Fast or Slow Bass - article and discussion

[D-MAN]

Here is an article that pretty much lines up with my experience.

I have a ES crossover with a phase angle differential that has a profound effect when the high frequency units are hooked up with an out-of-phase polarity (not as easy to determine as one might reasonably expect!). I've experimented with it both ways to find out what the effects are, and this article is pretty much spot-on.

Connected out-of-phase, the top end has the effect much like TUBES. Smooth, liquid midrange, drops out the bass to a degree - did I mention "smooth"? The soundstage becomes deeper and possibly more-well defined, but the bass is slightly lacking and somewhat "softened" and definitely "slower" compared to the other polarity setting.

The high freq. units and high-freq crossover are out of phase (but evidently not all the way), but the blend is still, well, sort of "nice". I haven't got an explanation other than it is not actually 180% out-of-phase which is easily detectable by ear. It is some other value, so it is NOT as readily noticable to the ear. Sort of weird.

Wired "correctly", the bass comes alive, but the soundstage changes. The bass is the main clue for me, but in reality, I could actually listen either way.

Here's the article: http://www.soundstage.com/maxdb/maxdb061999.htm

Having had this experience, I wonder how many commercially available speakers actually make some use of this technique on purpose or by accident.

DM

[BS Button]

Dman,

Can't really speak to this subject much, but I can tell you that I've just found out what amplifier damping is all about, first hand. A new (used) MX1 Yamaha takes control of your woofers...no rumble, just plain "SLAM" STOP, SLAM, STOP.

[Duke Spinner]

I wonder how many commercially available speakers actually make some use of this technique on purpose or by accident.

DM

Altec 604 comes to mind

[JBryan]

"...Connected out-of-phase... The soundstage becomes deeper and possibly more-well defined..."

An extreme slope xover (What is considered 'extreme'?) may give you more "play" with the phase but in my experience, with the signal "in-phase" the image is spot-on in the soundstage. As the phase drifts, the image becomes less defined.

Maybe you are hearing the slightly diffuse image and perceiving the signal as "deeper" and "smooth".

Just a thought.

[D-MAN]

You would think so, I agree.

But the imaging is still there because both high-ends (right and left speakers) are "in phase" with each other from the 600Hz point on up - the high ends are a bit out of phase with the woofers and that is only at the crossover point more-or-less...

This means that the soundstage is intact in its presentation.

The woofer is cancelled at the crossover point (being an ES at 600Hz) a little, but not all, in other words.

It's a rather strange experience, but nothing is noticably "wrong" with the presentation other than the bass is "slow" and is rendered "softer" in volume somewhat. This effect is quizical to me in that there is no reason for any cancellation of low frequencies to be caused by an out-of-phase top-end (mid and high freqs) to cancel out any of the low frequencies, so it has to be an electrical issue in the ESN crossover, as far as I can tell, but I'm guessing here.

DM

[nuclearay]

So many adjectives to describe bass! Most are beyond me. When I was setting my system up I think I was shooting for what sounded the most "natural" to my ears. I know what a piano, upright double bass, kickdrum, bassoon, bass guitar, etc all sound like in person... It took a lot of twiddling of the knobs and such but I finally I found a setting that was very pleasing -- I can close my eyes and I am "there" with fantastic soundstage and sound reproduction, and especially visceral in the low end.

What I find interesting about the article is the critical adjustments of phase alignment. For my room, this was ironically non-critical and I attribute that to the dimensions, and placement of speakers and furniture.

[jpm]

I set things up the same way. As real sounding as possible. I wouldn't know where to start with a spl meter or the like, nor am I remotely interested. The way things sound to my ear is important to me though.

[PrestonTom]

You would think so, I agree.

But the imaging is still there because both high-ends (right and left speakers) are "in phase" with each other from the 600Hz point on up - the high ends are a bit out of phase with the woofers and that is only at the crossover point more-or-less...

This means that the soundstage is intact in its presentation.

The woofer is cancelled at the crossover point (being an ES at 600Hz) a little, but not all, in other words.

It's a rather strange experience, but nothing is noticably "wrong" with the presentation other than the bass is "slow" and is rendered "softer" in volume somewhat. This effect is quizical to me in that there is no reason for any cancellation of low frequencies to be caused by an out-of-phase top-end (mid and high freqs) to cancel out any of the low frequencies, so it has to be an electrical issue in the ESN crossover, as far as I can tell, but I'm guessing here.

DM

DM: The effect is quizzical, but lets look at a couple of things.

First, there is only cancellation when you are at the midline. In other locations the waveforms may be adding or subtracting incompletely.

Second, you will only get a real cancellation if the waveforms are in opposite phase AND the waveforms are at a comparable level. You may be getting some room modes that are differentially changing the effective level of the two waveforms. If so the they may be in opposite phase but not equal in level. This, of course, would also give an incomplete canceling. Actually you may be (in a sense) altering the location of the room modes also.

To the degree that the crossover slopes are steep, one would think that there would be too little overlap to make this phenomenon very audible. I suspect it would not happen in a room that had a great deal of bass absorption. But that is just a guess.

Good Luck,

-Tom

[DrWho]

Phase is only an issue when you have multiple drivers playing the same frequency - so it's only an issue in the crossover region. For the on-axis response, changing the phase of one driver relative to the other just acts like a simple EQ, variable from 0dB of boost down to -infinity.

However, you must also consider the off-axis response where you start to introduce time-delays due to propogation length differences, the end result being comb-filtering. And as we all know, this off-axis sound bounces around the room and eventually makes it to the listener - slightly delayed from the original signal. The early reflections will get summed in our ears as being part of the same sound and change the timbre of what we're listening to. If we had no comb-filtering, this change in timbre could be described as part of the spatial cues of our room - giving everything you listen to a signature sound. But since we have comb-filtering, we don't get the same spatial cues that we expect (based on other sounds made in the room) and we end up with something percieved as blurred because we aren't hearing the spatial cues that we expect. But after a while our ears become accustomed and we block out those timbre differences.

But everything I've just talked about occurs only in the crossover region. So the quick conclusion would be to go with as steep of a crossover as you can find so as to minimize the region in which these effects can occur. But just blindly doing so ignores other attributes of obtaining a smooth transition between drivers. And again it comes back to the off-axis response. Typically, the lower frequency transducer starts to exhibit a narrowing polar response in the higher frequencies. Likewise, the higher frequency transducer start to exhibit a much wider polar response in the lower frequencies. For a practical example, you might be down to a 30 x 30 degree horizontal dispersion with a 15" driver at 500Hz, whereas a MF horn would be closer to 100 x 45. If we were able to construct a very very very steep filter exactly at 500Hz, we could easily obtain a flat on-axis response. However, our off-axis response will make it seem like the woofer is much much too quiet - so when the off-axis response arrives at our ears, we again are confused by wrong spatial cues and it will seem as if there was a reduction in the 400-500Hz range...all the way down until the polar response of the woofer widens up again. Many people will find this pleasing and percieve a reduction of those frequencies as an increase in clarity. Heck, the Bose acoustimass systems don't have any reproduction at 300Hz whatsoever and most people don't notice. So by choosing the slope of the crossover, you are directly changing the off-axis response. And again this is something that only occurs in the crossover region.

So in an ideal scenario we'll use driver combinations that allow steep crossovers that also obtain flat power responses. But since we don't have ideal listening environments, the new ideal would be to custom tune the crossover regions so as to better integrate the speaker into the room. And this is something that I know klipsch does - the only problem is they have to target the average listening room (which includes the display floors too). As you can imagine, compromises quickly come into play so as to reach a larger audience, and this is also why speaker positioning is so important.

And don't forget that different crossovers introduce different changes in the phase and impedance, which just further compound the issue. And then there is the wonderful time-delay issue.

But I agree entirely that the integration of the lows to the mids has the greatest influence on the percieved quality of the bass.