Test track recommendations to reveal the effect of bass driver polarity

[ToTo Man]

I have inverted the polarity of the bass driver in one of my Celestion Ditton 66s and I think I prefer the new tonal balance with woofer polarity inverted; the midrange now sounds slightly more forward with less upper-bass bleed. I have posted measurements here.

I don't yet hear any downsides to the inversion, apart from perhaps the "outside of sweet spot" tonal balance no longer being quite so even (I've not yet measured this so it's purely a hunch), but this doesn't really concern me as I always listen in the sweet spot.

I've been trying to detect changes in transient response between the two wiring configurations but am struggling to hear a difference. I'd therefore be grateful for a recommendation of a test track(s) that makes it most obvious for an unskilled listener to hear the difference between sucking and blowing (...and no links to video content, please! )

 

[ToTo Man]

C'mon, don't pretend I'm the only one interested in hearing the effect of polarity?! This is the Audio (Nervosa) section of the forum, isn't it?

 

[sq225917]

What's the xo point between mid and bass find something busy in that region. One will sound right, the other will sound shit, assuming the xos were half decent in the first place.

 

[Rug Doc]

When doing polarity checks etc I simply use a test tone generator...

Usually on my iPhone or computer and streamed so I can be in the seated position and get the central image focused.

BUT.

Before you get into reversing the polarity inside the cabinet I’d personally use the master setup speaker setup to get the speakers in the right place first (if you haven’t already or aren’t faced with domestic restrictions!)

https://forums.stevehoffman.tv/threads/sumiko-speaker-setup-protocol.187416/

This has always been one of THE most important things in setting up any speakers in my room and the difference it makes is huge!!

Have fun. Setting up a system and getting it to to image well and play bass properly is one of my favourite pastimes.

 

[Tony L]

How does the 66’s ABR work with regards to phase? I have a feeling this may be blurring any clear difference between polarity as chances are it won’t be moving even remotely in phase across its whole range. The next aspect is obviously time-alignment. With a multi-way speaker as tall and widely-spaced as the Ditton 66 you will inevitably be getting some comb effects etc.

Given both of the above factors I’d be interested if you still preferred the out-of-phase bass with the speakers properly floor-standing rather than on plinths. Any time-alignment, Allison Effect etc that was designed-in by Celestion will have been done expecting that usage context. They expected the 44 to sit on a stand, but not the 66.

PS As for source material I’d pick a variety of jazz, classical and electronica, stuff where there is a lot of movement in the bass. You are obviously looking for even runs with no notes being prominent or missing (unless intended, hence use a wide selection). One of my favourites is actually Miles Davis Bitches Brew as it is a real murky mess of a thing and if you can follow the basses (there are more than one) on say Spanish Key then you are likely doing ok!

 

[ToTo Man]

The LF/MF crossover frequency is approximately 500Hz, the woofer's slope is 24dB/oct and the mid's slope is 12dB/oct but the mid is a 2-inch dome and has a natural roll-off below 500Hz which makes the slope steeper.

My main complaint of the 66 is a bloated sound between 300Hz and 500Hz and, on this particular pair, the midrange has a dip in its response at 750Hz that my other pair of 66s don't have, which compounds the effect of the 300Hz-500Hz bloat.

Replacing the T2169 bass drivers with slightly less efficient T1600 drivers helped to reduce the upper-bass / lower mid-range to the same level as my other pair of 66s with T1600 bass drivers, but the dip at 750Hz means this pair still sound slightly warmer / fuzzier / laid-back than my other pair.

Inverting the polarity on the bass driver cleans up the sound and gives a flatter frequency response at the listening position. In fact the response through the midrange is now incredibly similar to that of my Tannoy Edinburghs. I'll post graphs of the response at the listening position shortly.

I think it's safe to say that 66s aren't as phase coherent as more conventional designs, and yes the spacing of the drivers no doubt creates more time-alignment issues. You make an interesting point about the 66 being designed to operate as a true floorstander, Tony. I hadn't considered the impact that elevating them would have on the low frequency phasing / time alignment. Sadly removal of the plinths isn't a viable option; I listen too close (2.25m) and sit too high (1.05m) to place them on the floor; doing so results in audible cancellations around the MF/HF crossover frequency.

 

[cooky1257]

From a distance..by reversing the polarity of one bass driver you will in one 66 be creating a deep cancellation around the bass/mid crossover region-the other will be as is, also you could well be going some way to eq out(by phase cancellation) the mid bass hump you dislike so much. What you will lose is the tight central bass image for a rather a wishy/phasey and unfocused but spacey one. If you were listening in mono it'd probably be fine but for me at least, I cant live with the phasey effect when I try it on my active JBL's, it's very inconsistent too and is very track dependent-some good/interesting others 'where's my bass gone?"....

 

[ToTo Man]

From a distance..by reversing the polarity of one bass driver you will in one 66 be creating a deep cancellation around the bass/mid crossover region-the other will be as is, also you could well be going some way to eq out(by phase cancellation) the mid bass hump you dislike so much. What you will lose is the tight central bass image for a rather a wishy/phasey and unfocused but spacey one. If you were listening in mono it'd probably be fine but for me at least, I cant live with the phasey effect when I try it on my active JBL's, it's very inconsistent too and is very track dependent-some good/interesting others 'where's my bass gone?"....

Just to be clear, I'm not listening to the speakers as a stereo pair with one woofer reversed (that does indeed result in a "where's my bass gone situation"!), but rather I have been evaluating each loudspeaker individually (by creating mono versions of familiar test material) and repeatedly panning left and right to decide which loudspeaker wiring I preferred. This evening I "completed the job" by reversing the polarity of the woofer in the other speaker, so now I am able to listen to both loudspeakers as a stereo pair with both woofers wired in reverse polarity. I'm quite captivated by the new presentation, the imaging and dynamics are much more precise now to my ears. I don't like using cliches but I now feel performers are "in the room" with me instead of being filtered by a smokescreen. The before and after measurements do indeed show a wide null around the crossover region where there used to be a broad peak, but my ears don't seem to notice an obvious absence of output in this region (perhaps this is because my Tannoys exhibit a broad dip across the same frequency range and my ears have become accustomed to it). Out of interest, why would the bass sound phasey on some tracks and not on others, assuming both woofers are wired with the same (inverted) polarity.

 

[Sue Pertwee-Tyr]

This evening I "completed the job" by reversing the polarity of the woofer in the other speaker, so now I am able to listen to both loudspeakers as a stereo pair with both woofers wired in reverse polarity. I'm quite captivated by the new presentation, the imaging and dynamics are much more precise now to my ears. I don't like using cliches but I now feel performers are "in the room" with me instead of being filtered by a smokescreen.

Is it possible that the speakers were originally wired ‘incorrectly’ and you’ve inadvertently fixed them? What you describe does sound more like a properly functioning speaker.

 

[SteveG]

Is it possible that the speakers were originally wired ‘incorrectly’ and you’ve inadvertently fixed them? What you describe does sound more like a properly functioning speaker.

Or something else in your system is badly broken or badly performing and this is making up for it? Or there are room issues which a drastic reduction in bass is helping with?

 

[Sue Pertwee-Tyr]

Or something else in your system is badly broken or badly performing and this is making up for it? Or there are room issues which a drastic reduction in bass is helping with?

It’s not a reduction in bass due to cancellation, as both bass drivers are in phase with each other though.

 

[ToTo Man]

Regarding the ABR, it is a completely passive device that moves in reaction to the woofer's movement (and thus relies on an air-tight cabinet) and so, AIUI, when the woofer moves outward the ABR moves inward and vice versa.

I think there are two issues here, the first being the potentially adverse affects of inverting the woofer with respect to its integration with the midrange unit, and the second being the potentially adverse effects of inverting the woofer with respect to preservation of absolute polarity from the amplifier output.

From the comments above, I get the impression that the latter is a more trivial issue, since not all listeners can hear the difference when the absolute polarity of a system is inverted. Also, I believe some loudspeaker manufacturers consciously invert the polarity of the woofer (e.g. Tannoy) to no ill effect apparently. There are however some listeners who swear about the importance of maintaining absolute polarity and even go as far as tagging every album they own with a polarity marker(!).

 

[Tony L]

Also, I believe some loudspeaker manufacturers consciously invert the polarity of the woofer (e.g. Tannoy) to no ill effect apparently.

That is an attempt to time-align as the compression driver is substantially behind the bass driver. Same goes in reverse for many flat-baffle multi-way speakers where the tweeter is in effect a fair distance in front of the bass driver, especially when one considers a fair bit of mid-band comes from the inner area of the cone or dust cap. I’d have expected a company as credible as Celestion to have done the math in all areas including mechanically defining the frequency the ABR rolls-off (it just won’t be able to keep up with the bass unit above a certain frequency due to the natural air-spring in the box). I’d actually expect the ABR to be out of the mix somewhere well below 100Hz as I don’t see it could move any faster given the compliance of the air-mass.

 

[cooky1257]

Out of interest, why would the bass sound phasey on some tracks and not on others, assuming both woofers are wired with the same (inverted) polarity.

It wouldn't as they are in phase with each other, I'd misunderstood your original post(which you cleared up).
The xover circuit has the bass and mids in positive polarity with a reversed tweeter. What you may have done is tilted the main lobe in a more favourable orientation to your listening position.
The Tannoy comp driver polarity is inverted as it is facing( and with a +ve input would be moving) in the opposite direction from the bass driver. The physical displacement and the crossover slopes used go some way to time align the drivers but they each have virtual 'sources' rather than just the mid point of their respective voice coils.. The drivers sum correctly at crossover.
Tannoy used to follow the -ve absolute polarity convention in their K series(SRM etc)studio monitors.

 

[ToTo Man]

I have now uploaded farfield measurements showing the effect of the woofer polarity reversal.

 

[ToTo Man]

Now that I've acclimatised to my 66's newly-revealed sonic signature and still very much enjoying it, I decided to add my subs back into the mix. I had a brief moment of panic thinking they had developed a fault in their fortnight of inactivity, as their output was much quieter than I remember. I checked all settings on the back in case any knobs had been turned accidentally but no, they were set as they were before...

It then dawned on me that by reversing the polarity of the 66 woofers, the phase of the passive ABRs had also reversed and were effectively acting as a black hole and consuming my subwoofers' output! Quickly swapping the speaker leads at the rear terminals of the 66 (thereby restoring the woofers and ABRs to original polarity but making the mids and tweeters inverted polarity) has restored the 15dB suckout in my subwoofers' output around 22Hz! I suspect I would have achieved the same result by turning the phase knob on the subwoofers to 180 degrees but I was curious to see the result of doing it at the 66 terminals instead. Doing so has had no effect on the frequency response above 40Hz, which is good to know.

 

[mega lord]

Where did you get to with this if I may ask ?

I have experience with reversing the polarity of woofers but in a very different situation. I can echo your thoughts from the opening post though the clarity of bass is much improved with my set up and room interaction.

I have doped 12" isobaric (no K in isobaric) woofers in a clamshell configuration with outer magnet facing up and no electrical filtering. A separate enclosure sits on top of this housing two 5.25" mid bass doped drivers (no electrical filters) with one firing up (cone out) and one firing down (magnet out and touching the 12 inchers magnet) along with five tweeters firing forwards, upwards, backwards and to both sides.

Occasionally I have messed about with polarity along my diy adventure and always dismissed it as not sounding quite right. But I have just tried it again recently and I am thoroughly enjoying listening to my 12" isobaric pairs with reversed polarity. I will have to analyse what changes I have made since I last tried it.

I have effectively three drivers all outputting bass frequencies although the isobaric 12 inchers go lower than the two 5.25". There must be some cancellation going on but as I am not the measuring kind I have to rely on ears, assumptions and experiments. The tightening in the bass is mostly evident on bass heavy music such as Portishead Dummy, Air Moon Safari, Gorillaz Demon Days, Morcheeba Big Calm, The Comet is coming, Jamiroquai etc etc. At times especially with certain Portishead tracks I could rattle things in the room where as now it doesn't happen.

I need to play a much greater range of music and switch back at some time to see if I am imaging things.

 

[ToTo Man]

Where did you get to with this if I may ask ?

I have experience with reversing the polarity of woofers but in a very different situation. I can echo your thoughts from the opening post though the clarity of bass is much improved with my set up and room interaction.

I have doped 12" isobaric (no K in isobaric) woofers in a clamshell configuration with outer magnet facing up and no electrical filtering. A separate enclosure sits on top of this housing two 5.25" mid bass doped drivers (no electrical filters) with one firing up (cone out) and one firing down (magnet out and touching the 12 inchers magnet) along with five tweeters firing forwards, upwards, backwards and to both sides.

Occasionally I have messed about with polarity along my diy adventure and always dismissed it as not sounding quite right. But I have just tried it again recently and I am thoroughly enjoying listening to my 12" isobaric pairs with reversed polarity. I will have to analyse what changes I have made since I last tried it.

I have effectively three drivers all outputting bass frequencies although the isobaric 12 inchers go lower than the two 5.25". There must be some cancellation going on but as I am not the measuring kind I have to rely on ears, assumptions and experiments. The tightening in the bass is mostly evident on bass heavy music such as Portishead Dummy, Air Moon Safari, Gorillaz Demon Days, Morcheeba Big Calm, The Comet is coming, Jamiroquai etc etc. At times especially with certain Portishead tracks I could rattle things in the room where as now it doesn't happen.

I need to play a much greater range of music and switch back at some time to see if I am imaging things.

I still currently have one pair of Ditton 66 with the bass drivers connected in inverted polarity and another pair of 66 with the bass drivers connected in 'correct' polarity. I have however since changed my listening arrangement, and the speakers now sit on the floor instead of on 250mm stands (I now have the speakers tilted backwards to compensate for the drop in tweeter height).

In this new arrangement, the 'bloat' I previously heard in the 300Hz-600Hz region with 'correct' polarity bass drivers isn't as objectionable as it was with the speakers on the 250mm stands, and doesn't need as much EQ'ing to make it sound 'transparent'. The inverted polarity configuration by comparison now sounds somewhat dipped/thin in the upper-bass/lower midrange. The casual listener probably wouldn't notice it as the dip spans a fairly narrow range of frequencies, but it is noticeable when you A/B the two configurations side by side.

In this new floor standing arrangement, I prefer the presentation with the bass drivers in 'correct' polarity but with a broad Q parametric EQ applied to reduce output around 500Hz. If I didn't have access to EQ I'd probably still prefer the inverted arrangement.

Note that my impressions focus on the tonal aspects of the sound. Having listened to both polarities, I honestly cannot hear any difference in the way the bass 'times' with the midrange and treble, either way sounds the same to me. This may be specific to my particular speakers or perhaps I'm not as skilled in hearing timing aberrations as I am in hearing tonal aberrations. Your speaker configuration is more complex and way beyond my level of knowledge and experience(!), so I best defer to someone who knows what they're talking about to speculate about the effect of inverting the polarity of Isobarik drivers...

Normal vs inverted woofer polarity (no EQ applied):

Normal vs inverted woofer polarity (EQ applied to normal woofer polarity):

 

[Tony L]

Note that my impressions focus on the tonal aspects of the sound. Having listened to both polarities, I honestly cannot hear any difference in the way the bass 'times' with the midrange and treble, either way sounds the same to me.

That really surprises me, I’d expect to hear all manner of issues in that dip as half of any transient attack was split off in both directions, e.g. a rack tom bass guitar or whatever trying to simultaneously push and pull. I’d expect it to sound very odd on a good solo piano recording too, some notes sounding uneven and undynamic in the left-hand range. I seem very sensitive to phase though and hear errors when others don’t seem to care. By saying that there may be all manner of other phase error in the speaker so it may not be a simple 180 flip in practice.

 

[ToTo Man]

That really surprises me, I’d expect to hear all manner of issues in that dip as half of any transient attack was split off in both directions, e.g. a rack tom bass guitar or whatever trying to simultaneously push and pull. I’d expect it to sound very odd on a good solo piano recording too, some notes sounding uneven and undynamic in the left-hand range. I seem very sensitive to phase though and hear errors when others don’t seem to care. By saying that there may be all manner of other phase error in the speaker so it may not be a simple 180 flip in practice.

Perhaps I'm just not labelling what I hearing correctly. I can definitely hear a 'dip' in this area, but it's the same kind of 'dip' I hear if I were to pull an EQ slider down to create a dip. I call this a tonal aberration but maybe I should be calling it a phase aberration?