Jump to content

Subconscious Auditory Effects of Quasi-Linear Phase Loudspeakers


Chris A

Recommended Posts

38 minutes ago, Islander said:

...while the effect you describe may be difficult to describe and impossible to measure at this time, it may not always be that way.

I don't believe that I said that I couldn't measure it.  I can.  The problem is describing the subconscious/threshold effect that I experienced.  That was the part that I can't put into words. 

 

When I say threshold--I mean just that.  Once it's attained, the analogy that comes to mind is like pushing catamaran or other light dinghy into planing: everything smooths out suddenly, and the controls become very sensitive to changes such as changing direction by wiggling the tiller or sheeting in or out a click or two to adjust the driving force: all the controls suddenly have "power steering".  It can  also be described like someone just released the parking brake while driving down the road.  Once you decrease the phase growth and group delay to below a particular threshold, the effect suddenly appears and the satisfaction level in listening just goes through the roof.  I'm talking 50% or more subjective improvement in listenability for a very small decrease in group delay/overall phase growth.

 

It's not subtle--but difficult to describe.  It's also a bit startling to experience, too. 

 

Chris

  • Like 2
Link to comment
Share on other sites

On 3/14/2019 at 6:17 PM, Chris A said:

 

It's not subtle--but difficult to describe.  It's also a bit startling to experience, too

 

I substituted a word previously... o O (explain) … and omitted the 2nd half, which further touched on my belief, for which I still am unable to post... (to eliminate off topic rhetoric)  ... so I just wanted to convey this form of non-scientific explanation... explain = conceive... as in form or devise a plan or idea, but essentially born, as in invented in the sense to "explain" the unknown, but the unlikeliness is not the unusual, but where sounds evolved...so... God ---> Nature ----> Love.

Link to comment
Share on other sites

On 3/14/2019 at 4:17 PM, Chris A said:

I don't believe that I said that I couldn't measure it.  I can.  The problem is describing the subconscious/threshold effect that I experienced.  That was the part that I can't put into words. 

 

When I say threshold--I mean just that.  Once it's attained, the analogy that comes to mind is like pushing catamaran or other light dinghy into planing: everything smooths out suddenly, and the controls become very sensitive to changes such as changing direction by wiggling the tiller or sheeting in or out a click or two to adjust the driving force: all the controls suddenly have "power steering".  It can  also be described like someone just released the parking brake while driving down the road.  Once you decrease the phase growth and group delay to below a particular threshold, the effect suddenly appears and the satisfaction level in listening just goes through the roof.  I'm talking 50% or more subjective improvement in listenability for a very small decrease in group delay/overall phase growth.

 

It's not subtle--but difficult to describe.  It's also a bit startling to experience, too. 

 

Chris

 

I think I may need to read Toole's book to fully grasp what you're saying, since some of the terms used in your posts in this thread are either not fully familiar to me, or not fully understood by me, or completely unfamiliar.  This is not my field of expertise at all, but it does seem intriguing, and learning about some of these concepts could enrich my appreciation of music and how our equipment reproduces it.

 

Thanks for sharing your knowledge and experience here.

Link to comment
Share on other sites

All I'm really talking about is the time-based behavior of loudspeakers/rooms and how well they respond to impulse input signal--all the way from the source of sound in the loudspeakers, to your ears. 

 

Most performance data published on loudspeakers is in the frequency domain--SPL vs. frequency.  Phase can be plotted on the same SPL vs. frequency curve (but is rarely done by manufacturers) to give you an idea of the loudspeakers' time-based response in addition to their steady state frequency response (shown in the figure below: the top trace is SPL vs. frequency, the bottom trace is phase vs. frequency):

 

Chris A's Right Jub SPL and Phase.JPG

 

If certain frequencies produced by the loudspeaker itself arrive before others, then the loudspeaker has phase distortion. If you mathematically take the first derivative of the resulting phase vs. frequency curve from the loudspeaker, then you have the loudspeaker's group delay curve--a curve of rate of change of the phase vs. frequency:

 

Jub-TAD4002 Group Delay Using 1st Order Xover Filters and shorterned down K-402s.jpg

 

It turns out that humans can directly hear group delay distortion (above a certain threshold--with a different threshold value at different frequencies), but only if nearby early reflections in the room don't significantly interfere with the direct-arrival acoustic energy from the loudspeakers. 

 

Additionally, the loudspeakers themselves need to be able to control reflections vs. frequency off of the room's walls and nearby acoustically reflective objects.  It does this through "directivity" vs. frequency.  You can see this directivity in the following polar directivity chart of the K-402-MEH, which has consistent directivity down to slightly below 100 Hz:

 

K-402-MEH horizonal normalized sonogram.jpg

 

So to hear this effect that I'm talking about means you need three things:

  1. Control of early reflections in-room from right around the loudspeakers (within a yard or metre of the loudspeaker itself), which are usually controlled by use of absorption that can equally control mid-bass, midrange, and higher frequency reflections (i.e., broad-band absorption).  [Diffusion panel approaches usually can't handle the "broad band" requirement very well.]
  2. Loudspeakers having good/consistent polar directivity control vs. frequency for all frequencies down to ~100-200 Hz and all the way up to 20 kHz (which basically rules out loudspeakers having direct radiating drivers--especially woofers.)
  3. A way to reduce or increase the phase distortion and its linked group delay distortion of the loudspeaker.  This is pretty easy to do using a DSP crossover, but almost impossible using a passive crossover network, unless you're substituting complete crossover networks out in A-B fashion.

Most people haven't gone to the trouble to do all three areas simultaneously (and effectively).  Until you do all three, you probably won't hear the effect that I've been talking about. 

 

The "full-range" loudspeaker crowd, planar dipole loudspeaker crowd, and headphone crowd are usually chasing this effect using loudspeakers having many limitations in their output capabilities.  I've found that there are too many compromises using any of these alternative loudspeaker types to be viable for my home hi-fi listening use.  YMMV.

 

Chris

  • Like 2
Link to comment
Share on other sites

On 3/16/2019 at 9:26 PM, PaulBlossom said:

Do you think phase correction of higher order crossovers, via FIR, would accomplish the same result? 

Yes, if both the phase and group delay are basically flat through the crossover interference bands.

  • Like 1
Link to comment
Share on other sites

I forgot to add: ...assuming that the pre-rise on the step response is also held to a very small value (also referred to as "pre-ringing").

  • Like 1
Link to comment
Share on other sites

Hello Chris,

Thanks for the content and plots you posted in this thread.

 

Regarding the pre-rise of the step response seen with FIR filters usage, what is your obsevation? Do you have some best practices tips for obtaining a small pre-rise? Number of taps maybe (more or less ?, if the hardware can take it), windowing?

 

I do my filters with rePhase and clearly hear that using FIR filters the sound is not so 'clean'/'crisp' than using simple IIR filters in EqualizerAPO engine. But then again, the phase is more liniar using liner-phase FIR filters. 

 

Link to comment
Share on other sites

 

Hi...and welcome to the forum.

 

My observations come from IIR filtering using HF channel delay to reduce overall phase growth a little vs. first order...not FIR filtering. 

 

In A-B testing that I did recently, I could hear more harshness/hardness of sound of a "squashed" step response third-order Bessel vs. a conventionally set up first order that minimizes pre-rise (but lengthens phase growth a bit).  Some of this subjective difference could be from confounded sources, and I can't really help you with the sound of FIR filtering, per se, but I can help a bit with the implication that ~1 ms step response pre-rise is probably audible.

 

813599076_BMS4592NDonK-402horn(yellowBessel18vs.redfirstorder).thumb.jpg.77c4a51a7159be59ba734603bc4ccd9b.jpg

 

Chris

 

  • Like 1
Link to comment
Share on other sites

  • 2 weeks later...

Thanks Chris for the explanation, I thought you were referring to FIR pre-rise time.  

 

I am about to start fabricating a K402 clone. Thinking about what CD to buy, so this next question is: Did you try also 1st order crossover with normal 2" CD or just with the coaxial BMS that you use? 

 

Thanks!

Link to comment
Share on other sites

I haven't used first order crossover filters with anything other than the TAD TD-4002s on the Jubilees and the BMS 4592ND on the MEH.  The only consideration is where to put the crossover frequency in order to control the extreme low frequencies being audibly reproduced by the compression driver below ~350-400 Hz.  For the BMS 4592ND relative to other metal dome-type compression drivers, the crossover frequency can be lowered by 100 Hz since it is capable of being crossed at 300 Hz--according to the literature that BMS produces (and that seems to be true in listening tests).  I believe that my Jubilee crossover frequency is at 550 Hz, but the shallower slope increases the width of the polars at the crossover band and actually helps to broaden the inherently narrow polars of the Jubilee in the 300-600 Hz band due to the truncated dual-mouth bass bin "dual slit" effect by using the controlled directivity of the BMS-4592NDs on the K-402 horn to fill the lower midrange polars.  This improves lower midrange and upper mid-bass sound substantially, I've found.  The phase of the lower bandwidth of the TADs relative to the bass bin woofers seems to support this broad handover (interference) band quite nicely.

 

If there is an issue with the first-order shallow slopes and with a local rise in driver response in the stop band, I just add an extra PEQ to attenuate any driver stop band peaks.  This seems to work quite well without introducing phase shifts typical of crossover filters.  It's the flatter overall phase response that we're getting using the 1st order filters, and adding an attenuating PEQ doesn't divert the overall phase like minimum phase crossover filters do.

 

Chris

  • Like 1
  • Thanks 1
Link to comment
Share on other sites

  • 4 weeks later...
  • 2 weeks later...

Well, I learned a lot over the past two days.  Here is the new step response of the same Jubilee as posted above compared to a Danley SH-50.  The Jubilee is in yellow trace and the SH-50 in cyan trace:

 

2052892131_TADTD-4002Jubileevs.DanleySH-50stepresponse.thumb.jpg.f58ccfdb109a9e34f0a8b3ac90a83266.jpg

 

I guess we can legitimately call Klipsch Jubilees by the moniker "studio monitors" since their step response beats anything that I've seen posted from other studio monitors.  Additionally, the phase and group delay plot shows the same sort of amazing flatness:

 

1386355050_TADTD-4002Jubileevs.DanleySH-50phaseresponse.thumb.jpg.9609882a62112235dbcc16fd4b640a42.jpg

 

218615819_TADTD-4002Jubileevs.DanleySH-50groupdelayresponse.thumb.jpg.32536747d5ae7394d025af8f5c0bbc1f.jpg

 

Chris

  • Like 1
Link to comment
Share on other sites

Yes, I couldn't really believe the phase plot when I looked at it, too, so I also looked at the impulse spectrogram plot.  Note the flatness of the peak energy time curve, which looks as if FIR filtering was used (which it wasn't).  This is just the natural response of the horn/driver combinations within the Jubilee (TAD drivers):

703494743_TADTD-4002JubileeImpulseSpectrogram(FractionalOrderCrossoverFilters).thumb.jpg.144a2d2808a0e6c500fd4b2a218236f3.jpg

 

  • Like 1
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...