John Curl on PWK, K-horns, phase and time distortion

[sunnysal]

I was over at AA and saw these comments mentioning PWK and klipschorns:

"In audio, I first saw this with 'time delay'. 30 years ago, the ear was considered 'phase deaf'. This was then known as: 'Ohm's law of acoustics'. Today, we align our speakers to within millimeters, and yes, the is an improvement. I was there when Richard Heyser stated that 2 feet of path difference might be audible, and Paul Klipsch formally protested. Now you have to understand, both Richard Heyser and I owned one K-horn each and we were professional friends with Paul Klipsch. We both liked and respected Paul Klipsch, but Richard was working on this time align thing, and he felt that we were ignoring something important. Later, in Sept 1975, Manfred Schroeder wrote a paper in the 'Proceedings of the IEEE', titled 'Models of Hearing'. This paper had an entire section on monaural phase sensitivity, and showed a number of tests that showed sensitivity of the ear to monaural phase, and contradicted the existing 'envelope hypothesis' of hearing. This one paper changed our total perception of how the ear works. Over the following years, much more time and effort has been given to 'phase aligning' loudspeakers systems, compared to earlier years, because our 'paradigm' had shifted. Later, both Richard Heyser and I both gave up our K-horns, because we both found the time distortion to be annoying, once we owned up to it, and it was almost impossible to phase align a K-horn, even though it had other positive qualities."

I asked a few question around AA about how one might approach fixing these defects in k-horns (via electronics from TacT or electronic delays on each driver, etc.) but have not gotten too far. any thoughts from the peanut gallery? regards, tony

[rigma]

I have done this using a Crown IQ USM-810, which is a active crossover, with time delay and parametric equalizer all programed via your computer. This is done by Tri- amping the K-Horns. Another advantage is direct coupling the woofer amp, without going through the passive crossover thereby taking advantage to my Crown amps high damping factor. Sounds fantastic.

[WMcD]

It is perenial question.

The tweeter is at the front, the midrange driver is about 20 inches behind, and the bass driver is several feet further back. It seems so wacko that there is not some gross problem.

Several years ago on the forum a fellow was asking me for suggested time delays to be used on the tweeter and mid to delay the signal so that everything would wind up at the front at the same time. I didn't see any report from him.

I'd love to have the time and equipment to investigate myself.

My recall is that PWK used a system of pipes to get things into "time alignment." There was some difference in sound. But using equalizers to correct for that change, things sounded the same.

I've not read the Manfred S. paper and should get a copy. It may pertain to the hotly debated issue of polarity, instead of phase. The polarity issue is an interesting one. Given the length (number of mechanisms) of the recording and play back chain, we really don't know if absolute polarity is maintained. More specifically, if the a pressure wave pushes on the microphone, is our speaker or headphone diaphragm pushing out, too, the same way. Some amplifier designs invert polarity. And there is a phase issue, which is more complicated.

There was some discussion about whether the ear can hear polarity inversion. This requires an asymetrical signal or musical tone. Evidently, if you look at the output of some musical instruments, might be a sax, there are sharp peaks going positive and gentler peaks on the negative. (Of course the should even out to zero, long term.)

So the question was, what can anyone's ear detect if polarity is inverted, with the sharp peaks go negative. Like hooking up the speaker with the negative lead to the positive and vice versa.

It was reported that on some waveforms, humans could hear the difference. Of course this is a situation where there are no frequency response issues, or time issues. How much can be inferred from this is anyone's guess. It did show that popular belief that polarity could not heard, at all, was not correct.

It is a far cry from the physical setting of the bass, mid, and treble, being displaced in time. And also a far cry from most music.

It is fair to point out that Heyster did a time arrival analysis of the K-Horn in his review for Audio magazine. He heard problems in the mid to treble range and concluded it was because of the offset of those drivers. There was nothing from him about the bass to mid "problem".

It is interesting to note that if the audable issue is mid to treble (per Heyster), then it is not too difficult to move the tweeter to the back of the system in a cabnet of its own. Further, the long awaited, always expected two-way Jubilee will not have this problem. (With apologies to Mr. T. Williams.)

I have wondered whether, in PWK's time, there was so much phase, polarity, and time distortion in the recording, storage, and playback chain, that K-Horn issues were far fewer. Therefore, if everything else gets to be "perfect" maybe there will be audable issues in the K-Horn to be corrected.

Gil

[j-malotky]

Tony

I am not an expert on this subject, but years ago I heard about this as well. I was told that time phase is only audible if you are within a few meters of the Khorn. Once the sound matures at a distance it is no longer a problem.

I have always had my Khorns in large rooms 22 to 25 feet wide and have not noticed this in my systems.

JM

[rigma]

Yea, William that was me, I set the delays and slope based on input from that thread and everything sounds great. I think alot has to do with driving the woofers with 500 watts of Crown power and 3' of 10 ga cable with no passive xover in the path, the bass is very tight and controlled.

[jdm56]

I used to fantasize about digitally delayed,infinite slope crossovers for tri-amped k-horns; the "holy grail" being a perfectly time and phase coherent wave launch from a full-range, three-way horn system. But I really wonder, even if it could be done, just how different would it sound? Would it even sound better? And at what price? Perhaps there is a good chance that all the extra dsp and hardware required to accomplish such a thing might do more sonic harm than good.

The fact that Heyser and Curl had to to replace their klipschorns after focusing on this supposed problem, may say more about the human psyche than it does about the audibility of the problem itself. Perhaps they were, as engineers, trying to satisfy their intelect more than their music loving heart that probably led to the k-horn purchase in the first place.

[rigma]

You may be right..LOL.. I am an engineer and love to tinker! I at least think it sounds good, but who knows, at least I am having fun. I have an RTA on order so I can do some serious testing.

[jdm56]

An RTA?! Oh, now you're really gonna be miserable!

The bass in my room sounds great to my ears most of the time (I confess to sometimes juicing the bass k-nob rightward a bit.), but my trusty yet humble ol' RatShaq spl meter shows me large dips at app. 50 and 100 hz that I have only been able to minimize - not remove. And so it haunts me. I suspect the cure will only come in the form of a couple of SVS subs.

[sunnysal]

I had supposed that digital time delay was the answer to this issue, but I will probably never know since triamping and a digital crossover are beyond my reach, both economically and mentally/technically. three great stereo amps, good speaker cables for all that, the digital speaker control...grrrr...guess I will play with SET for a while before moving into the 20th century, lol! tony

[rigma]

I really bought the Crown IQ processor for the Crossover feature so I could Tri-Amp my Klipschorns. The parametric EQ and time delay was part of this so I figured why not use it. The system sounds great, but I am not sure if any of that is due to time aligning or not. It sure did not seem to degrade the sound in any way.

[sunnysal]

I've got this bipolar personality...half of me wants to plug a TacT box into a Bel Canto digital amp and some Revel Studios or B&W 801s and see what happens, the other wants a Marantz 7C into a 300b SET into my k-horns...until I win the lottery I have to take the shorter path....SET here I come! LOL tony

[Klewless]

Everybody knows,

1. What you build yourself sounds much better than anything anybody else can make.

2. What you bought yesterday is crap compared to what you are going to buy tomorrow.

3. You now wish you had kept what you just got rid of yesterday.

4. Stereo is visual.

5. etc

Conclusion?

You can convince yourself of anything you wish.

He, he

[Istari]

I'm sure that Paul flashed his yellow BS button on this one!!!

[j-malotky]

With The sound comming out of your speakers at 750 mph (speed of sound at sea levil, 68F normal air pressure) does 22" or whatever distance between drivers really make much difference.

I don't think so. I agree with HOBBIT. Pauls yellow button sums it up.

JM

[TBrennan]

JM---Yes, it can make an audible difference. I refer you to John Hilliard's (the Zeus of The Great Horn Gods) study of the problem after the famous Eleanor Powell "double-tap" tap-dancing monitoring incident at MGM in the early 1930s.

This led to Hilliard's development of the Shearer Horn, which had much less mismatch than the WE systems, and eventually to Hilliard's development of the VOT. Oh yes, it can make a difference. However there are worse things a speaker can do, PWK chose his tradeoffs and decided he could live with the time mismatch problem.

[HDBRbuilder]

Well...here I am yet again trying to interpose myself as knowing something about this...but in reality, I am just gonna give a bit of food for thought re: time delay and phasing!!

we are sitting here in this thread trying to come up with an optimal phasing/time delay notion...but the truth of the matter is that it, as PWK would likely have said "doesn;t make a dime's worth of difference". Why do I say this? Simply because the music itself as it is recorded is NOT in phase or in time to begin with!! Oh yes...now you are all thinking..."how the hell did this idiot come to this conclusion?" Well...here goes!

Example: A trumpet player, a flute player, and a tuba player all read and perform their respective notes at EXACTLY the same time, yet the SOUND from the flute is the FIRST to exit the instrument(short wind path), the sound from the trumpet is next to exit the instrument (medium wind path), and the sound from the tuba is last to exit the instrument(long wind path)...and yet, even it all of these instruments are used in the same recording, and miced at the same exact distance from the instruments themselves, the phase will NOT be the same...neither will the time delay!! The same can be said for the recording of a miced acoustical guitar, a string bass, etc...so...just what damned difference does it all REALLY MAKE, after all? IMHO...NOT MUCH!! Since music is inherently time-delayed and NOT in-phase from its inception, then the little bit involved in horn-loading "doesn't make a dime's worth of difference" either way!! Think about it!!

[TBrennan]

Builder--- You forget that a single instrument will have frequencies reproduced by both bass and treble drivers; this will cause a time delay within the sound of a reproduced instrument: you seem to be talking about delays from one live instrument to another, an entirely different thing.

In the "double-tap" incident at MGM the sound of tapdance clicks was reproduced with an echo; two taps where there should have been one. Hilliard traced this to the path length difference to the listener between the drivers of the bass and treble horns used in the 2-way WE monitors. When the treble horn driver was moved back into the same plane as the basshorn driver the double-tap went away. This is indisputable.

Hilliard experimented and found that the effect was frequency dependent and that with a 500hz crossover a delay of 3ms (about 3 feet) was tolerable. Thus PWK seemed to be skirting right on the edge of what Hilliard found acceptable. But Hilliard was never at ease with the idea of ANY time delay, thus his abandonment of his Shearer Horn and his development of the VOT. The time delay thing is most certainly a flaw in the Khorn. But it has strengths that are more important.

[HDBRbuilder]

Tom...no...I was just drawing an analogy...that since the time delays from each musical instrument in a recording of music are inherently out of whack...then when you combine that with the distance factor to where the microphones are when recording this ALREADY delayed response of the instruments COMBINED with other resonance characteristics involved...such as studio/concert hall dynamics, etc...then even though the drivers reproducing these recordings may be out of time-phase somewhat, it is highly likely that this same out of phase-ness MAY actually CORRECT some of the natural out of phase tendency of the original music produced ...instead of AMPLIFYING the DIFFERENCES...even the slower moving bottom end soundwaves in music are more out of phase the farther you are from the source when recording them!! Sure it is true that two wrongs don't NECESSARILY make a right...but in some cases it may work in reverse!! Just think about it for a few minutes!

In any case...in the instance of the k-horn...the midrange DRIVER is only about 18" to the rear of the tweeter driver...and less than half that distance from the bass bin driver, anyway!...when looking at the drivers as seen on a vertical plane.

[rigma]

I hear what you are saying HDBRbuilder but my thoughts have always been the fact that two drivers at or near crossover are producing the same note but with a time delay just as TBrennan described.

[HDBRbuilder]

I agree with the time delay in relation to two (or more) different drivers presenting different portions of a particular musical instrument's or voice's sounds...but that also occurs naturally in listening environments through reflections, partial absorptions, and eiger waves...all depending upon both the vertically planar equality of the drivers, the listening environment, the position of the ears in relation to the directivity of the separate drivers, the distance FROM the drivers to the listening position, and the source material itself. There are many other factors besides just the planar equality of the drivers that affect this, as you can see. Even the time delay itself can be entered in the source during the recording of the material if the recording is done in other than a monaural original mix!!...IOW...for example the tap dancer...is to the left in a stereo mix...but the right mic records the tap with a time delay...because of the distance from the tap source...get my point? The same can occur from a multiple channel recording before mix-down...A low microphone on the left would pick up the tap at a different time than the HIGH mounted microphone on the left!...BUT in its ENTIRETY of sound...thereby providing an echo effect from the source material itself once it is mixed down to just two channels.

Either way...time delay is involved in MORE than just planar differences in equality of drivers in speakers. That is MY point. When you add in recorded reverberations and such...it makes the issue even more complicated...even in todays home theater systems this use of time delay and signal processing is exactly what gives you differences in your "cathedral", "concert hall", etc. configurations available in your H/T systems for EFFECTS...see what I mean?

If we all lived in just a monaural world listening solely to monaural original recordings, then this issue would have a much more profound meaning than it tends to have when presenting anything BUT monaural recordings through a single speaker system firing from only one point at us!