How to figure how much delay is needed

[The Dude]

I am thinking out loud here, but with my new system I am working on a active crossover the minidsp. I here this and that about the horn length were the folded bass horn and the strait hf horn need some delay. Is there a way to calculate this based off of specs or in my situation, would it just be measure and testing.

[Chris A]

You can calculate based on sound traveling 1132 feet/second (13584 inches/second) at room temperature (about 72 F or 22.2 C): this is 73.6 microseconds per inch of physical offset between the acoustic center of the drivers. Calculate centerline path length of the horns: don't forget to add the driver housing delay due to the diaphragm being farther back from the interface of the horn mouth and the driver face. Add the digital delay to the miniDSP channel connected to the shorter-length channel (e.g., the tweeter channel if synchronizing the midrange/tweeter).

Once you dial in that number of milliseconds delay between, for instance, the midrange and tweeter, I recommend adjusting the delay up and down in increments of 20 microseconds per step and listening to the soundstage effects.

You should be able to hear when the soundstage "snaps in" becoming large and coherent and the timbre of the speaker changes to become more neutral. This will happen when you get within 1/10 of a wavelength of the crossover frequency: in this case, this is usually about 1/4 inch at tweeter/midrange crossover.

Setting the delay between the bass bin and the midrange horn/driver is typically 10x less sensitive due to the increased wavelengths at these crossover frequencies. It's also a bit more difficult to hear when this happens. Calculate the theoretical delay and dial that in for the midrange/tweeter channels. Then listen for changes in overall timbre and pureness of tone in the "tuning fork" or middle C range. Add steps of delay to the midrange channel of about 200 microseconds each step. Make sure that you change the delay on the tweeter and midrange channels by the same amount.

Chris

Edited April 23, 2014 by Chris A

[The Dude]

Chris, thanks for chiming in. I always value your input, but I don't have a tweeter, just the ev hp 640 horn on top of the Jamboree. So what you are saying about it not being as sensitive with the Mid horn to the Bass horn, does that mean it takes less delay.

Edited April 24, 2014 by duder1982

[mustang guy]

In short, add a millisecond for every foot a folded horn is longer than other horns. It's all about the distance the sound wave travels to get to you. In a 10 foot folded horn, there is about a 10ms delay just coming out the mouth.

[The Dude]

So I would have to find out the path length of the folded horn.

I have read that I should try and eq the lf and hf first then set the crossover. But I don't even know what needs equing at this time do to lack of measurement devices. But I started with just setting the crossover to 600 hz with a 48 db Linkwitz-Riley filter. I thought it sounded decent right off, Then I dropped it down to 24 db, it sounded muffled not very good. Then I just tried 12 and things seemed to blend better sounded smoother. I was reading on linkwitzlab.com about some basic stuff on crossovers, as i am just trying to understand. I thought I would try and set some stuff up just to see how good it could sound without equing.

[Chris A]

...So what you are saying about it not being as sensitive with the Mid horn to the Bass horn, does that mean it takes less delay.

Well, no, not really. You can calculate the total centerline path length of the Jamboree horn from the compression slot to its mouth exit plane. Then do the same for the mid/high horn. Calculate the difference in path length in inches between these two horns, then multiply that differential path length by 73.61 to find the first cut on delay for your EV horn and its driver, in microseconds.

I'd go ahead and crank that delay in to the mid/high channel then go back and EQ the two horn/drivers relative to each other, since time misalignment alters the timbre heard from the loudspeaker. Once you select your crossover slope (12/24/48 dB/octave L-R, etc.) then go back and step through delays in 200 microsecond steps in both positive and negative directions to the calculated delay setting, until you hear the effects that I identified above. (But I wouldn't go more than about 1000 microseconds in either direction during this trial.)

Once you find the point of maximum soundstage and tonal balance, write the miniDSP settings down. You can then re-EQ the loudspeaker at that point.

Voila! No measurement equipment required.

Chris

Edited April 24, 2014 by Chris A

[The Dude]

Since you said this,

since time misalignment alters the timbre heard from the loudspeaker

after this

EQ the two horn/drivers relative to each other

I would think are talking about the EV horn being relative to the Jamboree horn. one thing, do most horn companies have general eq settings to be used. I ask as I am new at understanding all of this and have tried to research all of this on the net but only seem to find things on car speakers. All so I ask do to the fact that I shouldn't need to measure anything.

So if I am understanding this needs to be tackled in this step,

1 Time alignment,

2 Eqing,

3 Crossover setting.

Thanks, I think you will find out this won't only help me but other members who are trying this whole active thing out. Your big help around this place.

[Rudy81]

If you want to jump into the deep end of the pool, you might take some time and review this thread. https://community.klipsch.com/index.php?/topic/112615-active-crossovers/

Lots of good information on alignment, phase, etc. Been in your shoes and took a while to get it, but with everyone's help I am able to tweak things to my satisfaction.

I would recommend some basic gear to help you with accuracy. A basic mic, sound card and laptop can work wonders. I use REW software as my main suite for this type of thing. You can also find lots of information on this topic with a google search. It can be overwhelming at times.

Have fun!

BTW, after using the steeper 8th order L-R crossover (48dB), I went to 24dB. Rane has some great papers on this issue. http://www.rane.com/note160.html

Edited April 24, 2014 by Rudy81

[JL Sargent]

You might find that time alignment has absolutely no affect whatsoever on how you hear the music. I have used many different active setups with folded horns, short horns, direct drivers, etc. I have found that only EXTREME manipulation of the time really affects the way I hear the music. PWK stated that most people do not hear an improvement with corrected time alignment. I suspect he was correct because KHorns, Belles and La Scallas shipped from Hope Ark. did not/do not have it.

Edited April 24, 2014 by JL Sargent

[Rudy81]

You might find that time alignment has absolutely no affect whatsoever on how you hear the music. I have used many different active setups with folded horns, short horns, direct drivers, etc. I have found that only EXTREME manipulation of the time really affects the way I hear the music. PWK stated that most people do not hear an improvement with corrected time alignment. I suspect he was correct because KHorns, Belles and La Scallas shipped from Hope Ark. did not/do not have it.

I'm no expert on the subject. However, my experience is that moving the delay back and forth affects the frequencies at and near the crossover point.

I have found that if I use my mic to measure the db level at the crossover point, and I move the timing of one driver back and forth, I register a change in db level. Basically, as the driver phase comes into alignment, the db level is at max for a giving input. If I move the timing in either direction from that max, the db level decreases until a minimum when the drivers are 180 degrees out of phase.

In my system's case, I have had to pay attention to the timing of the drivers in order to not produce an out of phase dip at and near the crossover frequency.

Let me add that I don't want to confuse the time alignment and phase issues. I realize you can also achieve the same effect by messing around with the driver phase.....but that doesn't mean they are time aligned. I start off by using my impulse response function to obtain the delay of each driver, then determine the difference and start there in order to fine tune by obtaining maximum db level at the crossover frequency.

Also, from what I have learned, IF you eq after having time aligned, you will need to go back and double check the driver phase alignment.

Edited April 24, 2014 by Rudy81

[Chris A]

You might find that time alignment has absolutely no effect whatsoever on how you hear the music. I have used many different active setups with folded horns, short horns, direct drivers, etc. I have found that only EXTREME manipulation of the time really affects the way I hear the music. PWK stated that most people do not hear an improvement with corrected time alignment. I suspect he was correct because KHorns, Belles and La Scalas shipped from Hope Ark. did not/do not have it.

Note that the Heritage line of speakers was marketed and sold during the years when digital active crossovers were not economically available. Therefore, PWK couldn't do much about time alignment of horns/drivers in the Heritage line. By the time that good digital crossovers were widely available at reasonable prices, PWK had passed. Nowadays, the Klipsch professional/cinema line of loudspeakers publish model delay corrections, and have been doing this for at least a decade.

I've also found that some just have a difficult time hearing differences in different areas of loudspeaker performance. This situation usually takes care of itself without anyone having to get involved.

I can hear time misalignment very clearly in the center Belle between my Jubilees. Before I got the time delay settings correct, I was ready to dump the Belle as a center between Jubs. It has been a pleasure since to have the Belle there. It's very easy to demonstrate what happens when you don't correct the time misalignments in the Belle: it's not subtle.

Chris

Edited April 24, 2014 by Chris A

[JL Sargent]

I have a pair of Belles and honestly have never felt the need to align them. I have not tried to introduce a center Belle between a pair of Jubilees or own a movie theater so I guess you got me there. LOL.

[Chris A]

EQ the two horn/drivers relative to each other

...Do most horn companies have general EQ settings to be used.

So if I am understanding this needs to be tackled in this [order of] steps,

1 Time alignment,

2 EQing,

3 Crossover setting.

Thanks, I think you will find out this won't only help me but other members who are trying this whole active thing out. Your big help around this place.

Thanks for the kind words. Note that I answered one question above: Klipsch publishes this info for their cinema line.

I do try to think of other readers when I attempt to answer questions since I believe that the questions are really tips of the iceberg. I suggested the order of steps above as my approach to the problem--but it's clear that it's not the only way.

I try to keep things simple; in this case, using measurement equipment is a "nice to have" to check out results. Rudy's approach is even easier if you have a hand-held SPL meter available and a microphone stand to mount it on to make sure that the SPL meter doesn't move around between measurements.

But I find that dragging out the equipment and getting it set up sometimes is more than what I want to do. I'm basically "the proverbial lazy engineer".

Chris

Edited April 24, 2014 by Chris A

[The Dude]

Guys some good info here. Rudy I will have to spend so time reading your thread. One thing I will say is I didn't care for 24db step, could this change once things are dialed in. Maybe this is one of the reasons you crossover after fine tuning everything.

I do own the umk1 Mic and a sound card my laptop craped out on me this weekend. So either need to find a way to get some extra long usb cables or do this the old fashion way.

I tried this once and failed but this time I aim to figure it out, as I am up for the challenge.

[Rudy81]

Guys some good info here. Rudy I will have to spend so time reading your thread. One thing I will say is I didn't care for 24db step, could this change once things are dialed in. Maybe this is one of the reasons you crossover after fine tuning everything.

I do own the umk1 Mic and a sound card my laptop craped out on me this weekend. So either need to find a way to get some extra long usb cables or do this the old fashion way.

I tried this once and failed but this time I aim to figure it out, as I am up for the challenge.

I think learning how to do this is well worth the effort. In the aforementioned thread, pay particular attention to the comments by Dr.Who. He and Mike Bentz really helped me out as I figured out how to do things. It is a bit of a learning curve.

I originally tended to use the 48db slope, thinking a steeper slope would allow each driver to act more independently and this avoid some of the crossover issues. However, after reading many opinions on the subject, one issue that stands out is the problem of HF horn vs. LF section characteristics. If there is a big difference in each section propagates sound, you are more likely to notice this difference with a steep slope. The shallower slope has a wider overlap, so might present a smoother transition.

Best thing is to try various option, always optimizing the phase and time alignment, and seeing which you like best. If your crossover has a memory function, use it to create various setups and then just load the various options.

[The Dude]

You can calculate the total centerline path length of the Jamboree horn from the compression slot to its mouth exit plane. Then do the same for the mid/high horn. Calculate the difference in path length in inches between these two horns, then multiply that differential path length by 73.61 to find the first cut on delay for your EV horn and its driver, in microseconds.

Ok if my measurement is correct I should have a horn path length on the Jamboree about 42"( I don't know how accurate I need to be but I could probably subtract a couple inches but at least its a good place to start.). Then the Ev is 19" so 42-19 is 23 the multiplying the difference by 73.61( where does this number come from), equals 162.71 is that milliseconds. Or if that is microseconds then it would be .16271 milliseconds.

Edited April 25, 2014 by duder1982

[The Dude]

If you want to jump into the deep end of the pool, you might take some time and review this thread.

I have started to read that thread, I will try to take in as much as possible,

Chris we should see about adding a tutorial on how to set up a active crossover with your pinned thread on Active bi amping. Your and Rudy, Dr. Who and anyone else could write it up.

Edited April 25, 2014 by duder1982

[The Dude]

Ok I was wrong, I still don't know how accurate I am on the bass horn but lets say it is 42". Then the EV is 17.5, so now 42-17.5=24.5*73.61=1803 microseconds or 1.8 milliseconds. I will plug that in tomorrow, I all so found the eq settings for that horn. I know I will still need to take some measurements to see how it produces, and to get everything correct. But it may be a good starting point.

Chris do you still use Roy's eq settings for the K402 since you have the TADs or did you just use it as a starting point. Reason I ask I have a different drivers on the EV horns and wonder how much it would even make a difference.

[Chris A]

...I still don't know how accurate I am on the bass horn but lets say it is 42". Then the EV is 17.5, so now 42-17.5=24.5*73.61=1803 microseconds or 1.8 milliseconds. I will plug that in tomorrow, I all so found the eq settings for that horn. I know I will still need to take some measurements to see how it produces, and to get everything correct. But it may be a good starting point.

Chris do you still use Roy's eq settings for the K402 since you have the TADs or did you just use it as a starting point. Reason I ask I have a different drivers on the EV horns and wonder how much it would even make a difference.

The path lengths that you quote look about right. You might find that the Jamboree is a couple of inches shorter in path length from my readings on the forum, so I'd recommend beginning iterating the 1.8 ms delay to be a little bit less to start with: perhaps 0.1 to 0.2 ms less to start with.

I use Roy's EQ and crossover settings above 270 Hz, but I have modified his settings below that point via in-room measurements using REW and a DEQ2496 with calibrated microphone, which you can see below in the unsmoothed FR plots using REW. The light green trace is the raw (un-EQed) Jub bass bins, and the dark green trace is the EQed Jub bass bins with TH-SPUDs.

Your midrange horns and drivers (...which ones?...) would be much different since your midrange horn looks to be a "collapsing polars" midrange horn. I'd start with anything that you can find on Google with regard to EQ settings for that horn-driver combination, if you can find them. If you can't find anything, I'd start with no EQ filters since many non-constant coverage midrange horns of a few years vintage were designed to require little EQ due to their collapsing polar design.

Edited April 25, 2014 by Chris A

[Chris A]

Chris we should see about adding a tutorial on how to set up a active crossover with your pinned thread on Active bi amping. Your and Rudy, Dr. Who and anyone else could write it up.

I'll take a look at Rudy's thread when I get a chance to look for stuff to plagiarize: it may take a while.

My guess is that more detailed steps might help folks using active crossovers. I think that this is a subject that really results in big improvements in sound quality and reduces overall costs for "audiophiles" who are constantly searching for the right high-dollar/low power class A (...or class D, AB, etc.) amplifiers to drive their horn-loaded loudspeakers. It takes a lot less power per amplifier when bypassing passive crossovers by using active crossovers.

Chris

Edited April 26, 2014 by Chris A