Is there any timedelay compensation in the Khorn between the bass re the mid re the tweeter

[jlossint]

Is there any time-delay circuit delaying the signal from the K-77 relative to the signals from the K-55/K-400 horn. Also is there any delay circuitry introduced compensating the signals between the K-55/K-400 horn and the bass horn.

It is likely to affect any lobing and therefore the image.

Any information about this topic?

[djk]

Only possible with digital delay and three amplifiers.

[sunnysal]

richard heyser´s article and testing for audio magazine in the 80´s showed significant (though everyone seems to differ on what constitutes "significant") time delay between woofer, sqwuaker and tweeter on the k-horns. delay is up to 8.4ms max between the woofer and the tweeter, some feel this is too much. the only way to remove these time differences in via an electronic crossover that adds delay to the squaker and tweeter. regards, tony

[Klewless]

Relatively speaking

The midrange and bass horn are in phase at crossover (400 Hz) and the mid horn leads by one cycle.

The midhorn to tweeter represents the largest delay because the tweeter is several wavelengths ahead. Often they are brought in phase by reversing the tweeter polarity.

Paul Klipsch has demonstrated many times to skeptics that the time disparities are not significant. At least to him.

I knew a fellow who moved his tweeter out of the box and set it on top of the khorn. Personally I couldn't tell the difference it made.

It would be a very expensive venture to

A. Resolve the issue with amps and delay units

B. Redesign the Khorn

Also I don't think that bringing the drivers into time alignment would resolve the lobing effects. These are created by the fact that more than one driver is reproducing the same frequencies. Ie, the areas around the crossover points.

[boom3]

PWK addressed this question many years ago. He always felt that the delay was insignificant and the electronic delay devices then available were were too noisy and distorted to be useful in any event. It would be interesting to see an experiment with current technology.

Time alignment (a registered phrase, I seem to recall) is one of those elusive phantoms of audio. My experience has been that when all other factors are accounted for in a balanced design, then temporal alignment should be included if it can be included without compromise of other, more important criteria. There were many designs of the 70s that started with temporal aligmnent as the controlling critera and worked backwards to other criteria. The result was a flock of odd-looking and unbalanced designs. How much temporal shift between drivers can be tolerated is very controversial. The usual criteria is 2 milliseconds, although that is highly subject to frequency, listener position, and subject matter. I own a pair of Dahlquist DQ-10s which were some of the first speakers to be designed for temporal alignment in the US. I have a more comprehensive review on my website, but in sum, I think the DQ-10's performance stems from factors other than temporal alignment. In any event, they can't compare with my CW IIs, which were designed to a different criteria.

[colterphoto1]

I understand time delays and phase differences, but help me here. What are 'lobing' differences?

It seems from the discussion that if the Bass and Mid horn are fairly close timing wise, that simply moving the tweeter on top of the box so it aligns with mid driver (so long as you reduce diffraction effects with top of cabinet) would be a pretty inexpensive, simple fix.

Michael

[DRBILL]

How does the symphony compensate for the tuba in the back row and the violin in the front row?

They do nothing. The ear is very forgiving. "Try listening to the music instead of listening to the speakers". That was from the mouth of PWK!

DR BILL

[Deang]

"I think the DQ-10's performance stems from factors other than temporal alignment."

Absolutely. We played with the baffle alignments in the late 70's, and no matter how we did it -- nobody could tell a difference. The drivers are lined up via the voicecoils, and we shifted them so the faces of the drivers were evened up -- no difference at all that we could tell.

My Klipschorns sound very good to me. If alignment was half the issue that some claim it to be -- I don't think they would sound so good.

[D-MAN]

Scientific testing dating from the 1930's (the famous tap-dancing scenario) resolved that the time delay as indicated by distance between the bass horn and the upper frequency horn is that less than 3 feet of distance between the drivers cannot be determined by the human ear.

Distances of over 3 ft. result in a discernable difference.

PWK has decided that the time lag in a Khorn between the bass diaphragm and the midrange diaphragm as being undiscernable, but technically it exists, although I agree with him that it is not bothersome.

The overall length of the Khorn bass channel is 64 inches (measuring center-of-channel). The length of the mid horn is (what - about 1-1/2ft?) 64 - 18 is approx. 46 in.

That is 10 inches over 3 ft, so there is a potentially discernable time lag. But 10 Inches is not much as sound travels that distance in a micro second or two.

How accurate is your hearing? I have never heard a time corrected set of horns, so I really cannot say.

I would ESTIMATE that it is not a problem worth alot of money to try to "fix".

DM

[sfogg]

"Distances of over 3 ft. result in a discernable difference."

If the earlier posted info is correct about a 8.4ms delay then the K'Horn is well over this distance. 1 foot is roughly 1 ms of delay.

" PWK has decided that the time lag in a Khorn between the bass diaphragm and the midrange diaphragm as being undiscernable, but technically it exists,"

I don't think it is discernable as a time lag directly (at least it isn't on my Las), however it is audible through the interference between the drivers blending region on my Las. Cross the bass and mid horn over much steeper with something like Al's ES 600s or Deqx digital crossovers and the vocal range gets much coherent from the reduced interaction between the two. Has any of the single 'full range' driver guys ever heard a system with very steep crossovers? I wonder how they would feel they compare?

Shawn

[D-MAN]

That is because the overall horn length to the diaphragm of a LS/Belle bass bin is 30 inches total length minus the length of the mid horn to the driver diaphragm. Not over 3 ft in difference, therefore, we can't discern it.

The distances between the bass diaphragm and the tweeter are not as questionable because there is no sonic correlation between low bass notes and high (tweeter-level)frequencies that you would recognise for judgement, although it could be artificially produced. Put another way, that mixture would always resolve to 2 separate tones, and a timing issue between them would not be apparent to us. It's the tones that blend together or that occur at the cusp of the crossover point(s) that we are sensitive to in the sense of determining timing issues.

Low notes (i.e. bass notes) are usually not a concern with timing issues anyway. It's the higher frequencies, such as with the famous tap-dancing where each tap was being produced on the tweeter and midrange horns, with 8 feet of separation between the respective diaphragms causing the audience to question the accuracy of what they saw on the movie screen with what they were hearing. Traditionally, the bass horn and overall length thereof is not considered to be quite as critical to overall delay timing issues as are the upper frequency horns.

DM

[Parrot]

----------------

On 12/27/2004 12:49:35 PM DRBILL wrote:

How does the symphony compensate for the tuba in the back row and the violin in the front row?

They do nothing. The ear is very forgiving. "Try listening to the music instead of listening to the speakers". That was from the mouth of PWK!

----------------

Great post, Father Bill.

[3dzapper]

Every capacitor and inductor in the x-over affects the timing and phasing of the signal. Nobody outside Klipsch knows what all those caps and inductors are for in the AK-4 network but I would bet that this subject has been addressed to a degree. They definately aren't necessary for splitting the signal into three.

All TVs have a passive delay circuit in the audio section to compensate for the longer signal path of the video signal. Because there are both visual as well as audio components to TV this is necessary to eliminate the "karate movie syndrome".

Rick

[Deang]

I think that's what the thick grillcloth is for. I'm sure it slows the highs down a little.

[D-MAN]

----------------

On 12/27/2004 7:11:08 PM DeanG wrote:

I think that's what the thick grillcloth is for. I'm sure it slows the highs down a little.

----------------

Ok, put Dean on the Naughty list for next year!

DM

[Al Klappenberger]

Guys,

Time delay and "group delay" are pet peeves of mine! You got me going!

Compensating for time delay between drivers is difficult. It has to be done

using electronic equipment employing digital techniques with a separate amp

to each driver. There is no passive circuit that will proved enough flat

delay to equalize the propagation time difference between driver in a Khorn

that will fit in the average living room! Doing it with passive components,

that is lumped-component delay lines, is impossible.

The fact is that it is a red herring anyhow! Time delay error between drivers

is equivalent to phase differences between them. If you consider the varying

wavelength of all the frequencies in the tweeters range, which can be less

than one inch, the propagation delay represents many wavelengths as the

squawker driver is at least a foot or two behind the tweeter. This means the

phase relation between tweeter and squawker is virtually random.

I did an experiment a while ago which employed three different sine wave

signal generators each tuned to represent the components of a square wave.

Each was adjusted as closely as possible to the fundamental, 2nd and 5th

harmonic of some frequency. 1000 Hz maybe. I don't remember exactly. The

phase relationship between these was constantly varying from what it would be

in a true square wave where the phase relationships would be "locked". I

could see the waveform squirming around on a oscilloscope but was unable to

hear any change in the tone quality of the combination whatsoever when

listening to it through headphones even while looking at the oscilloscope for

visual clues. The bottom line is that you CAN NOT HEAR phase relationship

errors between the components of a waveform! Your ear is totally deaf to it!

Since the three drivers in a three-way speaker are reproducing different

frequency ranges of the complex waveform called music with delay (i.e. phase)

errors between them you can conclude that you can NOT HEAR the delay

difference between the drivers casue by the different propagation length

between them.

HOWEVER--> When two sources are generating the same sound there is an

interaction between them if they are not aligned in time. This is where time

alignment comes into it. The different path delays will represent different

wavelengths at every frequency and will add up and null out at different

places in the room. Not only will the phase error sums and difference

positions in the room change with frequency but will also be different at

every place at any given frequency because of different path lengths between

the drivers at different locations! You CAN hear this! So, unless the two

drivers occupy exactly the same place at the same time there will be uneven

distribution around the room. This is the argument for a two-way system. Of

course that means a single horn and driver with uniform frequency response

and distribution form 400 Hz to at least 15 KHz. That isn't easy to do!

My solution--> Here's where my extreme-slope networks come into it. If each

driver only makes a sound within it's own range and there is no overlap in

frequency between the drivers there will be no interaction. It's true that

sharp filters generate phase errors, but like I proved with the three random

phase signal generators, you can't hear it! Plus the delay generated by even

the sharpest filters is minuscule compared to the total phase shift caused by

the propagation delay between the speaker and your ear and between each

driver and your ear. When the interaction is removed the stereo image comes

into much better focus and the "sweet spot" virtually disappears!

There is one situation where phase is critical though. That is the

relationship between the phases of a single sound between two sources. That

is, your left and right speaker! This is part of how your brain computes the

stereo image. This is why crossover networks should be matched left and

right. Consider too when several drivers on each speaker are generating the

same sound. Your ear / brain combination gets all confused as to where the

sound is coming from!

Ok.. I will now leave the soap box, stage left!

Al K.

[DrWho]

----------------

On 12/27/2004 12:49:35 PM DRBILL wrote:

How does the symphony compensate for the tuba in the back row and the violin in the front row?

----------------

After being in various orchestras for the last few years, I must point out that the tuba in the back row does indeed play well ahead of the beat. I don't want to get too much into the topic as it's a mute issue because the point being made is the same...just that the analogy would be disregarded by someone of a different viewpoint.

[DrWho]

In regards to the timing issue, in the old warehouse at work we had an audio system composing of some JBL Control 29's scattered around all the workstations with just one single Klipsch subwoofer (15" downfiring with an 18" passive i believe) to fill in the rest of the warehouse. Well as we moved everything to the new warehouse, the last of the sound system was the last to leave and we ended up with two speakers on a ledge over in the corner and then the sub which we put next to the computer/amp rack in the middle of the warehouse. It was extremely annoying hearing the bass of the kickdrum hit and then to hear the snap of the kickdrum a few ms later. Now I know these distances were extremely large (The HF and MF sections were about 30 feet away...yes, those C29's could get freaken loud and cover the warehouse). I believe though that this same issue would happen with the khorns...

With the khorn, you would hear the slap of the kick before hearing the boom of the kick. I am not sure how much later the boom comes and with the natural reverb of the room and the nature of the sound, I doubt you'd really be able to notice the difference. I am sure however that there are certain sounds where this will definetly be audible, but until one has heard otherwise, then there is no way to distinguish it. I can see this being an issue with extremely fast twangy bass lines...you'll hear the twang before the oompf. I have no idea how dramatic the effects of this are with the khorn and thus the true audibility of it. It seems there were measurements done already showing an 8ms delay which I believe would be very audible...even moreso for sounds shorter than 8ms (like a drum beat).

But in the grand scheme of things it doesn't seem to be too large a compromise for the other benefits gained by the design.

[Ray Garrison]

Dr. Bill,

That's sort of the whole point. WHen you're at a concert, you can tell that the violins, violas and cellos are in front of the tuba - it sounds like they are. There is a depth to the orchesta (more or less, given the arrangement of the musicians, the hall characteristics, and where you're sitting.) The idea of home reproduction is to replicate the sense that the players are laid out in space. AMoung the many different parameters that contribute to effectively recreating that sense of depth, time alignment has a role to play. Personally, I find that room treatments, proper positioning of the speakers, and an accurate amplitude response are of more significant importance, but if those are taken care of, a speaker that is time aligned (say, a Thiel or a single driver like a Lowther in a non-horn loaded cabinet) does a more credible job of positioning players in space.

[jdm56]

I'm wondering as I read Al's post how much, if any of our ability to hear phase or time differences is in any way dependent on the orientation (vertical or horizontal) of the two sources? For instance, we can all hear phase differences between two speakers - speakers are always set-up L & R. But drivers are almost always arrayed vertically and we apparently can't hear phase differences between them. Of course, 99.9% of the time the drivers will never reproduce the same frequency at the same level, but I'm just speaking hypothetically here. What if you put one speaker on the floor, and one on a tall stand directly over it. Would we still be able to tell when they were in phase and when they weren't?

...yeah, i need to get out more, i know.