Corner-Horn Imaging FAQ

[Chris A]

There are a couple of ideas in play here...

 

Quote

...are there any that would be "invisible" to the Khorn bass[?]...

 

Basically, you don't want to absorb frequencies that don't have anything to do with improving corner-horn imaging. I use Auralex Sonofiber tiles; they do the job well but more discussion on that, below. If you look at this material's absorption properties v. frequency, you will find that it controls the desired frequencies well. Almost nothing will attenuate frequencies below ~150 Hz except room modes, flexing walls (extreme lf below ~25 Hz) and active cancellation arrays (ref: the "double bass array" concept).

 

Quote

...since PWK warned that the wall should be unobstructed for 4 feet out from the corner...

 

I think that PWK was saying the same thing that I've tried to say earlier - you need smooth and obstruction-free side walls and front wall around the corner horns (i.e., further than 4 feet). If you don't have that condition then you need to use absorption material to control any early reflections.

In fact, reflections of less than 0.7 ms need to be attenuated as a rule - that's about the first 9.5 inches or 25 cm from the midrange horn's mouth. I'd use absorbent material in these areas no matter what kind of wall is in proximity since these very early reflections actually distort the stereo image instead of merely changing the overall timbre and muddy percussive sounds. For the Khorn, that's why I recommend putting some fuzzy material on top of the top hats and letting it extend over the front a couple of inches - the Khorn K-400 horn loses vertical pattern control at about 1700 Hz, and that energy winds up refracting off of the top hat and the front of the Khorn.

 

I use about 2 feet of absorption on the side walls closest to the mouths of the K-402s and Jubilee bass bin. It works amazingly well. I haven't yet tried this for my Khorns (Shinalls) upstairs yet - I'd guess that I'd use about the same amount if I was time-aligning the tweeter/midrange drivers. If not time-aligned, then we have more issues to deal with.

 

Roy D. said that there should be no strong midrange reflectors between the near-field corner horns. This is especially true if the corner horns are less than ~22 feet apart or less than about 11 feet from a center-channel speaker cabinet. Midrange early reflections negatively affect corner-horn imaging and timbre balance, especially for speakers like the Khorn which spreads its midrange energy around the near field a great deal more than the K-402 horn (or the K-510 horn above ~1700 Hz).

Maybe one of the reasons why PWK wanted the long dimension of the room for Khorns was to avoid the near-field midrange reflectors such as furniture, equipment racks, and center speakers. Just placing the speakers further apart increases their imaging performance, once they are further apart than Haas zone reflections.

 

Chris

[emmvette]

"What Can I Do About Ceiling Bounce of Klipsch Heritage Midrange Horns (I.E., K-400 Series Horns)?"

Here is what I did; I had this issue, but didn't think about using accoustic treatments so I just raised the height of my room. It was so much easier.

Kidding of course, but it is nice to see this thread. It's been interested to move my Khorns around. They ended up in this room because they sound the best there. Lots of room, high ceilings, carpet, but not perfect corners. I had to do a little fiddling around with the back, but they do sound good and make decent bass as well.

[Chris A]

You've got a pretty good example there of a Khorn-friendly room (after the tree is recycled, of course [8-)] )

Chris

[WMcD]

"What about Amplifiers and Corner-Horn Imaging Performance?"

I've saved this juicy subject up to this point for a good reason: in order to understand the effects of different type of amplifiers on corner-horn performance, you need to understand the effects of "early reflections", discussed above, and the treatments available to recover your stereo imaging performance in rooms with cluttered areas between the speakers and non-smooth front and side walls. Once you understand the psychoacoustic effects on imaging of corner-horn speaker midrange horns/drivers, then a more rational discussion on amplifier effects can be achieved.

"What Are the Issues Related to Amplifiers and Corner Horns?"

Amplifiers that exhibit high output impedance have the effect of providing a "reverb effect" in-room, especially if the room is small and relatively live acoustically. What kind of amplifiers have relatively high output impedance? Tube or valve-type amplifiers.

"Why is This an Issue...What is Happening?"

The reverb effect is due to strong room reflections back to the horns/drivers themselves, which are much more efficient than direct-radiator speakers at converting electrical energy into acoustic energy - and back again (...i.e., they are acting like microphones). The Khorn, for example, converts about 10% or more of its input electrical energy into acoustic energy (according to PWK's own calculations and measurements), while cone-type speakers typically are only 0.1% efficient or less. Planar speakers, like electrostatic and Magneplanar-like speakers, are even less efficient. BTW: the same reverb effect happens with headphones. [special thanks to Bob Carver on explaining this phenomenon.]

"So Why are We Talking About the Efficiency of Corner Horns and Imaging with Some Tube Amplifiers?"

Because the horns/drivers themselves are 100x more efficient at converting acoustic reflected energy back into electric energy (to your amplifier's output terminals) than direct radiators are. This means that the room's reflected energy goes right back to the amplifer's output terminals.

"So What's the Issue with Amplifiers and Horn-Loaded Speakers?"

Nothing, as long as the amplifier has low output impedance (like virtually all SS amplifiers and most higher-forward-gain tube amplifiers with feedback--like multi-stage push-pull designs). But if your amplifier has relatively high output impedance (i.e., "zero feedback" and SET-type tube amplifiers, particularly output-transformerless or "OTL" types), what happens is that the amplifier tries to "hold the driver diaphragm in place" against the reflected acoustic-electric energy. The net result is a "reverb" effect that is sensitive to SPL coming back to your corner horns from the room.

"So What's Wrong With That?"

Well, for starters, it's artificial and you can't turn it off unless switching to a lower output impedance amplifier, or switch to a much larger listening room with high ceilings. PWK talked about measuring speakers with a "rubber yardstick" when you start to depart from a "live music reproduction" standard. That conversation is apropos here. If accurate reproduction is the standard by which PWK designed and built his speakers, then amplifiers with high-output-impedance used with very efficient corner horns (where the reflected acoustic waves tend to have the highest amplitude and tends to pile up) will lead to distortion - non-harmonic distortion - which is the worst kind of all.

"So Why Do So Many Corner-Horn Owners Use Tube Electronics That Have High Output Impedance?"

I believe you can answer that question yourself now. Another way to achieve the same effect is to place a reverb unit on in front of your amplifier input terminals - at least you can turn it off when you get tired of it.

"So What Other Issues are There With SET Tube-Type Amplifers?"

Low Power SET-type tube amplifiers usually have too little amplifier power headroom--even for 105 dB/W-M corner horns, thus leading to fast-transient "soft clipping" that some tube enthusiasts apparently like. PWK's "rubber yardstick" comments are apropos here. It's preferable to not have any clipping effects at all if we are to retain an accurate-sound-reproduction yardstick. Also, SET amplifiers exhibit a much larger amount of harmonic distortion (i.e., even harmonics only). It is preferable to not have these harmonics added to our stereo's output, since the magnitude of harmonic distortion is an indicator of the magnitude of intermodulation distortion (IMD), which is non-harmonic and very detrimental to the quality of sound reproduction.

"Soft Clipping" of Tube Amplifiers that Masks Their Clipping Distortion but Introduces Harmonic Distortion

"So Why Do Class "A" Tube-Type Amplifiers Sound So Good With Corner Horns?"

Well, there another effect taking place. The acoustic effect is a "softening" on the hard-edged sound that tends to accompany multi-stage BJT solid state amplifier output. When the exact psychoacoustic effects are identified, we will have a much better understanding of why many people prefer a "tube sound" when using horn-loaded midrange and tweeter loudspeakers. The net effect is to soften the objectionable element of these types distortion products (and just-plainly-harsh recorded sound). Why is that recorded sound not "softened" by the recording, mixing, and mastering engineers in the first place (...by placing their microphones in different locations...)? Probably because most monitors used in recording studios are apparently direct radiating types (e.g., the Yamaha NS10M).

"Why Does the Use of Direct Radiator Monitor Loudspeakers in Mixing/Mastering Control Rooms Matter?"

Probably because direct radiator speakers partially "knock the edge off" the music they reproduce - just like low-power SET amplifiers coupled to horn-loaded speakers with their transient-soft-clipping characteristics, except that mixing and mastering engineers probably don't hear the resultant hard edge that horn-loaded speaker owners do later with SS amplifiers.

"Can We Get Tube-Type Sound Without Distortion?"

The answer is yes: there are new JFET and MOSFET single-stage amplifiers made by Nelson Pass ("First Watt" series) that sound like tube amplifiers, but are solid state AND they have relatively low output impedance AND they measure very cleanly with respect to measurable distortion like THD, IMD, noise, and other figures of merit. More on that subject in another thread.

There are also very good tube-type designs - usually those with some feedback to lower their output impedances. They are very expensive compared to basic SET designs.

"So What's the Bottom Line on Amplifiers and Corner-Horn Imaging Performance?"

Amplifiers with very low distortion, low noise figures, and low output impedance but with high slew rates - like suitably designed tube amplifiers and well-designed single-stage MOSFET and JFET amplifiers will provide the most uncolored AND pleasant listening corner horn imaging performance, retaining acoustic "detail" so dearly sought after by many audiophiles, and providing the original dynamic range of the actual performance due to the horn-loading advantage of corner horns over direct radiator speakers.

Next up: discussion on the Khorn K-400 and Klipschorn Jubilee K-402 midrange horns...

Thank you Chris for a very insightful description. I missed it the first time around.

The reverb effect is interesting. Let me point out some things that I understand differently.

Assuming the reverb effect is significant:

My understanding is that a starting point in transfer efficiency, that is into the horn, is that the generator's impedance is to be the same as the load impedance. We have to look at the acoustic impedance of the source and that is a complicated matter of Bl, the voice coil resistance, and the electrical impedance of the output of the amplifier. The equations in the Bell Lab paper shows this.

An issue there is the electrical impedance of the amp and there is an adjustment factor for it. (Though there is something in the paper (not before me) on an optimal. Something maybe you can help me with.

Anyway, the major design factor is to match the driver impedance (on an acoustic level) to the throat impedance. We see that the variation in acoustic impedance of a finite length horn and try to put the effective driver impedance in the middle of the wobbles. From memory, it takes a 6:1 or 1:6 mismatch to lose 3 dB and therefore we can live with a peak to trough ratio of 36 in acoustic impedance to get plus or minus 3 dB. Not the point here but I'll mention it.

Let me switch over to transmission line theory, which is pretty much the above. But the other part of transmission line theory has to do with echos. Let's look at the situation where there is a mismatch at the load end of the transmission line (mouth of the horn). There is a reflection. And I think it is the same case if there is an extraneous signal injected into the mouth of the horn (from the room, in your discussion).

I believe that the variations in throat impedance of a finite horn is due to these reflections.

But what happens to the energy in any case. Lets consider the driving source and throat on the left and the mouth on the right. The horn is the transmission line in the middle.

Let's consider the reflection and/or room source creating a pulse into the right side of the transmission line. It goes back to the left and encounters the diaphragm and the electrical system. If we've set it up for maximum transmission efficiency (pulse output to the right), the acoustic impedance is the same as the characteristic impedance of the line.

This means that the left-traveling pulse (reflection) encounters a perfect acoustic load of the diaphragm/voice coil/ amplifier system, and is absorbed. The pulse does not get reflected back down the transmission line. Therefore there is no reverb.

In my understanding, this is the dual role of the left side of the system driving the transmission line. A good match of generator impedance to transmission line impedance provides optimal energy transfer into the line and also absorbs reflections. And we're not yet to the point of discussion.

You mentioned various amplifiers and their characteristics. But let me point out that the acoustic impedance of the driver system includes the voice coil impedance (more on this later).

I agree that various amplifier designs have different electrical impedances. And this is part of the "n" factor in the Bell Lab paper for the optimal throat size, which is really to say the acoustic impedance of the system on the left or diriving system. Note that Keele's paper on using T-S parameters mentions it and assumes that n=0 which is to say zero output electrical impedance of the amp.

But . . but . . but. it seems to me that your description isn't quite accurate in a few ways.

My first thought is that the output impedance of the amp is considered, through "n" when we design the system. While somewhat complicated, the electrical output of the amp adds to the voice coil impedance in the overall consideration of the effective acoustic impedance of the driver. Therefore it is not quite corect to say that a low output amp, or a high output amp is superior. It is just a matter of what it is designed for in the horn driver combo.

There is also something I don't quite understand in your description of how an amp with a high output impedance "holds" the diaphragm. That is to say, an amp without a feedback loop.

An amp with a feedback loop will "hold" the voltage at at it's output terminals to zero (assuming no music input to the amp) even if the speaker terminals are pushing or pulling current due to reflections from the transmission line and the motor of the driver (acting like a microphone). You're writing is that an amp with a higher output impedance and no feedback loop can do that. As the song says, "But I don't think so." Smile.

In my view, any amp can not hold the diaphragm. Even if the amp has a feeback loop, it can't control the diaphragm because the voice coil resistance gets in the way. The resistance gets in the way.

OTOH, I think we can agree that a speaker driver with a very high Bl product can restrict movement of the diaphragm. But this is just a matter of the resistance value of the acoustic characteristics discussed above.

- - -

Thank you for considering my comments, if nit picking.

I very much enjoyed your long posts and very much appreciate your devoted efforts to make things clear, technical accuracy, intellectual approach, and superlative insights.

Maybe we can get together sometime.

WMcD

.

[Chris A]

Quote

There is also something I don't quite understand in your description of how an amp with a high output impedance "holds" the diaphragm. That is to say, an amp without a feedback loop.

An amp with a feedback loop will "hold" the voltage at at it's output terminals to zero (assuming no music input to the amp) even if the speaker terminals are pushing or pulling current due to reflections from the transmission line and the motor of the driver (acting like a microphone). You're writing is that an amp with a higher output impedance and no feedback loop can do that. As the song says, "But I don't think so." Smile.

In my view, any amp can not hold the diaphragm. Even if the amp has a feeback loop, it can't control the diaphragm because the voice coil resistance gets in the way. The resistance gets in the way.

 

Hi Gil,

 

This is the description that Bob Carver used in his description of tube-type amplifiers (implying high output impedance amps). You have to wade through a few minutes of this video before his interview comes around to his prior "Silver Seven" amps:

http://twit.cachefly.net/video/htg/htg0029/htg0029_h264b_864x480_500.mp4

 

I believe that your entire post above is concerned with this subject, so I'll wait until you have a chance to view this before I comment on anything else.

 

The one thing that Bob didn't say was that higher efficiency speakers (like horns) multiply the reverb effect due to their, well, higher efficiency.

 

I immensely enjoyed his related discussion on his developmental "six foot headphones" that obviously were not highly efficent, so to absorb the acoustic energy bounce off human eardrums.

More comments can be found here: http://community.klipsch.com/forums/p/140333/1428680.aspx#1428680

 

Chris

[WMcD]

Okay, I watched the interview. The last minutes are what we are discussing and I watched it twice and made some notes.

It seems to me that Carver is making some condradictory statements and therefore it is diffuicult to interpret.

1) Starting in the great amp shoot out and in a later section he is saying that SS amps and tube amps can be made to sound the same and indistinguishable. In the case of his amps they sound, -- he used some superlative. Let us say, very wonderful.

1.1) In the shootout, A-B-X testing was used to vindicate that Carver's amp sounds the same as a big Mac, or others. In my view, this is flirting with the concept that all accurate amps sound the same. It may be that all accurate amps sound harsh, dry, analytical, etc. A-B-X testing does not enter the discussion again.

2) But later, he is asked to comment on tube amps and he says he likes them better, or some such words. This seems to contradict 1) and we're not hearing about 1.1 either.

3) He comments on a feedback loop in in tube amps. But he says the delay in the loop is relatively small - it is the delay in the room reverberation which dominates. This is hazy.

4) It looks like he ducked the question about SS amps and feedback loops and switched to output impedance -- as if to imply that SS amps don't have this . Nelson Pass is a widely respected designer of SS amps. He speaks of what he calls a "hall of mirrors" effect of feedback loops. He too searches for what makes an amp sound good or bad. So Pass thinks there is something going on in SS amps with maybe too much feedback (and bad) . . . but Carver implies this sort of thing is indiginous to tube amps (and good).

5) In the interview Carver has mentioned "spaciousness" and reverb in several situations. Like his speaker and then tube amps. He is saying that listeners like it. A fair reading of Carver's preference for tube amps is that they and a speaker induce reverberation. This may be well and good but now we have a situation where amps are subjectively better because they allow artifacts to be introducted into the system. If this is so, we have to find a new metric for inducing good artifacts. Accuracy is then a bad thing.

6) Chris it talking about horns and reverb and whether a given amp allows them. In my view this requires a careful consideration of the system, per my post. But Chris is relying on Carver on a technical point regarding reverberation (fair to say a bad thing?) and Carver in another matter likes reverberation.

The above is why I find the contradictions so great that no conclusions can be reached. Smile.

WMcD

[Chris A]

Quote

The above is why I find the contradictions so great that no conclusions can be reached. Smile.

 

Gil, I'm sorry that this particular concept (reverberation effect from high-output-impedance amplifiers and high-efficiency speakers) is so difficult to understand: it's pretty clear to me. Carver has added a very insightful observation to my list of insightful observations.

 

But it's okay, since it's all just audio. To each, his own...

 

Here is another article of related interest on acoustic feedback...in this case a thesis:

http://publications.lib.chalmers.se/records/fulltext/107789.pdf

 

Chris

[artto]

Well, there are two fundamental issues here.

Accuracy, and Illusion.

IMHO, at our current level of understanding and knowledge of
the laws of physics, we have yet to figure out a way of placing two material
objects in the same place and time. The same goes for sound, and sound
reproducers. And it also applies to the capture (recording) of, and “attempted”
re-creation of the sonic event. Yeah, yeah, I know, we’re entering the realm of
quantum physics here, but let’s keep the idea simple right now if for no other
reason than understanding or to make a point.

Over the past several decades there has been substantial debate
on the tube verses transistor “sound”. Since I was personally involved with
this at a very early stage (late 60’s, early 1970’s) I’ll throw my two cents
in.

First the fundamental issue with the devices themselves.

When I was attending college at U of I, Urbana, I got into an argument with some young
“audiophiles” about the same age as I. They kind of pissed me off. After all, I
already had what I thought at the time was some of the best audio gear on the
planet (Thorens turntable, Crown preamp, Crown DC300, Cornwalls). In
particular, these guys made a point of the DC300 and Cornwalls being a very bad
combination, and that I should be using a vacuum tube pre/power amp with them.
These guys also happened to be audio dealers working from a house selling Fulton speakers and Futterman
OTL amplifiers.

After leaving in frustration and feeling somewhat insulted,
I talked to one of the engineers/owners at the company where my wife worked.
They made amateur radio telecommunications equipment. I told Bill my story and
to my surprise he said, “Well of course! Tubes have a naturally more linear
response in the audio range than transistors.” At the time I didn’t know
exactly what this really meant, but I then sought out someone at the electrical
engineering department who was more versed specifically in audio and got exactly
the same answer!

Second, as I later discovered, transistors had much more
gain than tubes. And this additional gain allowed audio designers to do things
with transistors that they couldn’t do so easily with tubes. One of those “things”
was to use the extra gain in the feedback loop to reduce distortion and thereby
make them operate more “linear” in the audio range.

(conceptual example: vacuum tube gain = 25dB, if 20% (5dB) of
that gain is used for feedback to reduce distortion you’re left with 20dB gain.
Transistor gain =100dB, if 80% (80dB) is used for feedback which in theory
reduces distortion even further, you still have an amplifier with 20dB gain)

The problem with this is that with simple “static” test
signals, and by “static” I mean something like a simple sine wave or two,
non-changing in frequency, amplitude or phase, this idea of introducing ever
higher amounts of feedback for reducing distortion works very well. However, as
Nelson Pass pointed out, this can create a “house
of mirrors” when a signal such as music, which is constantly changing in
frequency, amplitude and phase are introduced to the amplifier. Part of the
music signal which is already done and gone, and on its way to the speaker, has
been “fed-back” to the input and mixed with what is now a “different” signal.
And this happens over and over repeatedly thousands of times per second. What
you have in the end is truly a “house of mirrors”.

So the first problem is one caused simply by transistors
allowing designers to do things (mistakes, if you will) to a degree not
possible with tube amps. The equipment “tested” very well, but sounded awful with
music. Now, if we could somehow delay or fast forward the signal
process and time align the “feedback” with actual part of the signal from which
it came, the story would be much different. Isn’t that kind of where “digital”
is really supposed to take us?

Then there’s the issue of the amplifier/speaker interface.
Transistors, naturally being a lower impedance device allowed amplifier
designers to eliminate the output transformer thereby “direct coupling” the
amplifier to the speaker more easily with the drivers of today than tube amps usually do. While this certainly has its advantages depending on
the setup, consider this: What is a speaker, really? It has some kind of “voice
coil” or motor. What happens when this “motor” tries to change direction,
amplitude, or frequency, or just plain stop? Well, the fact of the matter is “it
can’t”, at least not perfectly. It continues to move to some degree, doing what
it “was” doing. And when it moves it also acts like a motor, and that motor is
producing energy. And where does that energy go? Why back to the amplifier, of
course. And there the signal can get “re-introduced” into the “direct coupled”
amplifier, and of course, back into the feedback loop.

What I’ve presented here is very “conceptual” and obviously
there’s more to it than that. But in the early days that was basically what was
happening with a lot of solid state amplifiers. Today the situation doesn’t
seem to be as bad as we have apparently learned from our mistakes. Yet sonic
differences remain. And then there are other issues such as overload/clipping
characteristics, harmonic and frequency intermodulation distortion content, noise
characteristics and other things.

Accuracy or illusion? Which is it that we really want? Are
they both one in the same? Or are they something that can never be the same? What
about the component which has “distortions” that result in creating more of an
illusion of the real thing?

In either case, I don’t think we’re there yet. But, this is all
about creating the illusion of a live
performance, isn’t it? If it can create the illusion, is that accurate? Or does
truly accurate reproduction of what’s on the recording destroy the illusion? Until
we can create something like the holodeck on Star Trek, something that can be
observed consistently by everyone, I think it’s something we all have to decide
for ourselves.

For further reading on the subject of recreating the original
“aural event” I suggest that if you haven’t already done so, go back to the
beginning. These ideas and problems have already been investigated long ago in
the 1930’s by Bell Laboratories in their Symposium on Auditory Perspective
included in the Klipsch Audio Papers. It will give you a true understanding of
what is required to “re-create” an auditory event (without the recording process
getting further in the way of things) and what the limitations are, none of
which have moved beyond that to this day.

And as far as Bob Carver goes, PWK (under the alias of O.
Gadfly Hurtz), author of “The Ultimate LSH Loudspeaker” is one of the few
people allowed to poke some fun at him in the AES Journal regarding one of his
more stupid moments. Its things like this that lead me to believe Bob Carver is
not capable of engineering anything audio however his products tend to be very
cool looking, and like Amar Bose is obviously an excellent businessman. And as
far as Carver being able to make a solid state amplifier sound like a tube amp,
well, I have yet to hear it.

The Symposium on Auditory Perspective can be found here in “The
Paper” starting around page 220.

http://bit.ly/lzGIa

The Ultimate LSH Loudspeaker is on page 37.

[garyrc]

...

Second, as I later discovered, transistors had much more
gain than tubes. And this additional gain allowed audio designers to do things
with transistors that they couldn’t do so easily with tubes. One of those “things”
was to use the extra gain in the feedback loop to reduce distortion and thereby
make them operate more “linear” in the audio range. ...

The problem with this is that with simple “static” test
signals, and by “static” I mean something like a simple sine wave or two,
non-changing in frequency, amplitude or phase, this idea of introducing ever
higher amounts of feedback for reducing distortion works very well. However, as
Nelson Pass pointed out, this can create a “house
of mirrors” when a signal such as music, which is constantly changing in
frequency, amplitude and phase are introduced to the amplifier. Part of the
music signal which is already done and gone, and on its way to the speaker, has
been “fed-back” to the input and mixed with what is now a “different” signal.
And this happens over and over repeatedly thousands of times per second. What
you have in the end is truly a “house of mirrors”.

So the first problem is one caused simply by transistors
allowing designers to do things (mistakes, if you will) to a degree not
possible with tube amps. The equipment “tested” very well, but sounded awful with
music. Now, if we could somehow delay or fast forward the signal
process and time align the “feedback” with actual part of the signal from which
it came, the story would be much different. Isn’t that kind of where “digital”
is really supposed to take us?

What is a speaker, really? It has some kind of “voice
coil” or motor. What happens when this “motor” tries to change direction,
amplitude, or frequency, or just plain stop? Well, the fact of the matter is “it
can’t”, at least not perfectly. It continues to move to some degree, doing what
it “was” doing. And when it moves it also acts like a motor, and that motor is
producing energy. And where does that energy go? Why back to the amplifier, of
course. And there the signal can get “re-introduced” into the “direct coupled”
amplifier, and of course, back into the feedback loop ....

Accuracy or illusion? Which is it that we really want? Are
they both one in the same? Or are they something that can never be the same? What
about the component which has “distortions” that result in creating more of an
illusion of the real thing? ...

If it can create the illusion, is that accurate? Or does
truly accurate reproduction of what’s on the recording destroy the illusion?

Is the point you made about excessive feedback within the amplifier essentially a problem of TIM distortion, or is there more to it than that? Wouldn't the time delay be so short that the problem would be heard as distortion, rather than a echo like "hall of mirrors?" Later, once the "motor" that is the speaker starts sending back an altered signal to the amplifier, then I can see how we could hear that as delay + distortion. Does damping factor have something to do with all this? I've heard that high damping factors are more useful in advertising than in music reproduction.

If the recording engineers are unaware (as someone on the forum said) that their microphone diaphragms, boards, studio acoustics etc. are subtlely distorting the sound because they are listening through speakers that veil the sound and make it prettier, then I would say that accurate reproduction of what's on the recording might destroy the illusion, and make the sound more unpleasant when played with articulate speakers at home.

[Chris A]

Quote

Is the point you made about excessive feedback within the amplifier essentially a problem of TIM distortion, or is there more to it than that? Wouldn't the time delay be so short that the problem would be heard as distortion, rather than a echo like "hall of mirrors?" Later, once the "motor" that is the speaker starts sending back an altered signal to the amplifier, then I can see how we could hear that as delay + distortion. Does damping factor have something to do with all this? I've heard that high damping factors are more useful in advertising than in music reproduction.

 

The degree of delay in feedback signals strongly affects perception as noted above, but it isn't a linear function of delay. There are breakpoints at 0.7 ms, 10 ms, and somewhere in the range of 20-40 ms depending on content and background. In each of these delay regimes, perception of delay effects is strongly differentiated.

 

By the way, here is an interesting article written by Nelson Pass on the subject of harmonic distortion in amplifiers as a function of increasing feedback.

 

This is the first article that I've seen on this subject that has actually measured the negative effects - a relative increase in higher order harmonics.

 

Also note that Earl Geddes has written on the subject of audibility of higher-order harmonics, but mostly related to higher order harmonics in "waveguides" (i,e,, horns).

 

Chris

[Chris A]

dp...

[Chris A]

Quote

By the way, here is an interesting article written by Nelson Pass on the subject of harmonic distortion in amplifiers as a function of increasing feedback:

https://passlabs.com/articles/audio-distortion-and-feedback

 

I've converted this article to pdf format so that it is a little easier to read (enclosed).

 

Chris

 

Audio distortion and feedback - Nelson Pass.pdf

[Chris A]

An interesting excerpt from the above Nelson Pass article:

 

"An important thing about distortion[including harmonic distortion] – when you run a signal through a device which is even slightly non-linear, you have changed the signal forever. You can use various techniques to reduce distortion after the fact, but you can't go back. The problem is greatly compounded when complex signals consisting of many frequencies all travel through the gain device at the same time, or when a simple signal is passed through a number of nonlinear gain stages in series."

 

Harmonic distortion turns into AM distortion (...and FM distortion, as it turns out). The resulting output gets very noisy and opaque at SPLs well below actual concert levels. Mr. Pass has said a mouthful here. I recommend careful consideration of his entire article, embedded as a pdf file above.

 

"Many audiophiles believe that 2nd harmonic is to be preferred over 3rd harmonic. Certainly it is simpler in character, and it is well agreed that orders higher than third are more audible and less musical. However when given a choice between the sound of an amplifier whose characteristic is dominantly 2nd harmonic versus 3rd harmonic, a good percentage of listeners choose the 3rd.

 

I have built many examples of simple 2nd and 3rd harmonic "types" of amplifiers over the last 35 years. When I say "types" I mean that they used simple Class A circuits described as "single-ended" versus "push-pull" and so tended to have a 2nd harmonic versus 3rd harmonic in the character of their distortion, but were not made to deliberately distort. Anecdotally, it appears that preferences break out roughly into a third of customers liking 2nd harmonic types, a third liking 3rd harmonic, and the remainder liking neither or both. Customers have also been known to change their mind over a period of time.

 

However the issue is partially obscured by the fact that the 3rd-harmonic type amplifiers usually have lower total distortion. Third harmonic usually appears with a negative coefficient, resulting in what we think of as "compressive" - the example in figure 3. It's worth noting that odd orders on nonlinearity also can be seen altering the amplitude of the fundamental tone - something a distortion analyzer doesn't ordinarily display.

 

Audiophiles have been accused of using 2nd- or 3rd-harmonic distortion as tone controls to deliberately alter the sound. I suppose that there are people who like it that way, but I don't think this is generally the case. For reasons which will become clearer when we talk about inter-modulation distortion,high levels of any harmonic become problematic with musical material having multiple instruments, and the argument that 2nd or 3rd adds "musicality" doesn't quite hold up."

 

Chris

[garyrc]

Hi Chris,

How do you hold your Auralex Sonofiber tiles to the wall? Are your walls plaster, paster on sheet rock, or what? I thought of Velcro, but the sticky stuff isn't sufficient unless you distribute the pull over a very large surface area. I'll use glue if I have to, but would rather not. If you use glue, what kind?

Thanks for all of your help in the past, especially your posts in this thread.

Gary

[Chris A]

Gary,

 

My walls are of drywall construction--"sheet rock". I use push pins to hold up the tiles presently, but sagging of tiles is evident, so I'm also looking for solutions that don't mess up the walls. The adhesive that Sonex sells is truly awful stuff - sort of like urethane foam that doesn't foam up...or harden.

 

I suppose that mounting the tiles on thin wood/plywood backing and then hanging up using a picture hanger might work, but it would be subject to rattling when the subs are pushing out LFE - so I just keep re-pinning the tiles on the wall from time to time. I'm beginning to get dissatisfied, however.

 

Chris

[garyrc]

Thanks, Chris.

We were thinking of digging a few little holes in the absorbing material, and embedding a few screws with neoprene washers, then pulling the fuzzy stuff across the screw heads and neoprene. Screws do invade the wall, but would be easier to remove than horrible glue. There is a layer of 3/4 ply just behind our sheetrock, so the screws could reach into that. The trouble is, we would like to cover the fuzzy stuff with Soundsuede fabric so the fabric would have to be fitted around the absorber and tucked in behind it after the screws are in.. We will experiment with this with a small piece, and also with solutions involving elastic.

I agree with you about the Sonex glue. We had it in the audio room in our former house, and tried to peel / scrape / tear the Sonex off before selling. No way. The Realtor attempted to hide just how aghast she was. We ended up leaving the Sonex up, covering it with fabric, and calling the room a Music/Mixdown/Meditation room. We put an ad in Mix Magazine and at the nearby private college. In the write-up, we encouraged people to use a flashlight (provided) to peer through the fabric and see the anechoic wedges. It sold for more than we, or the Realtor, ever expected (before the decline).

Given all of the acoustical adventures, I'm thankful I've never had a WAF problem. She is very accepting, thinks the Klipschorns look marvelous, and is very creative in solving music room problems. She layed the floating floor in or old house's music room. Years before, she helped me construct broadband membrane absorbers (with absorbing surfaces as well) to hang in a large bathroom that we used as a voice booth. The bathroom we used was in my parent's house long after I lived there. More than a decade before that, the same bathroom was made dark, with an expanse of black cardboard with pinholes in it placed over the window to simulate a starry night sky for a Super 8 film I was making. So I guess I didn't have a MAF problem either.

[Chris A]

"Why Does the Use of Direct Radiator Monitor Loudspeakers in Mixing/Mastering Control Rooms Matter vis-a-vis Corner Horn Imaging and Timbre Balance?"

 

Probably because the typical direct-radiator speakers used in mixing and mastering control rooms partially mask the music they reproduce - just like very low power SET amplifiers with their transient-soft-clipping characteristics do when coupled to horn-loaded speakers, except that mixing and mastering engineers probably don't hear the resultant hard edge that horn-loaded speaker owners do later with their much more efficient speakers.

 

Also note that horn-loaded speakers also reproduce higher-order harmonics of multistage SS amplifiers more apparently, which tends to put a hard edge on those frequencies typically rolled off using underpowered tube-type amplifiers. For instance, recordings of jazz saxophones and double reeds, such as oboe, English horn, bassoon, tend to have a harder edge on horn-loaded speakers (due to their much lower modulation, harmonic, and driver compression distortion than direct-radiating speakers), driven by lower quality SS amplifiers of multistage design using moderate levels of feedback (which, according to the Nelson Pass article, produces the most higher-order harmonics at low amplifier output levels).

 

In the control room of the mixing or mastering engineer, these hard edge frequencies might otherwise be manually rolled off if the mixing or master engineer were using horn-loaded speakers with multistage SS amplifiers susceptible to higher-order harmonic distortion.

 

The solution to this second problem, of course, is to use a sufficiently powered single-stage amplifier with very low harmonic and modulation distortion, such as the aforementioned FET amplifiers by First Watt, or push-pull tube-type amplifiers of much higher quality (in terms of freedom from distortion) than any SET type amplifier.

[garyrc]

Probably because the typical direct-radiator speakers used in mixing and mastering control rooms partially mask the music they reproduce - just like very low power SET amplifiers with their transient-soft-clipping characteristics do when coupled to horn-loaded speakers, except that mixing and mastering engineers probably don't hear the resultant hard edge that horn-loaded speaker owners do later with their much more efficient speakers.

Were we better off with recording studios using the old studio monitors that at least had horn loaded midrange/highs? I think maybe we were. One of the very best studios in San Francisco used Altec 604Es to monitor (I think the enclosures were bass reflex, though). One of the best in Marin Co. used JBL drivers with custom wooden horns. I think the mixers were quite aware of any hardness, subtle microphone distortion (hints of very brief diaphram crashing?), traces of mixer distortion (like some of the old mic preamps that overloaded easily), and took steps. A good number of my newer CDs seem to reflect steps not taken.

[Chris A]

Quote

Were we better off with recording studios using the old studio monitors that at least had horn loaded midrange/highs? I think maybe we were.

 

That is what I was trying to convey.

 

I find that those mastering labs that still use horn-loaded midranges and tweeters, such as Doug Sax's older "The Mastering Lab", produce recordings that sound very good on horn-loaded speakers. His newer recordings in this studio are mixed on direct radiating speakers, and I can hear a difference in "listenability", I believe. I don't believe that is an accident.

 

Quote

A good number of my newer CDs seem to reflect steps not taken.

 

I also find that this is true.

 

I recently acquired a CD called "Beyond Words" by Oregon, and another called "Elements" by Ira Stein (I received them on the same day and played them) on advice found in a music-related thread over at DIYAudio. I loved both of these recordings, but both drove my wife out of the room.

 

The Oregon CD in particular was one of the cleanest recordings that I've ever played. But both recordings had a fair amount of conical bore instruments (sax, double reeds) apparently recorded without tweaking or rolling off their highs--that drove her out of the room. My first clue was that I also recently acquired a Stan Getz/Oscar Peterson CD originally recorded in 1958 that had all the saxophone highs rolled off, and she mentioned that she really liked that recording. It was much easier to listen to.

 

Chris

[garyrc]

I loved both of these recordings, but both drove my wife out of the room.

I wonder if women still have better high frequency hearing? I heard this difference was at least reduced by people starting to wear ear protection in high noise settings that used to be populated chiefly by men, fewer men smoking, but with a temporary increase in female smokers (late '60s, '70s).

But both recordings had a fair amount of conical bore instruments (sax, double reeds) apparently recorded without tweaking or rolling off their highs--that drove her out of the room. My first clue was that I also recently acquired a Stan Getz/Oscar Peterson CD originally recorded in 1958 that had all the saxophone highs rolled off, and she mentioned that she really liked that recording. It was much easier to listen to.

I love bright but low distortion recordings -- always have. I hope there is a good way to tweak without them artificially rolling off the highs, like backing off the microphones a bit??? I think we hear overload of some kind -- or extraneous resonance -- with "hard" recordings, but I'm not sure where it comes from. Whatever the cause, I think the mixers can hear problems more clearly through horn loaded speakers.

Is there a good way to quantify "veiling" or "masking," and the like? Is it measurably low pitch resolution, slightly inferior transient response, a frequency response dip, too much frequency modulation distortion, or what?

To me, horn speakers make some recordings sound much better, and others much worse, at least in the home. The odd thing is that I've never heard a "hard" recording in a studio with horns, and almost never in a movie theater, with their horns, going all the way back to early Todd-AO. But take a good movie recording and f*** it up in the DVD version, as they did with The Man Who Would Be King, and it can be hard as glass, a real knife in the ear.