linear headphone to compare the recordings.

As for interaction with components, what interaction? A good power amplifier with low source impedance and with adequate output devices to supply sufficient current capacity under the particular loads, makes this a non-issue. A good amplifier has no audible differences with differing loads(within reason). Only poor quality, or flawed designs exhibit such behaviour(s). Of course, I refer to the perceptual research once again, in reference to audible measured artifacts in relation to amplifiers. No one has yet shown credible data that suggests otherwise.

-Chris

I'll be slow to be convinced that an accurrate A/B test of speakers using SAME components is valid. Part of what we all run into in getting the most from our systems is appropriate component matching for the unique and sometimes "quirky" characteristics of speakers. Loudspeakers are not a ubiquitous sound transport mechanism and could never be tested like a set of connecting cables would. But for the moment, I'll suppose it's possible.

The intent of the thread which I started was to point out the error of broad and generic criticism of some manufactures such as Klipsch by unfounded negative statements which by inspection are in error because they don't even limit the comment to a particualr model. "Crap" if I remember correctly was in one of your posts, with no reference made to what, when, and where. If this statement has to be proven to you, then you are, in fact, representative of this type of post.

Ok, time to post now: which Klipsch models have you tested...against which other speaker brand and model, when, using what transport device or source material, what amp, how many "listeners", in what modes (digital/analog), size of the room, accoustic treatments, etc. I've really been patient in this thread so far........but now I want to see some facts before we hear another opinion.

 

[UFObuster]

We are really wandering off of the original thread topic now. But hey, it's interesting so far.

Double-blind testing speakers is laborious, but is possible, and done routinely. At NRC and at Harman, this is done by placing speakers on large turntables, behind an acoustically transparent curtain. The different speakers are rotated into place into the same position real-time via these devices. These studies are of course, level matched. These studies I reference, are peer reviewed by the world's leading acoustic/loudspeaker design experts. The JAES is the standard publication for these works. Like a medical journal, but for audio engineering.

The loudspeaker studies use a very large number of subjects and speakers, in order to minimize isolated error(s). The largest of such radiation/response tests conducted at the NRC, by Ian Paisely of Mirage a number of years ago, used almost a thousand test subjects, and many different speaker systems, in order to systematically deduct what was favored vs. what was not, under blinded conditions.

Some studies now use bin-aural recordings made in real room acoustics, so that tests can be conducted over headphones via software.

Personally, I do not [yet] have a turntable system. When I compare modification(s) or new ideas, I use a control and a variable. Identical except for the desired variable(s). Using a calibrated measurement microphone, I record one unit in a small anechoic chamber(or in a real room, if the test is supposed to take room effects into account), and then I make a recording of the 2nd unit. Or, I may use the same unit, before and after modification. I then level match and sync the recordings in software, then load the recordings into ABX software in order to judge for differences under blinded conditions. I use an extremely linear headphone to compare the recordings.

As for interaction with components, what interaction? A good power amplifier with low source impedance and with adequate output devices to supply sufficient current capacity under the particular loads, makes this a non-issue. A good amplifier has no audible differences with differing loads(within reason). Only poor quality, or flawed designs exhibit such behaviour(s). Of course, I refer to the perceptual research once again, in reference to audible measured artifacts in relation to amplifiers. No one has yet shown credible data that suggests otherwise.

-Chris

One more observation on this study technique: using the medical model as an example, this sounds a lot more like post-marketing research. Finding out about what sells....or more to the point...what the consumer wants...not what is good.
Given that most buyers will not put forth an effort to do well-considered component matching and that they will assemble "generic" systems or off-the-shelf receivers without consideration of their capabilities only their features not to mention using dedicated amps, etc, then setting up a testing system to find a set of speakers that work best in this environment sounds very sensible and VERY profitable. I would certainly buy into this approach, too, if I didn't already have a higher expectation of what I wanted in quality sound. Maybe we'll soon see a "consumer reports" issue on the top-ten best sounding speakers.....if this research is really true...then we should all NOT trust our ears. We can just look at the list....look up what we can afford (ie, find our "price point")......and then buy it! Happiness to all!

 

[WmAx]

I'll be slow to be convinced that an accurrate A/B test of speakers using SAME components is valid. Part of what we all run into in getting the most from our systems is appropriate component matching for the unique and sometimes "quirky" characteristics of speakers.

It makes not much sense for you to continue your disagreement on this matter, since you seem to not know about basic electrical interactions, nor the basic percpetual research and how it relates to such.

Ok, time to post now: which Klipsch models have you tested...against which other speaker brand and model, when, using what transport device or source material, what amp, how many "listeners", in what modes (digital/analog), size of the room, accoustic treatments, etc. I've really been patient in this thread so far........but now I want to see some facts before we hear another opinion.

I did not intend to elaborate, but this is still fun, so I'll play along. The post I linked, where I said the Klipsh speakers were deserving of the 'crap' badge, was in the spirit of fun as well. So I may as well continue.

I did not test against any other model speaker. I use a linear headphone(modified Beyerdynamic DT880 or a Sony MDR-7506) as a baseline, with the same source material in ordinary listening comparisons. I do not remember specifically which headphone I used at that time. But both units are a reference for me. This provides me for the most consistent comparisons in uncontrolled conditions(listening at some one's home or in a store). I use a standard compilation CD (consisting of various high quality acoustic instrument tracks that I have gathered from various CDs) I have used for many years, so that I am very familiar with the tracks. The CD is also loaded into a portable music player, which I use on the headphones. The model I referred to was a Klipsch Reference model: RF-83. About 6-12 months ago. It was in a room that appeared to be about 30' D x 20 W. Nearly the entire side walls (left and right side) had large diffusion characteristics, and the rear of the speakers were about 2' from the rear wall, and 4' from the side walls. Wall to wall carpeting. A pretty good environment for this particular speaker, considering the off axis behaviour. The amplification was one of Denon's upper model receivers, I do not remember which model. The source was a Sony SACD player, I do not remember which model. Such details are hardly important to me; Denon is always a good measuring linear product, and their is no reason to presume a problem with the Sony player. The resonances from the cabinet sounded exactly like, well, resonances(as I have heard and identified in proper controlled blind test scenarios). Timbre coloration. Substantial ones. In addition, the HF was extremely directive(being a horn tweeter), and I don't like that one bit, but it is not the primary basis of my dislike. But at least I did not subjectively detect any treble problems. Seems to be a well designed horn cavity essentially free of resonance issues. But the cabinet resonances that I perceived were unforgivable, for my standards. But I should point out, nearly all speakers do not come close to meeting my standards. I don't know of any for the low price (based on the price points of speakers I have heard that were acceptable) point that the RF-83 sells for, that I find acceptable for my personal use.

-Chris

 

[WmAx]

One more observation on this study technique: using the medical model as an example, this sounds a lot more like post-marketing research. Finding out about what sells....or more to the point...what the consumer wants...not what is good.

I don't see where you are in a position to be critical of the research, since you do not appear to know anything about the research.

-Chris

 

[billy p]

I don't see where you are in a position to be critical of the research, since you do not appear to know anything about the research.

-Chris

Well, he stated that he's in the medical field with a degree and that should be enough.

 

[WmAx]

Well, he stated that he's in the medical field with a degree and that should be enough.

How does him having a medical degree make him informed of loudspeaker perceptual research?

-Chris

 

[billy p]

Exactly!!

 

[Knight]

I'll say this much, I just got my degree in electrical engineering and some of the people I work with know very little about speaker other than simply how they work. I can analyze the hell out of a speaker to pick the best one but thats no good. Double blind test don't work here because the speakers aren't in YOUR room. One speaker might sound better in the test but not in your home. You gotta go by what your ears tell you. To me Klipsch has a unique sound that you will either love or hate. Listen to them and decide for yourself.

 

[rolyasm]

I like Klipsh. I compared some 82's against my 4 year old Polks, and I preferred the Klipsch. Up to this point, I had totally loved my laid-back Polks. However, with the Klipch, I could hear much more detail. I used an SPL to match levels, not perfect, but effective. Still, the Klipsh brought out much more background detail.

Also, Klipsh have a very high sensitivity compared to most speakers. I thought that this meant they played louder at equal amp ratings. Isn't it easier to drive a more sensitive speaker? Here is a graph I found online describing this point. To play the same "loudness" you need 75 less watts in the more efficient speaker. Am I missing something Knight? You said earlier that they were very hungry speakers for power. Please help me to understand this better.


Speaker Sensitivity ........... rating Power needed
................................ to produce a given volume
Speaker A 85 dB ........ 100 watts
Speaker B 88 dB ........ 50 watts
Speaker C 91 dB ........ 25 watts

 

[Knight]

Rolyasm, you are correct with dB think but when I sold them I found that they bleed amps dry. This was because of the one thing I do like about Klipsch, they are able to play very loud, louder than most. It reminds me of a pair of Cerwin Vegas I have a while back. They had a sensitivity of 104 dB and they needed little power to be loud but at very high volumes they eat tons of power. This comes from the idea that depending on the speaker design at high volumes the impedence of the speaker can drop which causes the amp to put out alot more juice. If the amp runs out of power at high volume the high voltage across the voice coil with its lack of power to push the cone out further can arc and the impendence drops alot. This can damage the speaker and the amp. To me, since I like it loud, when you have a speaker that can play loud at high volumes I like to have plenty of power so I don't damage the speaker. Now when I was using my Mirages as my front speakers I didn't have that problem because I know the speakers could play as loud as the Cerwin Vegas could so I never really cranked them up and even though they needed more power at the same volume the top end was never explored so I never bleed the amp dry. It's kind of a head game.

Another thing to remember is that the sensitivity of the speaker isn't always linear and some highly sensitive speakers become very unsensitive at higher volume.

I'm a big fan of buying speakers on the sound and I pay less attention to the numbers. When you think about it most recievers on the market are in the 90-140 watt range and if you have speakers that need alot of power you won't get too much from going from 100 watts to 140 watts. Your better to get a high current amp as they tend to produce better sound.

I will agree with your analysis that the Klipsch are very accurate and sharp for that matter but I'm just not a fan of horn tweeters. With the exception of the Cerwin Vegas I found that you can beat the hell out of them like no other and they will keep smiling and ask for more.

 

[tbewick]

On the comments of Knight:

http://bruce.coppola.name/audio/Amp_Sound.pdf

This is from Rip Van Woofer's list of links.

I remember reading that Floyd Toole recommended a 200 watt into 8 ohm/400 watt into 4 ohm power amplifier as a minimum for speaker testing. I would say that choosing accurate speakers, then compensating for acoustic considerations directly, through doing room treatments and using appropriate sound processing tools, is the most sensible way to proceed.

 

[Seth=L]

Rolyasm, you are correct with dB think but when I sold them I found that they bleed amps dry. This was because of the one thing I do like about Klipsch, they are able to play very loud, louder than most. It reminds me of a pair of Cerwin Vegas I have a while back. They had a sensitivity of 104 dB and they needed little power to be loud but at very high volumes they eat tons of power. This comes from the idea that depending on the speaker design at high volumes the impedence of the speaker can drop which causes the amp to put out alot more juice. If the amp runs out of power at high volume the high voltage across the voice coil with its lack of power to push the cone out further can arc and the impendence drops alot. This can damage the speaker and the amp. To me, since I like it loud, when you have a speaker that can play loud at high volumes I like to have plenty of power so I don't damage the speaker. Now when I was using my Mirages as my front speakers I didn't have that problem because I know the speakers could play as loud as the Cerwin Vegas could so I never really cranked them up and even though they needed more power at the same volume the top end was never explored so I never bleed the amp dry. It's kind of a head game.

Another thing to remember is that the sensitivity of the speaker isn't always linear and some highly sensitive speakers become very unsensitive at higher volume.

I'm a big fan of buying speakers on the sound and I pay less attention to the numbers. When you think about it most recievers on the market are in the 90-140 watt range and if you have speakers that need alot of power you won't get too much from going from 100 watts to 140 watts. Your better to get a high current amp as they tend to produce better sound.

I will agree with your analysis that the Klipsch are very accurate and sharp for that matter but I'm just not a fan of horn tweeters. With the exception of the Cerwin Vegas I found that you can beat the hell out of them like no other and they will keep smiling and ask for more.

I agree, the Klipsch Reference Series is one of those speakers that requires a very capable amplifier for those higher volumes. For the Klipsch to sound good they need a very linear amp that can handle all the way down to 2 ohms at times. This is one of the reasons that people love the sound from tube amps paired with Klipsch, because the Klipsch are efficient they can good loud with lower amounts of wattage, and because class A amps are very linear and can take some torture (transistors can't take much torture for more than a few seconds before they are toast).

 

[memonmz]

Klipsch owns, Mirage, Energy, Athena, Spherex (not familiar with them), and Jamo (and its subsidiaries).

Harman owns, JBL, Infinity, and Revel speaker companies.

I think API was a recent acquisition by Klipsch.

Hmmm Klipsch owns Mirage? interesting... To me atleast Mirage speakers sound completely opposite of Klipsch, Mirage is nowhere near as in your face sound

 

[memonmz]

I sold Klipsch and I don't like them at all. But what I can tell you is that I noticed while selling them there were a small group of people that liked them alot. The horn tweeters drive me crazy but some people love them. You should listen to them and no one else. Just keep in mind that they tend to be power hungry speakers but if you have a good amp you'll be ok. After you listen to them next to an speaker that has say a bipolar design you will figure out which style you like, at least that's how I decided.

Hmmm I actually did that compared teh RF-62 and the Def Tech 7006 side by side and to me I liked that the Klipsch provided me with a much more in your face sound while still allowing attention to details and intricacies in the sound

 

[Seth=L]

Hmmm Klipsch owns Mirage? interesting... To me atleast Mirage speakers sound completely opposite of Klipsch, Mirage is nowhere near as in your face sound

All those companies aside from JAMO are part of API group which is/was based in Canada. Klipsch bought them out or they combined forces recently.

 

[Joe Schmoe]

Hmmm I actually did that compared teh RF-62 and the Def Tech 7006 side by side and to me I liked that the Klipsch provided me with a much more in your face sound while still allowing attention to details and intricacies in the sound

A speaker that can be described as "in your face" is not a speaker I will ever like. I hate that kind of aggressive sound. I guess that is why I hate Klipsch.

 

[UFObuster]

How does him having a medical degree make him informed of loudspeaker perceptual research?

-Chris

Not at all...but what it does make me is an expert in exposing bogus studies that don't have excellent control. I'm not an expert on angiotensin receptor blockers either, but I can certainly tell if the studies that claim their effectiveness are well designed or not. The "perceptual" issue is what leads me to categorizing this type of info as more like post-marketing reasearch.....and now I'm assuming that you guys know the difference....more directed at how a product is perceived in its desirability or usefulness may not really ever tell us anything about its actual quality. It can tell us what a certain cohort likes or dislikes. But do I want to quantify speaker quality by an "average" group of listeners? I think not. Yes, it may be good information but it doesn't inform me much about the sound quality. Like I said in an earlier post, if you could really quantify this quality sound...then there is no argument on what is better....You go get the lastest list of "best speakers"....read down the list to your price point....buy it, take it home, plug it in to whatever generic system you have because the opinion is that the components don't matter and listen....you know it's good because someone told you it was!