Khorns Time Alignment

[ClaudeJ1]

Choosing crossover points should involve a little logic. And acoustic crossovers are far more important than electrical crossovers. My method is to let the acoustic components tell me where they liked to be crossed over. That usually involves quite a bit of data.

That is the approach I took with the K402/1133. It was smooth down to 300 Hz. and up to 5Khz., which dictated the xover points. Amazing sound and would be a great addition to a Khorn with time alignment on top of it's great natural response.

Edited September 27, 2014 by ClaudeJ1

[Chris A]

All this was made possible with an active digital crossover (an EV Dx38), REW software running on my laptop, and calibrated microphone/mixer.

Quoted text just to reference my posting above...This experience using an active digital crossover (EV Dx38) on the center channel was very revealing:

1. Loudspeakers using passive crossovers are crossed at certain frequencies and have certain resulting driver gains usually adjusted downwards for the tweeter and midrange relative to the woofer channels (using either resistors or increased input reactance at driver frequencies by the passive crossover network), but this doesn't mean that the passive crossover frequencies chosen actually sound the best. Once an active crossover is used, experimentation with crossover points and types (i.e., higher-order Linkwitz-Riley crossover filters, etc.), EQ to smooth FR, and most importantly delay to correct for time misalignment of drivers.  Once you have a digital active crossover, many combinations can be tried and heard effortlessly. 

   I found that the settings used for a passive crossover weren't anywhere near optimal for sound quality in the Belle and Cornwall and I used different crossover frequencies, added real time delay correction and played with relative driver channel gains, EQing each channel to be flatter.

    

2. Having the ability to change crossover frequencies and filter type, driver relative gain per "way", EQ filters, and lastly--delay settings, quickly and without effort to hear the differences gives the experimenter a feel for what's important and what's not so important.  This is probably the most eye-opening experience that I've had with loudspeaker design and implementation, a very enjoyable one at that.

   

3. The process of changing settings and listening for the effects of those changes dramatically helps to develop your hearing for certain loudspeaker issues and out-of-balance conditions--even without using a digital active unit, you can develop an ear for hearing loudspeaker issues with passive crossovers and analog active crossovers.

   

    

4. Putting a speaker of different configuration between two stereo loudspeakers, e.g., Khorns and a Belle or La Scala center, for instance, then setting up the center loudspeaker to sound the same (same "timbre") is also a very good exercise in using a digital active and in understanding what's important and what can differ while still maintaining timbre between different loudspeaker types.

   

    

5. Using a pair of loudspeakers whose HF response has been set flat inside an anechoic chamber as reference point really helps to set the concept of a "neutral" loudspeaker sound and allows the experimenter an instant reference to check back against while making changes on other loudspeakers in the array.

   

    

6. The effect of delay correction was more important than I would have imagined before hearing time-corrected loudspeakers vs. non-aligned.  In one case in particular, setting a sine wave tone at the center of the crossover band between midrange and tweeter, then dialing the time delay on and off alignment was much more sensitive than I would have imagined: just a few microseconds of misalignment would drive me out of the room. This setting is more important than the crossover filter type and center frequency.

   

    

7. Once everything was dialed in for each loudspeaker, my wife spontaneously mentioned that the system sounded much better than she had ever heard before, much better than she would have imagined in fact.  We can now hear detail, nuances, seamless frontal imaging from wall-to-wall, and surround effects in much greater detail and the naturalness of sound is now evident.

Edited December 14, 2014 by Chris A

[sunnysal]

I tried triamping and time aligning my system using an electronic crossover with driver and room correction based on DEQX equipment.  this was years ago and I ended up back with a "normal" system setup.  Was there an audible difference?  I think I can say yes, I think I heard some improvement in the areas mentioned by other posters (high end smearing, improved imaging) BUT I ended back to the "base" system because I simply didn't feel the extremely complex system provided enough of a difference to warrant the trouble, and frankly could not consistently be sure I was hearing a difference.  The measured responses were as near to perfect as one could ask for; impulse, frequency, phase, everything beautiful.  I am happy with the sound of my system now, the JBL tweeters and second, third and forth amps (I think I had monoblock bass amps IIRC) are up in a closet, the DEQX was sold, etc.  I would never think to suggest others should come to the same conclusion, especially because someone may come to the conclusion that all the extra equipment and effort DO provide them with significant aural improvements, just didn't work out for me.

 

Sometimes I get the itch to play around again, put back in the 2404Hs, maybe try a different mid horn and driver, etc.  but so far I have resisted!  warm regards, Tony

[Chris A]

I've not used a DEQX.  I have used the DSPs in the Crown XTi-1000 amplifiers (two of these units are now being used to power and EQ subwoofers in my main setup), EV Dx38s (one for center Belle, one for surround Cornwalls), and a Yamaha SP2060 for the Jubs.  In all my setups (except perhaps the XTi's) the results were significant--as measured by ear. The performance of the Yamaha is by far the best in terms of naturalness and detail, IMO.

 

If you run an upsweep of any Heritage loudspeaker with passive crossover--which is not time aligned--you will see really big phase shifts around the crossover points and sometimes big nulls at the crossover bands.  These aren't subtle disturbances and are at least partially responsible for the "smear" that you mention.  I recommend at least time-aligning the tweeter to the midrange manually by removing it from the front baffle and mounting it in its own small movable baffle on top of the unit at the back of the cabinet to align with the midrange driver below it.  As you know the Khorn has by far the worst time alignment issues: 6 ms for the bass bin to midrange transition at ~400 Hz, and 1.6 ms for the midrange to tweeter transition in the  4-6 kHz territory, which are huge delays at each respective frequency measured in wavelengths. 

 

One of the better Heritage loudspeakers in terms of time alignment is the Cornwall (in this case, with a Crites CT125 tweeter).  Below is a frequency response+phase plot taken at one metre (I couldn't squash all the phase into one plot because the total phase delay range is too big to handle in one REW plot).  Note also that the frequency response isn't what I'd call smooth:

 

Edited December 14, 2014 by Chris A

[Khornukopia]

I recently installed an Electro-Voice DC-One digital electronic three-way crossover for my Klipschorns. After I set the correct time delays, thanks to advice from forum members, I experimented by changing the time delay 0.08 milliseconds plus for the right speaker, decrease 0.08 ms from correct for the left speaker, which caused my speakers to sound wider apart. The inverse will make the speakers sound closer together. Also by playing a 500 Hz test tone through just one speaker channel and gradually increasing or reducing the time delay farther from the correct setting, the apparent source of the sound would shift around the room, incrementally and relative to the manipulated out-of-phase between the woofer and midrange driver.

 

 

That helps me understand why some arbitrary horn or crossover mods might make it sound like you have a wider or narrower sound-stage presence, and why some upgrades MIGHT NOT sound good if the physical or electrical parameters cause an out-of-phase condition that degrades the way things sound at your listening position. Could be part of the reason some speaker system modifications (even high quality parts and components) don't work as hoped for and are eventually removed by their owners.

 

 

The interesting thing is that plain old factory stock horn-loaded Klipsch speakers sound as great as they do, in spite of their horn paths, non-exotic parts and cost dictated design compromises. Makes me feel even more impressed with what Paul Klipsch did so many years ago.

[Bryant0086]

  Dude !! I seriously doubt you can hear 4.1 ms.

 

  Bryant 

[Chris A]

  Dude !! I seriously doubt you can hear 4.1 ms.

 

  Bryant 

 

Depends on frequency.  At 20 Hz, that'd be difficult to hear.  At 7 kHz, it's actually painful to listen to at the center of the crossover band. And it affects the entire performance of the loudspeakers, I'm afraid.  It's like having two bad notes consecutively on a piano.

 

I find that most people that deny that this is audible have never tried to correct it and listen A-B to the improvement.  I know that I'll never go back to using passives on horn-loaded loudspeakers again.

 

Chris

[Don Richard]

For grins, about a year ago, I decided to hook the passive crossovers up to see if there was any difference. You better believe there was a difference! I couldn't even listen to half of the song I was playing. I immediately reconnected the triamp setup, and all was well again. Not a subtle difference at all.