Hey Coytee / JC...

[DrWho]

Seems the thread was locked down before I could get around to your replies, but I was able to ****** a little bit of your posts using the search feature:

DrWho: I emailed Roy about that and hes not sure that I did it right.

He gave me a way to try it again using pink noise. Ill probably do it

again just to see if I can then tell a difference. Then again, is our

delay factor THAT great at something like 2271 uS? In other words, is a

delay factor of that long/short enough to be noticeable? As for how far

was JC sitting back (while at my house) Id say he walked around a fair

bit. The chair was probably in the neighborhood of 13/15 give/take but

he/they...

Well that's 2.271ms, which corresponds to about 2.5 feet. I'm trying to think of analogies in real life where you can hear the difference without speakers, but I'm having trouble finding something that correlates perfectly. Well...something that correlates perfectly that one might actually be able to experience in real life.

At 90Hz, I can dial time delay to within about 2ms by ear in a sound reinforcement rig. It gets easier when you go up in frequency because the phase interaction becomes more audible. I think it would also be easier in a home environment where the noise floor is much lower and the sound quality is simply better. My coworkers thought I was crazy at first so I had them dial it in by ear too. Even the "deaf" ones could bring it in to within 3ms. The differences are certainly smaller as you get closer to the ideal, so the importance shouldn't be blown outta proportion. I've heard it mentioned from a few different sources (like Roy and mas on the forum), that it pretty much gets inaudible once you get within one period of the frequency range (ie, 1ms at 1kHz). I'm not exactly sure how that plays out with the period of 100Hz being 10ms though...

Dialing in the time-delay only fixes the frequency response in a single axis though. The second you move your head up or down, you ruin the time-alignment since you're now a different distance from each driver. That's why a lot of manufacturers go to such great lengths to achieve coaxial designs...

However, I think there's more to the whole time-delay thing than just the frequency response. There's a change in timbre that happens due to the sound arriving at different times that isn't exactly related to the frequency response. Take for example a nice Hartke aluminum 18" bass guitar rig with a small conical horn tweeter. There's gonna be a huge gap in the midrange because the passband of each driver doesn't overlap...so no matter how misaligned the time arrivals are, the measured frequency response isn't gonna change. However, you'll notice a change in timbre with the bass guitar as you move the tweeter around in time...especially when playing slap bass. It's not as dramatic as the frequency response issues from overlapping passbands (like what happens in a crossover), but it's still there...or at least I think it is - I could just be crazy []

Richard. Yes it could be my new experience with two way. Don't know.

There were time at your house I found myself backing up. Not sure

exactly why. It could just be you have them in a small room and I was

trying to get an idea of the imaging or seperation of the "stereo". It

has been a little while so I can't recall exactly. The room is small so

there could be all kinds of speculation. May be a different once you

get them in a bigger room. It's almost like that "seperate" tweeter

coming from another

I had to eat crow on this topic with mas, but it seems there's a loose rule for how far away from the speakers you need to be before the sound gels together nicely. Apparently you need to be 10x the driver spacing away from the speakers - which with the Jubilee would put you around 20 feet back? I'm not exactly sure how this plays out, cuz that'd put it about 20 feet back for the khorn too...or a lot of other speakers we enjoy at smaller distances. I would imagine that it's somewhat Q dependant?

Nevertheless, I wonder if you didn't just feel like it sounded better the further back you got?

I'm kinda bummed the thread got taken down as I was looking forward to more discussion about tractrix horns too...

[Coytee]

I had to eat crow on this topic with mas, but it seems there's a loose rule for how far away from the speakers you need to be before the sound gels together nicely. Apparently you need to be 10x the driver spacing away from the speakers - which with the Jubilee would put you around 20 feet back? I'm not exactly sure how this plays out, cuz that'd put it about 20 feet back for the khorn too...or a lot of other speakers we enjoy at smaller distances. I would imagine that it's somewhat Q dependant?

Nevertheless, I wonder if you didn't just feel like it sounded better the further back you got?

<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

If you'll reflect then you will see one comment I've always been consistent with and that is "my" Khorns in 'my' room sounded 'to me' a WHOLE LOT better when I was standing about 25/30' away from them. I discovered that quite by accident.

I do not notice the change in sound 'quality' (bad word, I'm not sure which is the right word)...perhaps the sound coherence?

Anyway, I do NOT notice the change in sound quality of the Jubilees when performing that same exact experiment, which simply consisted of walking away while in the line of fire of a single speaker until I was at the base of the stairs. By going to that spot, another corner in the room (outside corner) ends up blocking the right side speaker (Khorn or Jube) and the left speaker is firing straight line to me.

When standing there, the Khorn simply sounded out of this world and noticeably more "lush" if I can use a stupid word.

Mike has suggested to me that what I'm noticing in my neophyte ways is the 'bubble' Roy keeps talking about and how with the 2 way the bubble comes together faster (closer to the speaker) than it does with a 3,4,5,... way which might need more time/distance for the sound bubbles to meld into a single bubble.

I don't know if that's poppycock or not but the explanation sure fits what I'm experiencing.

As for listening to what MAS has to say.... I'm starting an uncalled for (and perhaps vicious?) rumor that he's learned everything he knows at the beak of Mr. Parrot... [] []

[DrWho]

I think sound coherance is exactly how I'd describe it too [Y]

I believe the 10x driver spacing mas was referring to involved overlapping passbands between drivers - and the 'rule' was derived by taking a crap load of measurements. So in the case of the Jubilee, I wonder if it's only a concern throughout the crossover region....but at 24dB/octave that's not much of a range for the overlap so it's probably not very noticeable (especially that 500Hz region...not a lot going on there in terms of intelligibility). Or maybe Roy has just produced a product that defies the rule...there's always exceptions to every audio rule. Perhaps Roy is a magician too? [] Anyways, do you happen to notice any change in the crossover region?

Btw, I'm looking forward to your time-alignment comparisons if you ever get around to doing it again. Heck, I've got a Dx38 here that I can program and send you the config file so all you have to do is load it and not think about it.

[jwc]

Not sure how important me moving further from the source is gonna prove anything or not. I found myself doing that in Hope too.......I migrated to the opposite corner when listening to it (the solo jub with 402).

I don't seem to do that at home however my listeing may be 15 feet back. When I was at mark1101's house checking out his clone setup....I don't recall "moving" back. I stood there right in the middle of this unfinished room.

Yes.....I too am bummed at the loss of that thread. Down for good reason.....but I actually had to spend a lot of time trying to word my posts the best I could so that I wouldn't be misunderstood and trying to portray my feeling as how certain things sound. That is hard. These "new" products by Klipsch are on everyones mind as we are contemplating there placement in a Home.

There was a lot of other folks description of stuff too that was interesting.

I am completely ignorant with time delay. Can't really share my thoughts as my head is empty on that subject. I have heard passives on the jub and it sounds fine. Maybe there is some critical error that I haven't picked up.

jc

[Amy]

If there's anything you want me to fish out and send you from recently vanished threads, just let me know. It will be a lot easier to find the good stuff than to weed out the bad.

Send me an email: amy.unger@klipsch.com

[Coytee]

Anyways, do you happen to notice any change in the crossover region?

Btw, I'm looking forward to your time-alignment comparisons if you ever get around to doing it again. Heck, I've got a Dx38 here that I can program and send you the config file so all you have to do is load it and not think about it.

You're making me laugh...have you not yet inferred from my comments how clueless I'd be on your first question? I'd have NO idea of what to listen for. As I recall Roy's comment, I think he said to run pink noise, change the delay to zero and by doing that, I might be able to hear the difference somewhere. I do not recall his exact commentary and it's not here for me to refer back to (I fear I may have deleted it).

Regardless of all that... I'm not TOO concerned about it anyway. Heck, if they can make it sound like it did in Hope with no delay then I just got to think we're splitting hairs as well as shooting for "perfection" when in fact, (speaking only for myself) this dolt can't necessarily tell the difference.

hmmm... Who... are you near a phone? I just had something dawn on me and I don't know if I can explain it in words (though I'll try)...call me if you're around (or someone else who's knowledgable of this time delay technical stuff....well ..... ok, even you MAS...[6])

My work number is 888-531-0715

Here's something that just now popped into my head.

Scenario: I've got a LaScala downstairs that I use when I'm working down there. I've got a 100' cord going from it, upstairs and it ends near my 2102 and I simply swap wires out when I want to plug it in. The 2102 powers both K-402's and I use the left channel (doesn't matter, just trying to give the visual). Another tidbit...I've got a setting I use on the Dx where when I do this, I hit that setting and it sends a regular filtered signal to the RIGHT Jube (and includes the signal from my bass amp). The "LEFT" channel has the bass signal truncated so the left Jube plays nothing. I've taken the wires to the left K-402 and removed them from the amp, replacing them with the line to the LaScala. This channel on the DX has been changed to be unfiltered so the output of the DX should be the whole signal being fed to the MC-2102 then fed to the passive on the LaScala and the LaScala plays full range (have I said yet that I love this Dx-38 and what it can do?)

OK, here's the deal... I've got a setting on the Dx that plays the LaSclala ONLY, and another setting that plays the upstairs right side Jub with its appropriate signal, and it also plays the downstairs LaScala with a full signal, fed from the same amp that is otherwise playing only the HF signal for the Jube.

(with me so far?)

On the Dx, I've taken the settings and presuming I did them right, I've summed the R+L channels on the input side so the DX will have a mono signal to BOTH speakers.

(still with me?)

Here's the deal... when I have both speakers playing, there is a CLEAR and DISTINCT delay between the two speakers. I've since come to learn ...well..lemme tell you what I did first...

Just last night I reswapped the line from the 2102 going to the LaScala and replaced it back with the left side Jube K-402 line. Remember though that the Dx has internally KILLED the Jube bass bin on this side. So, during this process, I've now got a mono signal, FULL signal going to my left side K-402 and on the right side, a properly filtered signal feeding BOTH the 402 AND the bass bin (with its appropriate amp).

When I killed the right side bass bin (the only one playing since the left one is killed inside the Dx), the sound was the 'same' coming from each speaker (402), there was no delay nor echo. If I did NOT kill the right side bass bin, then I heard the same delay that I heard when I had the system hooked up to the LaScala downstairs. Mike (Beasley) thought it was just room echos but now, I've brought it into the same room so that's not it.

If I leave only the right side playing (HF and LF) the sound is there and great. If I play just the left side (HF only) the sound is 'great' given the constraints that it is a full signal going through the 402 only (LF cut out). BTW, I've only done this at LOW volume so as to not destry the driver with a full signal.

If however, I play BOTH HF's at same time (left side = full signal, right side = properly filtered signal) it sounds fine, but when I bring the right side bass bin into the equation, I get that echo.

So, it seems as though it's the BASS bin on the right side that is causing this echo to happen. What I'm finding real interesting is, I do NOT hear this echo with just the right channel playing (HF and LF) nor do I hear it when I have just the HF's playing (right and left with both bass bins terminated)

Sooooooooooooo.... if you've read this far and followed my story (I hate typing this kind of stuff out, it's hard enough to describe verbally, much less in print)

Anyway, could the delay built into the system be the reason I'm hearing this echo? Truth be told, I've tried to dinker with the delay setting and STILL can't get it ironed out. Using the downstairs LaScala with the upstairs Jube is such a disparity in location that it doesn't matter to me a GREAT deal, but it does irk me at times when I hear that distinct echo (like when I'm walking up the stairs)

anyone care to take a stab at that? if so, you're welcome to call me for further clarification.

[]

[Coytee]

If there's anything you want me to fish out and send you from recently vanished threads, just let me know. It will be a lot easier to find the good stuff than to weed out the bad.

If you can find all of my comments about your ear lobes...[:#] is the word

[colterphoto1]

Richard- that post made my brain hurt, can I go home now? [:S]

[Coytee]

Exactly why I gave my number and otherwise am not a big fan of typing out issues/situations. It's hard to get to the detail that might help someone for example, I was having issues with an amp cutting out (protection circuit) and it was finally through several discussions with Mark (Deneen) that he pegged the problem. There is no way all of that could have happened had it been strictly typed out. I would have probably left out a small typed detail because of the effort and as it turned out, he would have missed the single clue he needed to figure out my problem.

Besides... my wife says I'm full of a LOT of hot air... (or something like that) []

[DrWho]

Hey Richard, cell phone minutes are at a premium over here so if you'd

like, feel free to give me a call sometime after 9pm CST if you're

still awake. 630-715-6788

(I'll be out until 10:30pm tonight so call a bit later if you call tonight).

Nevertheless, let me take a stab at it and see if I'm understanding correctly. First, the signal path...

You've got your CD player going through your preamp and into channels 1

and 2 of the Dx38. The Channel 1 output that is normally connected to

your left 402 is getting a full range sum of IN1+2 with no delay.

The Channel 3 output is connected to the right 402 with the sum of

IN1+2, delay, corrective EQ, and high pass filter. And finally, Channel

4 is connected to the right bass bin with the sum of IN1+2, its

corrective EQ and low pass filter. When setup in this configuration,

you only notice an "echo" when the right side bass bin is turned on?

What happens if you add that 2.2ms of delay back into the Channel 1 output? I bet the echo goes away. Explaining why will be a bit more tricky, but I'll wait to see if that fixes the problem before I waste my breath. []

[DrWho]

I am completely ignorant with time delay. Can't really share my thoughts as my head is empty on that subject. I have heard passives on the jub and it sounds fine. Maybe there is some critical error that I haven't picked up.

Perhaps ignorance is bliss?

I wonder if perhaps the time-domain behavior of normal room acoustics doesn't trump the improvements of time-alignment with the speakers. Heck, with two walls, a floor and a ceiling you've always got four "extra speakers" on each side adding all sorts of misalignments. Perhaps it's easier to hear with PA since the only "extra speakers" you gotta deal with inside the Haas window are the floor and the stage volume...and you almost always got people soaking up the floor and you dial in the system before the band shows up.

[PrestonTom]

I am completely ignorant with time delay. Can't really share my thoughts as my head is empty on that subject. I have heard passives on the jub and it sounds fine. Maybe there is some critical error that I haven't picked up.

Perhaps ignorance is bliss?

I wonder if perhaps the time-domain behavior of normal room acoustics doesn't trump the improvements of time-alignment with the speakers. Heck, with two walls, a floor and a ceiling you've always got four "extra speakers" on each side adding all sorts of misalignments. Perhaps it's easier to hear with PA since the only "extra speakers" you gotta deal with inside the Haas window are the floor and the stage volume...and you almost always got people soaking up the floor and you dial in the system before the band shows up.

Mike,

You are probably on to something here. The recent research on the audibility of group delay (which is what we are talking about re: time alignment) suggests that the minimum delay required for detection will increase in a reverberant space (i.e., becomes harder to detect the group delay). So pehaps the effect (in this case the possible improvement) could be obscurred by the room.

Good Luck,

-Tom

[Coytee]

Hey Richard, cell phone minutes are at a premium over here so if you'd like, feel free to give me a call sometime after 9pm CST if you're still awake. 630-715-6788

(I'll be out until 10:30pm tonight so call a bit later if you call tonight).

Nevertheless, let me take a stab at it and see if I'm understanding correctly. First, the signal path...

You've got your CD player going through your preamp and into channels 1 and 2 of the Dx38. The Channel 1 output that is normally connected to your left 402 is getting a full range sum of IN1+2 with no delay. The Channel 3 output is connected to the right 402 with the sum of IN1+2, delay, corrective EQ, and high pass filter. And finally, Channel 4 is connected to the right bass bin with the sum of IN1+2, its corrective EQ and low pass filter. When setup in this configuration, you only notice an "echo" when the right side bass bin is turned on?

What happens if you add that 2.2ms of delay back into the Channel 1 output? I bet the echo goes away. Explaining why will be a bit more tricky, but I'll wait to see if that fixes the problem before I waste my breath. []

Seems I lied... I went down to change the delay. I've done it before and never noticed anything. I figured I'd try again & make you happy. I did it, no difference however, I did discern the left side being a touch louder than the right...that got me wondering.

I ended up looking into the R+L summation for both channels and did NOT have the right channel summed into mono. I changed it to mono via the Dx and the delay issue and the output imbalance went away.

[bodcaw boy]

hey coyotee-o,

time delays are a little harder to discern when you are using high slope crossovers. then you have to specifically pick source material that excites that band. the lower the slopes the more noticeable as the overlap in the crossover region is larger with lower slope crossovers. miketn, or was it foggy shawn, is correct in that simply removing the delay will create a hole in the response. then we have to move the crossover so that the dip is alleviated as much as possible. what you then get is above and below crossover point, amplitude problems because of the phase mismatchup. that is when you begin to notice the coherency begin to diminish. again, you have to excite the band that this is happening in.

hope this helps.

have a blessed night,

roy delgado

[Rivendell61]

The recent research on the audibility of group delay (which is what we are talking about re: time alignment) suggests that the minimum delay required for detection will increase in a reverberant space (i.e., becomes harder to detect the group delay). So pehaps the effect (in this case the possible improvement) could be obscurred by the room.

Tom can you point me to that source? Testing by Sony found the opposite--detection of delay, phase, and modulation artefacts was better in a minimally treated, "moderately" reverberant room, than when using headphones or in anechoic conditions.

Mark

[PrestonTom]

Mark, The reference was published in J Audio Eng Society about a year or two ago. One of the authors was Moore. I don't currently have access to a search function right now.

-Tom

[sfogg]

Mark,

"Testing by Sony found the opposite--detection of delay, phase, and modulation artefacts was better in a minimally treated, "moderately" reverberant room, than when using headphones or in anechoic conditions."

Doesn't that suggest that the audibility of the delay *by itself* is suspect. But that the secondardy effects of the delay (comb filtering with the attendant polar response changes...etc...etc...) may be what is more audible?

Headphones would be one way so delay would not be effecting comb filtering. In an anechoic chamber the off axis polar response changes from the comb filtering would be irrelevant on axis.

I have not played with this much in my system but I did not notice much (if any) of a difference in altering the delay of my high driver relative to my lower driver. In my system I am running crazy steep crossovers.... around 210dB/octave for 35dB or so. At the 500hz crossover point I am down like 15dB relative to each other within about a 50hz. My area of overlap is very small, so the potential area of comb filtering is also very small.

If I changed to first order crossover that may totally change the results.

Shawn

[PrestonTom]

The formal testing that was done, of course, did not use music as a test signal. I am not sure what the Sony guys were using, they may have not done formal testing.

Rather, the test signal was broad band and contained strong impulsive characteristics ( I forget the specifics, but as I recall the signal was a good choice). Using music is problematic since the specific selection may or may not be particularly revealing (and others cuts could be) so it is a bit of a crap shoot. Also in a formal test, you would want to minimize the role of memory in the discrimination procedure. Again, using music (especially a long duration cut) presents these problems. This is why some folks may or may not hear the differences.

For the uninitiated, the idea of "testing" a music reproduction system and not solely relying on music as a "test signal", may seem strange. However, you want to test for the discriminability of very small changes in sound, and a poor choice of signal can obscure the real differences. That is why the test signals are designed the way they are (it is certainly not an informal procedure).

Good Luck,

-Tom

[DrWho]

"strong impulsive characteristics" - that would be the kind of sounds I would be listening for when testing for it.

I am curious though how group-delay plays into everything because most drivers generally add more delay as they go lower in frequency...and they have a crap load of it too...and time-arrival offsets aren't going to be very different in their audible nature.

I wish I knew the delays involved with that one bass guitar player that was tweaking his sound by delaying his woofer...I coulda sworn it was on the order of a few ms, but who knows, maybe it was a lot more. I wish I had more time to play with this kind of stuff...

[Rivendell61]

Tom,

Thanks. If you ever stumble across it again I'd be interested. Ref. your second post: not quite sure your definition of formal but the Sony tests were ABX. Sources were various: sine, square, pink, and yes even music--depending on what was being diagnosed/investigated.

Mark,

"Testing by Sony found the opposite--detection of delay, phase, and modulation artefacts was better in a minimally treated, "moderately" reverberant room, than when using headphones or in anechoic conditions."

Doesn't that suggest that the audibility of the delay *by itself* is suspect. But that the secondardy effects of the delay (comb filtering with the attendant polar response changes...etc...etc...) may be what is more audible?

Headphones would be one way so delay would not be effecting comb filtering. In an anechoic chamber the off axis polar response changes from the comb filtering would be irrelevant on axis.

Shawn,

Well...I don't know the 'why' but I am sure you are on the right track. The Sony stuff I have seen does not really speculate on cause--except generally to state the obvious. The testing was mainly done (AFAIK) during converter testing/refining and also to investigate related things like ultrasonic audibility, etc. So the fact that the most sensitive method of testing also corresponded to the situation in which most people would be using the equipment is something of a plus.

They got some very interesting results like.....diagnosing a 1/4 sample delay problem between channels (L/R). Which is darn tiny stuff (about 5 microseconds--I believe a tad below the normally accepted audible limit) --and even sub-1/4 sample results.

Floyd Toole seems to have said much the same thing (re better listening test results in room) at his AES Master Class in acoustics but....apparently, while he had the data to support it he also had no explanation.

Mark