"Absolute Phase" and biamping 4343

I know, discussions on phase correctness drive me insane also. A contentious topic but I'm just curious if what I'm experiencing here is just placebo or not...

Ok so I'm biamping my 4343's with a 12db slope (Linkwitz Reilly I think) which I believe calls for the speaker wires for the high and low amps to be configured out of phase.

Now I originally experimented (months ago) by keeping the high amp wiring to the speakers unchanged and swapped the phase of the low freq amp only. So I tested Setup 1 vs Setup 2 (see below). I actually found that what had the best low end was the low freq amp in phase with the high freq amp (ie Setup 2 below). I kept it that way and have been listening to it for months like that.

Then I read about "Absolute phase" recently https://www.stereophile.com/content/...act-or-fallacy and decided to experiment further.

So just now I tried this: I kept the low freq amp unchanged and reversed the phase of the high frequency amp. So now the Low and High frequency amps are out of phase ( ie Setup 3 Below). Voila this sounded better than what I've been listening to for the last month (ie Setup 2). So to summarize this is what I tried:


Setup 1 (Worst, weakest bass)

- High and Low Freq amps out of phase.
- High freq amp black terminal of amp to black terminal of high freq on speakers
- Low Freq amp red terminal to black terminal of woofers.

Setup 2 (2nd best, better bass than Setup 1)

- High and Low Freq amps in phase.
- High freq amp black terminal of amp to black terminal of high freq on speakers
- Low Freq amp black terminal to black terminal of woofers.

Setup 3 (Best!, best bass performance)

- High and Low Freq amps out of phase.
- High freq amp Red terminal of amp to black terminal of high freq on speakers
- Low Freq amp black terminal to black terminal of woofers.

If we believe "Absolute Phase" this is my conclusion: Setup 1 sounded bad because absolute phase of the woofers was wrong, even though high and low amps were out of phase as they should be. Setup 2 sounded better as the woofers were now in "correct" absolute phase and this had a bigger net positive effect then having the high and low amps incorrectly "in phase" relative to each other. Setup 3 sounded the best because both absolute phase of the woofers was "correct" and a having the high and low amps "out of phase" was also correct.

Thoughts?

I do not think "absolute phase" is what you should worry about. I think you should worry about phase of the individual woofers/drivers not being in correct phase relative each other, as the crossover was designed to be.
You can try to play the signal of the crossover frequencies (if you know them) with a signal generator, and flip phase of the individual elements, and see where you get cancelation. That is the wrong phase-setting.

Quote:

---

Originally Posted by Goldjazz

I know, discussions on phase correctness drive me insane also. A contentious topic but I'm just curious if what I'm experiencing here is just placebo or not...

Ok so I'm biamping my 4343's with a 12db slope (Linkwitz Reilly I think) which I believe calls for the speaker wires for the high and low amps to be configured out of phase.

Now I originally experimented (months ago) by keeping the high amp wiring to the speakers unchanged and swapped the phase of the low freq amp only. So I tested Setup 1 vs Setup 2 (see below). I actually found that what had the best low end was the low freq amp in phase with the high freq amp (ie Setup 2 below). I kept it that way and have been listening to it for months like that.

Then I read about "Absolute phase" recently https://www.stereophile.com/content/...act-or-fallacy and decided to experiment further.

.............Thoughts?

---

Hi Goldjazz,

may in the mentioned
"...... The results were almost perfectly consistent with what we felt we knew about those transducers: The better they were, the less difference was noted when polarity was reversed! ........"

there is a suggestion about something of the used drivers, or DC offsets of the used amps, or active crossover characteristics.
I would suggest to measure THD of the active elements that are used for the tests (CD or computer, active crossover, power amps with added some capacitor or inductive loading, or the speakers) and possible DC offsets.

Not to mention that in multi-ways speaker system as the mentioned 4343, distances between the drivers would make different phase shifts relative to the listener...

regards
ivica

Quote:

---

Originally Posted by bubbleboy76

I do not think "absolute phase" is what you should worry about. I think you should worry about phase of the individual woofers/drivers not being in correct phase relative each other, as the crossover was designed to be.
You can try to play the signal of the crossover frequencies (if you know them) with a signal generator, and flip phase of the individual elements, and see where you get cancelation. That is the wrong phase-setting.

---

Do you mean flip the phase for the other individual drivers eg midbass, high and ultra high? I was flipping the phase of this group of drivers as they are connected by a passive network and difficult to get to inside the speaker. I was also flipping the phase of the individual woofer.

When I wired the 2420 on the longer L300 horn, and 2405 to the 2235, I have to run the Amp Black to the Speaker Red or the Yamaha reports them as out of phase with the system. Makes no difference to the Yamaha's automatic set-up routine how I wire the horn and tweet.

When I used the Behringer 61-band RTA to set the horn and tweeter phase, I seemed to get the smoothest curve in the room with the horn and tweet wired with the Driver Red to the Crossover Black and this is how I've been running them for years. I find that the horn and tweeter always seem to phase best together, but changes to the tweeter phase are very hard to hear regardless.

But recently when comparing speakers, to first recognize the deficiency, and after many hours of doing A/B/X comparisons with the wiring, (I can do that ;) ), I came to the conclusion that even though, the Behringer may have looked to be smoother across the band wired as it was, it sounded better the other way with the major advantage being intelligability and tonality of the vocals. I find this to be one of the hardest criteria for speakers to reproduce. So I changed the wiring for the better, and need to get into my center channel to do that one too.

BTW, I recently also found a way to "lift the veil" from the 2405 making it "sparkle" as it can and also adding intelligability. But that's for another post.

Quote:

---

Originally Posted by Goldjazz

Thoughts?

---

Here are a few random thoughts.

What you were experimenting with is not "absolute phase", but rather affecting the performance at the crossover frequencies due to changing the relative phase between drivers.

By using 12 dB slopes you are introducing a 180° phase shift at the crossover point. A Linkwitz Riley crossover is 24 dB/octave and introduces a 360° shift. Then it gets more complicated as various horns add there own phase shift. There really is no perfect solution... especially with multi-driver systems.

Experimenting with the relative phase of the various drivers is definitely useful... especially once you understand what you are doing and how it affects the system's sound.


Widget

Quote:

---

Originally Posted by Mr. Widget

Here are a few random thoughts.What you were experimenting with is not "absolute phase", but rather affecting the performance at the crossover frequencies due to changing the relative phase between drivers. By using 12 dB slopes you are introducing a 180° phase shift at the crossover point. A Linkwitz Riley crossover is 24 dB/octave and introduces a 360° shift. Then it gets more complicated as various horns add there own phase shift. There really is no perfect solution... especially with multi-driver systems.Experimenting with the relative phase of the various drivers is definitely useful... especially once you understand what you are doing and how it affects the system's sound.Widget

---

Yeah well thats what I initally thought too, but the relative phase of the woofer to the internal passive crossover containing the other three drivers is opposite in setup 1 and also opposite setup 3 yet setup 3 sounds better. cheers.

Quote:

---

Originally Posted by toddalin

When I wired the 2420 on the longer L300 horn, and 2405 to the 2235, I have to run the Amp Black to the Speaker Red or the Yamaha reports them as out of phase with the system. Makes no difference to the Yamaha's automatic set-up routine how I wire the horn and tweet.

When I used the Behringer 61-band RTA to set the horn and tweeter phase, I seemed to get the smoothest curve in the room with the horn and tweet wired with the Driver Red to the Crossover Black and this is how I've been running them for years. I find that the horn and tweeter always seem to phase best together, but changes to the tweeter phase are very hard to hear regardless.

But recently when comparing speakers, to first recognize the deficiency, and after many hours of doing A/B/X comparisons with the wiring, (I can do that ;) ), I came to the conclusion that even though, the Behringer may have looked to be smoother across the band wired as it was, it sounded better the other way with the major advantage being intelligability and tonality of the vocals. I find this to be one of the hardest criteria for speakers to reproduce. So I changed the wiring for the better, and need to get into my center channel to do that one too.

BTW, I recently also found a way to "lift the veil" from the 2405 making it "sparkle" as it can and also adding intelligability. But that's for another post.

---

Very intersting. Sounds like you like to tweak your system like me. Ive found some good benefits from tweaking and trying different stuff. yeah post the findings on the 2405

Quote:

---

Originally Posted by toddalin

When I wired the 2420 on the longer L300 horn, and 2405 to the 2235, I have to run the Amp Black to the Speaker Red or the Yamaha reports them as out of phase with the system. Makes no difference to the Yamaha's automatic set-up routine how I wire the horn and tweet.

When I used the Behringer 61-band RTA to set the horn and tweeter phase, I seemed to get the smoothest curve in the room with the horn and tweet wired with the Driver Red to the Crossover Black and this is how I've been running them for years. I find that the horn and tweeter always seem to phase best together, but changes to the tweeter phase are very hard to hear regardless.

But recently when comparing speakers, to first recognize the deficiency, and after many hours of doing A/B/X comparisons with the wiring, (I can do that ;) ), I came to the conclusion that even though, the Behringer may have looked to be smoother across the band wired as it was, it sounded better the other way with the major advantage being intelligability and tonality of the vocals. I find this to be one of the hardest criteria for speakers to reproduce. So I changed the wiring for the better, and need to get into my center channel to do that one too.

BTW, I recently also found a way to "lift the veil" from the 2405 making it "sparkle" as it can and also adding intelligability. But that's for another post.

---

Hi toddalin,

length of the 2312 and 2420 internal horn would suggest of 1ms delay, and 2231/2235 driver in the L300 box including the network would introduce almost the same delay so theoretically VHF driver polarity inversion would be expected, but 2405 , due to the large distance (over 30cm = 12inch) cannot be compensated with any polarity , only proper delay (1ms for the beginning) would help in order to reduce comb filter effects.
From that point, 24dB LR network would suggest that almost no polarity change between bass and VHF driver would be expected.
May be square wave (low frequency) test signal and proper microphone and some 'real time trace software presenting'
( https://trueaudio.com/rta_abt1.htm )
can give much more precise ideas in such "phase adjustment experiments"



......"I recently also found a way to "lift the veil" from the 2405 making it "sparkle" as it can .".... ???????
I have no idea what kind of veil can be connected to the 2405 UHF driver (neglecting 'comb filer effects' previously mentioned)
regards
ivica

Quote:

---

Originally Posted by ivica

......"I recently also found a way to "lift the veil" from the 2405 making it "sparkle" as it can .".... ???????
I have no idea what kind of veil can be connected to the 2405 UHF driver (neglecting 'comb filer effects' previously mentioned)
regards
ivica

---

"But recently when comparing speakers, to first recognize the deficiency"

I would not loose sleep over it.

A lot of preamps and power amps invert the absolute phases anyway.

As the article indicates the loudspeaker needs to distinguish between linear and non linear wave forms.

Most loudspeakers even full range single drivers exhibit varying phase shift particularly at frequency extremes. This means the driver can't transmit the identical phase of the incoming signal.

Only a few loudspeakers will actually pass a square wave.

The 4343 is a multi way system, the biamp crossover point inverts the phase. The other passive crossover points have drivers in phase.

However each crossover point produces phase shift as do the varying distance of the voice coils.

It's quite technical and if you went thought it with a tooth comb you would get a head ache.

Now Jbl aren't completely stupid and in one of their papers on studio monitors they talk about what studio monitors are right for you?

The JBL/ Urei differences are discussed and Jbl point out that the Urie midrange projects a more forward sound stage (check the article) compared to their other multi way monitors.

The Urei was a time aligned monitor made famous by Ed Long.

https://m.youtube.com/watch?v=9utuBZL3cm8

Urei claimed their system produced a square wave.

The secret was the delay line in the bass section of the crossover that shifted the acoustic centre of the woofer back to align with the HF driver. It's basically a bunch of Bessel filters. Some people say that subjectively those filters sucked the life out of those systems.

I would be more inclined to listen to an old Sheffield lab pressing with some Stax headphones and have a switch to reverse the absolute phase.

Because those direct discs had very little intervention (relative) to most recordings the odds of detecting a change in absolute phase are much higher imho.

Quote:

---

Originally Posted by ivica

......"I recently also found a way to "lift the veil" from the 2405 making it "sparkle" as it can .".... ???????
I have no idea what kind of veil can be connected to the 2405 UHF driver (neglecting 'comb filer effects' previously mentioned)
regards
ivica

---

What I've done is to employ a 2nd order filter with a crossover frequency of ~7,300 Hz based on the N7000/N8000, (1.5 mfd/0.5mH/voltage divider/L-pad/20 ohms across tweeter) while simultaneously applying a 1st order filter with a crossover frequency of about 15kHz (choke/0.57 mfd/voltage divider/positive speaker terminal). But, they are attenuated at different rates such that the L-pad attenuates the 2nd order more than the first order as you turn it down. It still goes to "zero." Because of the choke, this is actually a band-pass filter. I don't know the value of the choke, but it would be very small and seems to my ear to make the system sound a bit smoother. (Looks to be 16 turns of 12 gauge on ~1/4" ferrite bobbin.)

With the L-pad all the way up, the 1st order adds nothing to the peak (at ~12,500 Hz), but flattens that area of the band, especially above the peak. As you turn down the L-pad, you attenuate the second order more quickly than the 1st order, retaining those nice crisp highs as you reduce the overall brightness.

The first order never even goes through those nasty L-pad windings before reaching the tweeter.

Quote:

---

Originally Posted by toddalin

What I've done is to employ a 2nd order filter with a crossover frequency of ~7,300 Hz based on the N7000/N8000, (1.5 mfd/0.5mH/voltage divider/L-pad/20 ohms across tweeter) while simultaneously applying a 1st order filter with a crossover frequency of about 15kHz (0.57 mfd/voltage divider/positive speaker terminal). But, they are attenuated at different rates such that the L-pad attenuates the 2nd order more than the first order as you turn it down. It still goes to "zero."

With the L-pad all the way up, the 1st order adds nothing to the peak (at ~12,500 Hz), but flattens that area of the band, especially above the peak. As you turn down the L-pad, you attenuate the second order more quickly than the 1st order, retaining those nice crisp highs as you reduce the overall brightness.

The first order never even goes through those nasty L-pad windings before reaching the tweeter.

---

Hi toddalin,

interesting solution, can you sketch such solution, to be more understandable.

regards
ivica

Quote:

---

Originally Posted by ivica

Hi toddalin,

interesting solution, can you sketch such solution, to be more understandable.

regards
ivica

---

OK, I've created a new thread.