Time Alignmnent in Cornwalls

[Clipped and Shorn]

I notice that visually the exact plane of the tweeter driver, horn driver, and woofer driver in my Cornwalls are not the same. Normally these differences would produce a time delay for the drivers that are further back from the listener. I assume Klipsch dealt with this somehow in the crossover design.

Does anyone know for a fact that the stock Klipsch crossover for the Cornwall did in fact compensate somehow for this apparent time alignment discrepancy between drivers?

If this was addressed in the passive Klipsch crossover design, how exactly is that achieved? I am aware that it can now be achieved in some active crossovers, but I am talking passive crossovers here.

Generally speaking, is there a way, in a passive crossover situation, to compensate for say a time delay which might exist if a tweeter driver is one foot or so in front of the horn driver ?

How complicated is this? Is it feasible to create some simple crossover elements which would adjust for a tweeter driver being so many inches ahead of a horn driver? Anyone here know how to do this, or is there a source for existing components that do this sort of thing?

If Klipsch did this in the Cornwall, how did they implement this?

C&S

[TBrennan]

Clipped---PWK didn't give a damn about driver time delay.

In the 70s a fella named Ed Long designed a passive crossover for the Altec 604 coax driver that compensated for driver delay, in fact Long trademarked the term Time Alingment. This crossver was used in the famous Urei 813 studio monitor.

Tom Danley of ServoDrive has designed pasive crossovers for his Unity horns that compensate for time delay, I've heard them and thay make a difference. Tom explained how he designed these but it went way over my head; suffice to say that without computers and his very sophisticated testing equipment he'd find the task near impossible.

[Clipped and Shorn]

Interesting. So it is not just putting a capacitor in there somewhere, hah hah.

Some active crossovers have a time delay adjustment, but it is usually on the lower end, not for the tweeter.

You usually see tweeters up front with everything else which means they are usually quite a bit forward of the horn driver. Intersting that PWK didn't worry about it. Maybe I won't either.

Would there also be a time alignment discrepency with the Klipschorns? How far back does the folded horn design make that woofer?

Does that custom Altec Box of yours does have the 515b back at the same plane as the Altec horn driver? Is that time alignment thought to be a known positive effect of that box design?

C&S

[Mike82]

While I've had both time-aligned and point-source speakers in my shop, I'd have to say both designs seem to make a difference, but I can't say whether the t-a or p-s design, by itself, is actually making the difference.

The crossover, or lack of in the case of the Lamhorn, probably contributes as much as anything. Some speaker designers will invert the phase of the tweeters to make the soundstage seem larger than it is. What really brought that home was the SP Test Disc where they invert phase. With most speakers, you could kinda of hear the difference, but with the crossover-less Lamhorn there was a huge hole in the middle of the speakers and you heard the bark about 3' to the outside each speaker!!

There are a lot more important factors in speaker design than time alignment and whether these incarnations make music more enjoyable is debatable.

I really like my non-t-a Khorns!

Mike

[sunnysal]

I was troubled in the past of mention to the "terrible" time and pahse problems with k-horns by people like John Curl, etc. But, as they say, ignorance is bliss. Since I have never heard (and probably never will) a time corrected k-horn I will never hear what I am missing...lucky me! I love the sound of my k-horns but I would sure love to see a design for a crossover that fixed those delay and phase issues. regards, tony

[Colin]

I have heard time aligned drivers, in backwards leaning cabinets, and the cohesion they have seems to remove attention from the loudspeaker, making them seem less colored or obvious. There is nothing subtle about the sound of big old horns though. Their sound is there. Perhaps this actually adds to the ¡§live¡¨ or ¡§real¡¨ illusion.

Maybe Tom Brennan, who never met a horn he couldn¡¦t take apart and test in some other configuration, could advise you how to physically time align the Cornwall drivers. I think you would have to:

h Remove the tweeter

h Mount it on top of the cabinet

h Move it back so the diaphragm aligns with the mid-range horn diaphragm along a vertical plane

h Tilt the cabinet forward so the tweeter, mid-range horn and woofer diaphragms are aligned along a vertical plane (15 degrees?)

h Raise the cabinet in the air so that the mid-range horn and tweeter aim directly at the listener¡¦s ears

h Reset the subwoofer volume and EQ since you now have a loss of mid and upper bass response

h Tilt your head back slightly so the holographic stereo image is focused at head level and not above or appears too tall

[sunnysal]

seems to me the only way to time align any of these speakers would be triamping and digital cross-overs+delay, those I suppose those TacT boxes get part way there by adjusting delay based on pulse response. tony

[j-malotky]

I still think you guys are nuts about the time allignment, but thats MHO.

Adding a delay into my line scares the crap out of me. After years of playing, I find the best sound comes from the "simplest" designs.

Anyhow, In my mind, I would think the best way to time allign is to remove horn drivers from you Cornwalls and mount them over the speaker with the compression drivers all in a line with horns sticking into the room.

JM

[rigma]

You can use a Crown USM-810, it has time delay, crossover, parametric EQ, mixing etc.. Top quality device that you can program from your PC.

[Al Klappenberger]

Hi Guys,

Time alignment is a lot of B-S! You can't hear it unless it is WAY out of alignment. By theat I mean by many FEET! The fact is you CAN'T Time align two drivers except at one point in space. The bigest bad thing it causes is peaks and dips all over the room as the wave from two drivers playing the same sound align in and out of phase. I am using extreme slope networks (120 dB / octave) to see that only a single driver is playing a single frequency. The distribution is as smooth all over the room with this scheme as I have ever heard.

In short, forget about the alignemt! Also, don't confuse "time" dealy with "group" dealy. They are NOT the same thing and one can NOT correct for the other!

Al K.

[TBrennan]

Clipped----Yes, with the VOTs you can align the compression driver with the bass driver. This ability was the main reason John Hilliard designed the VOTs; he was unhappy with the time delay with the Shearer Horns he'd previously developed. And the main reason behind the Shearer Horn was to improve the large time delay inherant in the Western Electric systems with "snail" horns and Bostwick tweeters. Though Hilliard experimented and found that 3 ms of delay was inaudible with a 500hz crossover any delay at all evidently rankled him, thus the development of the VOT.

The crossover also introduces phase problems and Tom Danley of ServoDrive has been working on this. In comparisons between Unity Horns with phase-corrected and non-corrected crossovers both Kurt Chang and I clearly heard an improvement with the corrected crossovers. Not a night and day difference; both systems sounded very good. But an improvement nonetheless. Look up Kurt's comments on AA.

Danley is a very astute professional designer of prosound loudspeaker systems and frankly, he's the greatest mind working in loudspeakers today. With his ServoDrive and Cyclone subs, and his "saloon-door" planar speakers he's developed some truely novel ways of moving air, the first novel means I can think of since the 1930s. His Unity Horns give coherant response from 100 to 20,000 hz with controled directivity all from a single, compact horn flare. His understanding of horns, drivers, crossovers and combined systems is so profound that for me talking to him is like talking to a Martian. :-)

The point being that when guys like Hilliard and Danley think about such things they bear thinking about. But there are more important things in speaker design, basic things like low distortion and good response. This time thing is icing on the cake. I'd rather listen to a Khorn, time delay and all, than to a single driver Fostex with no time delay but tons of distortion and poor dynamics.

[Clipped and Shorn]

Tom and Al, thanks for the interesting background and general information.

Mostly I need to move my bass reflex cabinets into the house, load them with the components, and fire them up. Until then, all of this is academic.

by the way I discovered an interesting company making high quality active crossovers. Less expensive than Bryston, who, btw, is now making a 3 way stereo for mucho $$$, the Bryston 10B LR23. Marchand sells components for DIYers. You can have your Lynx and still keep your Witz.

Marchand Electronics

c7s

[sunnysal]

the marchand tube 3-way unit (which I guess is the best for a k-horn) is $2,595 (the XM126-3AA)...a little rich for my pocket book...tony

BTW C&S what is that speaker in your avatar photo?

[Clipped and Shorn]

Vitavox

now I cannot find the link....

C&S

[Mike82]

C&S:

It's tough to tell from th epuny photo, but that speaker in your avatar is the Living Voice Air Partner (or Air Scout). About a $25K speaker.

[Clipped and Shorn]

Here is the picture of this beauty :

mmmmgood, I want some of these!

and design a new house to go with them......

C&S

[mace]

So how do orchestras account for time delay? If the guys in the back are playing in time with the conductor they'll be coming in late for the guy sitting in the front row? What about a rock concert with the huge array of speakers?

I just don't understand this time delay stuff. Speed of sound is 343.6 m/s, the horizontal driver distance from tweeter to squaker in the Cornwalls is about 0.3 m. So, the time delay is 0.3/343.6 = 0.87 milliseconds = 0.00087 seconds. Is that very perceptible? It sure seems small. In your average room the reflections from side and rear walls would seem more irritating than this 0.87 millisecond delay.

Mace

[djk]

The funny guy waving the stick keeps the people together, light goes faster than sound.

If you are used to playing on a small stage and then move to a large one your timing gets all shot to h3ll.

The small amount of time delay in a Cornwall on rim-shots and similar transients is plainly audible.

Not much can be done with the tweeter.

The mid delay may be fixed with an all-pass delay on the woofer (two caps and two inductors).

[WMcD]

This topic is a bag of snakes. Let me set out some general issues.

1) A very general test signal is a very narrow pulse. In the world of actual sound, it could be the pulse from a tap dancer's shoe, or a rim shot.

2) People quite correctly conclude that when it is split up into three signals in a cross over, the circuits often do not have an output which sum to recreate that pulse.

3) Also, if the pulse is fed to three drivers without filters, any physical offset of the voice coils will not allow perfect summing. However, speakers have their own internal reactances which can act like filters, altering phase. So the position of the voice coil is not that certain a marker.

4) There are some gross differences in the lenght of the paths in the K-Horn, LS, and Belle. Greater than 15 inches in any comparison.

5) Looking at the Cornwall, the voice coils of the drivers are not that much offset relative from each other in comparison to the front of the baffle. The midrange is behind the woofer and tweeter in front of it. OTOH, in a direct radiator system the offsets put the mid behind the tweeter and the woofer still farther behind. In each case, it is a matter of a few inches.

6) Classic crossover design looks at the phase and level relations in the passband where the woofer and mid, and then mid and tweeter are both sounding. It might be very narrow. The crossover design really does not try to delay things outside the cross over passband to accomodate physical offset.

6.1) It is certainly true that physical offset will affect phase, and therefore how sine waves sum, in the band where both speakers (e.g. a woofer and mid) are working at a common frequency. The resulting summed amplitute is affected, and can be heard.

7) How much can be heard is a matter of various studies. PWK maintained that amplitude (level) anomolies, arising from imperfect summing in the passband of the cross over can be heard. No one disagress with that. However, delay, unless it gets to a perceived "echo" is harmless. Harmless meaning that our ear brain combination don't detect it. When the echo effect can be heard is a matter of debate. It doesn't help that measured studies put it at greater than 2 or 3 mS (2 or 3 feet) and informal listeners say they can hear the effect of 2 or 3 inches with great certainty.

7.1) This whole delay issue does not create new frequencies. It is a form of waveform distortion in real time, but only that. He was concerned with distortions which cause spurious frequencies which our ear-brain finds annoying. There is reason to believe that horns do a good job of reproduction without creating these. Also, SET amps without feedback seem to be devoid of such annoying freqency products.

8) How much "time or echo" distortion goes on in the recording studio and the playback studio is a good issue. One has to wonder whether the relatively smaller time distortion in the speaker system comes into play, in the big sceme of things.

9) This is unrelate, but I throw it out. I read that theater film presentations have a built in time advance (at all frequencies). It was easy to do in the print because the sound pick up off the film is not exactly at the shutter of the projector where light goes though the film, to the lense, and off to the screen. Same issue though with advanced systems.

This accomodates the fact that the average listener is about 30 feet from the speaker behind the screen. There is, thus, about a 30 mS delay before the sound hits the popcorn. So when the lips on the screen announce, "Frankly Charolette" . . or "I'll be back" the sound better have been advanced 30 mS.

Question. Have you ever seen a transfer when this was not corrected?

Gil

7)

[TBrennan]

Mace---You don't understand, we're not talking time delay from instrument to instrument in the production of music; we're talking time delay within the reproduction of individual instruments, not say from snaredrum to violin but within the snaredrum and violins themselves.

I'll go back to the original "double click" incident at MGM in the early 1930s. A recording of Eleanor Powell tapdancing was being monitored on a Western Electric 2-way horn system with a 12 foot path difference between the drivers of the basshorns and tweeters. Instead of single taps double taps were heard, echos. This was traced to the path difference between the the drivers. First the tap was heard from the tweeter and 12ms later the tap was heard from the basshorn which was 12 feet farther away. This effect has nothing to do with the spacing between instruments on a stage or in a club.