A time-aligned top end - Part II

[Al Klappenberger]

Lee,

Thanks. I didn't know that! All I knew was that it increased. I didn't realize it was to that degree!

This illustrates how tough it is to go from 400 Hz to even as high as 15KHz with a single driver. I'm hoping with the right driver and network it might be possible.

Al K.

[Rudy81]

Dave, any chance you will have an audition pair of horns available when you settle on a production model?

[SSnyder]

Q- has any one of you run DBT for time alinegment or just theoryiezing???? I'm am very skeptacil

The B&C DCX-50 is a coaxial driver, meaning the Tweeter is located inside the midrange therefore the sources are the same distance from the listener. Therefore yes it is time aligned.

Good info here. http://www.gain11.com/ConsumerAudio/TimeAlignmentPartI/tabid/177/Default.aspx

THERE IS NO TIME ALINGEMENT IN CONCERT!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

I am not sure what you are getting at here, but there is a difference between production and re-production. Unless you are Milli Vanilli.

Dave

simpley this>>>> in live concert bass and trebel arrive at lisener at slightley diffrent times _____ BUT IT DOESNT' MATTER! and concert venyu is MUCH BIGGER ... . .. . .. so why would it matter in a LR

A SOLUTON TO A PROBLEM THAT ISNT'!!!!!!!!!

[sfogg]

"This is the key to what this is all about. A lot of these drivers will
specify the crossover with an asterisk * footnote. When you read it, it
will tell you something like 12 dB / octave minimum filter. This is
telling you that it can't handle even moderate low frequency energy
without distorting"

44xt

1000hz-20k Hz

320w peak power

160 watts program

80 watts continuos on pink noise ine the range specified

500hz to 20kHz

200w peak

100w program

50w continuous

The post about quadrupling throw when you go an octave lower is dead on.The flip side of that is the more efficient the driver is in the first place the smaller its diaphragm movement. Quadrupling the throw of a diaphram that is moving .1mm is a lot easier then quadrupling the throw of a diaphragm moving 1mm.

Shawn

[sfogg]

"> in live concert bass and trebel arrive at lisener at slightley

diffrent times _____ BUT IT DOESNT' MATTER! and concert venyu is MUCH

BIGGER ... . .. . .. so why would it matter in a LR

A SOLUTON TO A PROBLEM THAT ISNT'!!!!!!!!!"

You misunderstand the problem. In a concert you have dozens of instruments in space playing their notes. Each instrument is playing the entire frequency range of that instrument it is a single reproducer of its music.

To keep it simpler with one 2 way speaker you have a tweeter and a woofer trying to reproduce all the music from the concert. In a typical situation the woofer and the tweeter are both trying to reproduce some of the same material. If they are not time aligned with the listener (or crossed very steeply) the two drivers are interfering with each other and causing changes to the waves they are reproducing.... altering the signal from what they should be reproducing. With the path length differences in horns this is*easy* to hear in action if you do a simple test.

To get more complex... now consider the same thing occuring between all the drivers in the left speaker and all the drivers in your right speaker. You now have 4 drivers trying to reproduce the sound a single instrument makes in space. This causes many issues... again easy to hear in action with simple tests.

To get even more complex.... even if you are time aligned at the listening position unless you are concentric (or very steeply crossed) the response at other points in the room aren't time aligned and it alters the response at those points. You hear this too due to the reflections within your room.

Shawn

[SSnyder]

To keep it simpler with one 2 way speaker you have a tweeter and a woofer trying to reproduce all the music from the concert. In a typical situation the woofer and the tweeter are both trying to reproduce some of the same material. If they are not time aligned with the listener (or crossed very steeply) the two drivers are interfering with each other and causing changes to the waves they are reproducing.... altering the signal from what they should be reproducing. With the path length differences in horns this is*easy* to hear in action if you do a simple test.

I follow that you may not want tweter and woofer over-lappig butt dont' see that as time alignment!?!?!?!?!?

Why not single driver ala open baffel design???

so you have indede tested it personaly???

[sfogg]

"Here's one I think deserves a serious look at: http://www.greatplainsaudio.com/downloads/390series.pdf"

I believe that is basically the Altec 290. From what I have read most cross those over around 4500 or so so you will be more limited in what you use for a tweeter.

Shawn

[Guest David H]

"Here's one I think deserves a serious look at: http://www.greatplainsaudio.com/downloads/390series.pdf"

This driver keeps popping up, and it appears to be an excellent performer. I certainly hope it crosses higher than 4500Hz. I much prefer the midrange driver to get to 6k+

Dave

[djk]

http://alteclansingunofficial.nlenet.net/publications/techletters/TL_231.pdf

[sfogg]

"I follow that you may not want tweter and woofer over-lappig butt dont' see that as time alignment!?!?!?!?!?"

In all multiway speakers there is always some measure of overlap.

In first order crossovers the overlap spans octaves. Time aligning the drivers (either electronically or physically) makes it so that the arrival times from the two drivers are the same to the listener. That makes the overlap in phase to reduce the interference between the two drivers. There are still issues elsewhere in the room which is altering polar response which is also effecting the sound at the listening position.

The other way to reduce the effects of interference is to reduce the overlap. For example the crossovers I'm using drop over 30dB in about 1/6 of an octave. This has dramatically lower overlap which reduces interference effects everywhere in the room. With first order crossover there is nearly 5 octaves of music before they drop as far as my crossover does in 1/6 of an octave.

"Why not single driver ala open baffel design???"

Because that has its own set of trade offs. Dynamic range constraints, typically reduces bandwidth, uneven polar response (beaming up high). But like you said one of the benefits is the removal of that overlap between drivers which gives these a coherence that most multi-way speakers don't have.

"so you have indede tested it personaly??? "

Changes in audio due to changes in timing between the speakers is pretty easy to hear. If you have a test CD that has pink noise play that through one of your speakers. Move vertically compared to that speaker.. you are basically changing the relative path lengths to your ears. The character of the pink noise will change. How much will depend upon the slope of your crossovers and the physical alignment of your drivers.

Shawn

[DrWho]

"In the search for possible Drivers for use with the Eliptracs, I have run across a few I think might work well 2-way."

I'm still very happy with the Peavey 44Xt run two way with the K510.

Shawn

Did you post any curves of that arrangement?

[DrWho]

The B&C drivers all look like they are intended to be crossed over much higher than 400 Hz. The DE82 looks like it starts down way above 400 Hz on the official curves. Just because it's response will go that low doesn't mean it will do it without distortion. I found that out the hard way with the 1-inch JBL 2426h. I would need to test one and I think there are more logical candidates to consider first.

You have to consider the horns that were being used for those measurements....in other words, they aren't showing a plane wave response. I bet they might provide one if asked though.

[DrWho]

Q- has any one of you run DBT for time alinegment or just theoryiezing???? I'm am very skeptacil

THERE IS NO TIME ALINGEMENT IN CONCERT!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

I've done several DBT for time alignment and have never had a problem identifying the difference...even with cascaded 48dB/octave filters (so a true 96dB/octave). My lab partners thought I was crazy so they started trying to catch me off guard by playing the same thing twice or removing it the next day without me knowing while we were working on other stuff. I gotta say it's big enough that I was never faked out by their attempts, but it's small enough that I was surprised that I was able to hear it. It really comes down to how well you know your source material.

In fact, there have been several times when I tune it by ear because I don't have enough time to bust out the measurement gear and then when I come back later and measure it, I'm usually within about +/-10 degrees of phase at the xover frequency.

I've found that recordings with a good thwacky snare drum sound are the easiest to hear the difference with xovers in the 400-2kHz region. When you have your bandwidths misaligned in time, sharp transients get really soft and mushy sounding...this is because the sharp attack requires good phase linearity. If you're doing a tweeter (like 5-7kHz), then I would use a recording with a crisp high-hat.

I've also done experiments where you take a time-aligned system and then shift it back 360 degrees in phase because then you don't have the steady state frequency response or polar response issues. However, there is still an audible difference, even though you don't see it in the frequency response.

[Al Klappenberger]

FYI: I just got through testing the Faital Pro HF200 driver. It's a winner and looks like to will make a clean low distortion 2-way Khorn! Look at the other thread. I think it makes the B&C DCX50 and 9 KHz crossover a moot point for a time-aligned high end. It has about 1.26% harmonic distortion at 400 Hz at 105 DB SPL at 2 feet and goes clear to 20 KHz. The IM distortion is a squeak higher than the DCM50, but not by a significant amount!

Al K.

[DrWho]

It should also last longer than the ring radiator in the dcx50.

[sfogg]

DrWho,

I posted a number of different measurements on the 44xt at:

http://forums.klipsch.com/forums/t/84288.aspx?PageIndex=4

Shawn

[DrWho]

DrWho,

I posted a number of different measurements on the 44xt at:

http://forums.klipsch.com/forums/t/84288.aspx?PageIndex=4

Shawn

Maybe I missed it, but I was looking for a total system response. I was just wondering where you decided to xover and where you ended up on the voicing relative to the polars and all that, or did you just leave it with the ruler flat on-axis response?

[sfogg]

Mike,

I don't think I posted all of that, not sure I can find all the old measurements of the system. I ended up crossing at 500hz, from talking with Roy between 500-600hz the LaScala bass bin has about 90 degrees coverage so that is a good area to match the coverage to the K510.The K510 starts unloading below 500hz so it wouldn't have gone lower without EQ.

I did knock down the big peak in the LaScala (about 10dB around 170hz as I recall) but didn't completely flatten that out as I preferred the sound with a couple of dB lift in there. I cross to the subs at 70hz.

For the K510/44xt what you see in those graphs is how I left it and have been very happy with it. Might not work keeping it that flat in all rooms though, my room is extensively treated and very dead. I can tweak it with the bass,treble and tilt control on my Lex. if desired but almost never touch those controls as I just don't see the need.

Shawn

[Johnny's Jill]

. I think it makes the B&C DCX50 and 9 KHz crossover a moot point for a time-aligned high end. It has about 1.26% harmonic distortion at 400 Hz at 105 DB SPL at 2 feet and goes clear to 20 KHz. The IM distortion is a squeak higher than the DCM50, but not by a significant amount!

If this is the case, what crossover would you reccomend for use with the faital pro? Assuming i chose to go 2way on a lascala.

Craig 73

[Al Klappenberger]

With a LaScala there is no need to cross over down at 400 Hz. You can raise it to 500 Hz and use all sort of combinations. The minimum I would use would be my AP12-500. It will operate 2-way.

Al k.