• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

A Broad Discussion of Speakers with Major Audio Luminaries

What I’ve noticed here on ASR, though, is that a lot of people order speakers based solely on measurements, which is actually pretty blind and deaf.
Brick & mortar shops have dropped like flies, and most people don't live in the urban centers where the remaining shops still exist. The reality is that for the vast majority of people, it's either buy without listening or don't buy at all.
 
The reality is that for the vast majority of people, it's either buy without listening or don't buy at all.
I understand the issue.
However, the issue of "seeing while listening" versus "blind testing" is more or less irrelevant in this situation.
 
However, the issue of "seeing while listening" versus "blind testing" is more or less irrelevant in this situation.
Yes, because we now have very robust, reliable, repeatable, and effective measurements. So when people make extraordinary claims based on sighted listening, claims that are at odds with measurements but for which no blind listening data exists, a scientific-minded person will naturally dismiss them, because we know a great deal about how unreliable sighted listening is.
 
Howdy again, Dr. Toole. Hope you are well.

Routledge has a little discount going on their books at the moment, including the new 4th edition of Sound Reproduction, so I ordered a copy of the paperback edition today. Although I've read a number of your articles and some white papers, and watched some of your videos on this subject, I've yet to lay my hands on the actual textbook. So it should be a very interesting read.

I also watched your presentation for the Toronto AES on YouTube today. This and the chats with Erin and SoundStage have been good followups to your original CIRMMT talk in 2015.


Thanks again for all your contributions to the fields of home audio and acoustics. And for increasing our understanding of how to get better results from them.

My quest for some new speakers still continues. In the meantime though, I'm trying out some new headphones, namely the Sennheiser HD550 and HD560S (which are also both rather attractively priced at the moment). There is still unfortunately alot of "art" involved in interpreting and understanding headphone measurements. So I look forward to seeing what your book may have to offer on that as well. I wish we had more in-ear measurements of good loudspeakers in semi-reflective rooms to use for comparison though.

There appear to be two different versions of the HD560S btw. And the newer version, made after 2022 with the shorter cable and 1/4" adapter, appears to measure a little warmer in the bass and less bright in the treble than the original HD560S with the longer cable. More on this here.
 
Last edited:
Howdy again, Dr. Toole. Hope you are well.

Routledge has a little discount going on their books at the moment, including the new 4th edition of Sound Reproduction, so I ordered a copy of the paperback edition today. Although I've read a number of your articles and some white papers, and watched some of your videos on this subject, I've yet to lay my hands on the actual textbook. So it should be a very interesting read.

I also watched your presentation for the Toronto AES on YouTube today. This and the chats with Erin and SoundStage have been good followups to your original CIRMMT talk in 2015.


Thanks again for all your contributions to the fields of home audio and acoustics. And for increasing our understanding of how to get better results from them.

My quest for some new speakers still continues. In the meantime though, I'm trying out some new headphones, namely the Sennheiser HD550 and HD560S (which are also both rather attractively priced at the moment). There is still unfortunately alot of "art" involved in interpreting and understanding headphone measurements. So I look forward to seeing what your book may have to offer on that as well. I wish we had more in-ear measurements of good loudspeakers in semi-reflective rooms to use for comparison though.

There appear to be two different versions of the HD560S btw. And the newer version, made after 2022 with the shorter cable and 1/4" adapter, appears to measure a little warmer in the bass and less bright in the treble than the original HD560S with the longer cable. More on this here.
The Sennheiser most significant advantage is long term comfort. My son has gone through 3 pairs in over 10 years of constant use. I turned him onto the brand at age 11. Now at Indiana University with the headphones. He left his teddy bear at home.
 
  • Like
Reactions: ADU
I wish we had more in-ear measurements of good loudspeakers in semi-reflective rooms to use for comparison though
The notion that this defines what we perceive is superficially logical, but misses that important point that there is a brain between the ears that processes the BINAURAL signals, separating the direct sound from reflections and among those, the early from the late - and so on. In rooms the direct sound is a dominant factor in sound quality assessments, but it is not a dominant factor in room curves. A single in-ear measurement is really a room curve with HRTFs added. A single curve does not define how a loudspeaker sounds in a room so it cannot be a flawless reference.

Headphones remain a bit of a challenge, as the programs we listen to are stereo, not binaural, so the directional and spatial aspects are wrong. Headphones may have been involved in the mix/mastering process, but loudspeaker reproduction is the normal target experience. The headphone experience is not unpleasant - quite the contrary with respect to bass, which lacks room resonances. Only in-ear devices avoid the inevitable variations in external ears and ear canals which alter the acoustical coupling to the eardrum for all on- or around-ear designs. Fortunately, if there are no serious resonances, we adapt to many of these changes. Bass variations because of air leakage is a huge problem, as it can completely alter perceptions of spectral balance. Head motion tracking is needed for persuasive externalization, although some say that custom HRTFs help - in my limited experiences it was not a problem solver - moving the head destroyed any advantage. . There are likely to remain some strong individual variations. The fact that the eyes cannot see anything related to what is heard is a problem. Meditation may help :). In loudspeakers it is also true for the recorded material, but we see loudspeakers and a room which we can recognize. There is still a justification for live performances.

I only use headphones when there is no alternative, so I have not been motivated to chase the details. Sean Olive has done that and his summary in Chapter 15 is pretty thorough. .
 
Last edited:
Hey Dr. Toole,

I think it’s clear from your research that directivity is important to you.

How do you prioritize directivity performance? Constant directivity seems like a high priority if achievable, but is it more important to be constant or just continuous, without big jumps/dips? Do you feel the current trend of cardioid designs is worth the trade-offs (complexity in active systems, distortion in passive) or is there some level of “too much of a good thing” going on there? Or does the answer change depending on the quality of the room treatments?
 
Hey Dr. Toole,

I think it’s clear from your research that directivity is important to you.

How do you prioritize directivity performance? Constant directivity seems like a high priority if achievable, but is it more important to be constant or just continuous, without big jumps/dips? Do you feel the current trend of cardioid designs is worth the trade-offs (complexity in active systems, distortion in passive) or is there some level of “too much of a good thing” going on there? Or does the answer change depending on the quality of the room treatments?
The entire directivity issue is one that results, at best, in guidance, not a definitive target. The highest priority in a loudspeaker is to eliminate audible resonances, and that is still a challenge it seems. The target is a flattish and smooth on-axis curve, and smooth off-axis performance. This describes a loudspeaker that has a high probability of not fundamentally changing the sound of recordings being reproduced. The next layer of concern is spectral balance, and this is seriously complicated by two factors we cannot avoid: the listening room and recordings themselves. Because recordings are mixed and mastered in rooms and with loudspeakers that are different from yours or mine means that simple things like bass and treble balance are inconsistent. Many (most?) people ignore this and get on with enjoying the music, but fussy listeners may reach for tone controls or equalizers to improve things.

The off-axis performance - directivity vs. frequency - plays a role in this, but as these sounds are reflected or scattered (diffused) on their way to our ears the room is involved and they are all different. So, directivity is important, constant or smoothly changing seem to be acceptable options, but it is a secondary factor - direct sound matters more.

Cardioid loudspeakers offer the advantage of reducing the boundary effect of the wall behind the loudspeaker. It also alters the energy transferred to room modes. The former is probably widely advantageous. The latter depends entirely on the room geometry relative to the room mode/standing wave pattern. Chapter 14 explains this quite well, I think, but cardioid loudspeakers add a complication in that the angular/rotational orientation with respect to the room modes affects the energy coupling factor, as it does with dipoles. At low frequencies it is hard to avoid the advantages of multiple subwoofers that can be located where they need to be.
 
The entire directivity issue is one that results, at best, in guidance, not a definitive target. The highest priority in a loudspeaker is to eliminate audible resonances, and that is still a challenge it seems. The target is a flattish and smooth on-axis curve, and smooth off-axis performance. This describes a loudspeaker that has a high probability of not fundamentally changing the sound of recordings being reproduced.
Hi Floyd, would you agree that if 2 speakers both exhibit such characteristics, it would be prudent to blind listen to see which may be preferred by ears?

Cardioid loudspeakers offer the advantage of reducing the boundary effect of the wall behind the loudspeaker. It also alters the energy transferred to room modes. The former is probably widely advantageous.
Agree. This paper seems to suggest that lower LF decay times may be key to spatial perception as well. https://audioroundtable.com/misc/icad2005F96.pdf
The latter depends entirely on the room geometry relative to the room mode/standing wave pattern. Chapter 14 explains this quite well, I think, but cardioid loudspeakers add a complication in that the angular/rotational orientation with respect to the room modes affects the energy coupling factor, as it does with dipoles.
True, but with monopoles there is no such orientation option...if this helps.
At low frequencies it is hard to avoid the advantages of multiple subwoofers that can be located where they need to be.
Also true, but why not multiple cardioid subwoofers? If the AES site wasn't permanently broken I'd provide a direct link, but -
Controlling the mode excitation of rooms by using multiple low frequency cardioids in multichannel systems Convention Paper 6404
 
Hi Floyd, would you agree that if 2 speakers both exhibit such characteristics, it would be prudent to blind listen to see which may be preferred by ears?
Of course.
gree. This paper seems to suggest that lower LF decay times may be key to spatial perception as well. https://audioroundtable.com/misc/icad2005F96.pdf
Because room modes behave as minimum-phase phenomena, reducing the amplitude of a resonant peak also reduces the ringing. Evidence is that the frequency-response is more reliable as an audible cue. Section 9.6.2 in the 4th edition.
True, but with monopoles there is no such orientation option...if this helps.
Yes, monopoles make life simpler.
Also true, but why not multiple cardioid subwoofers?
If your budget allows it, why not? However, in the end I can't think whey there should be any difference in equally well engineered multiple woofer systems. Location matters most with both kinds of subs.
 
Because room modes behave as minimum-phase phenomena, reducing the amplitude of a resonant peak also reduces the ringing.
Right, I don't think any sane person argues against EQ. However, its usable for monopoles, dipoles and cardioids. Equally.
Evidence is that the frequency-response is more reliable as an audible cue. Section 9.6.2 in the 4th edition.
I will look at that section to try to make sense of your answer. The McGill paper showed the importance of room LF decay times with regards to spatial perception of correlated vs decorrelated LF.
It was shown that listeners were able to discriminate between spatial auditory images associated with multiple low-frequency audio signals emanating from subwoofers located in two reverberant reproduction environments, characterized as “home” versus “lab.” Whereas decorrelated subwoofer signals could not be discriminated from correlated subwoofer signals when the subwoofers were
positioned symmetrically to the front and back of the listener (the control condition), the discrimination was made possible when the subwoofers were positioned to the left and right of the listener (the test condition). Since the magnitude response measured at the position of the listener’s ears was held constant in all conditions, it was concluded that this discrimination was enabled by the interaural phase differences that were presented at the listener’s ears in the test condition, but not in the control condition.

Yes, monopoles make life simpler.
But with zero beneficial orientation capacity. Again, broken AES Low-frequency polar pattern control for improved in-room response Juha Backman
If your budget allows it, why not? However, in the end I can't think whey there should be any difference in equally well engineered multiple woofer systems. Location matters most with both kinds of subs.
Yes location matters but the cardioids (and dipoles) have lower decay times (paper in previous post), especially at frequencies where acoustic treatments are near useless or impractical. DIRAC purports to reduce times also.
 
I will look at that section to try to make sense of your answer. The McGill paper showed the importance of room LF decay times with regards to spatial perception of correlated vs decorrelated LF.
Don't bother, I just scanned the McGill paper and they are interested in spatial bass, not timbrally accurate bass. They are different, and there is a thread on this site discussing it at length. They simply found that below about 50 Hz decorrelated (spatial) bass was harder to discern in their small room - and this seems entirely plausible. Small rooms also present bigger problems for timbral accuracy. Ideally we want both timbral accuracy and some delivery of spatial bass. It has yet to be done in small rooms as far as I know.
s location matters but the cardioids (and dipoles) have lower decay times (paper in previous post), especially at frequencies where acoustic treatments are near useless or impractical. DIRAC purports to reduce times also.
The decay times are determined by the amplitude and Q of the room modes. The loudspeakers simply provide energy to these modes, some more than others, some are directionally sensitive, monopoles not so. If the energy is reduced, so is the ringing. If EQ is used the ringing also is reduced. It seems like physics. What am I missing?

BTW, I am leaving on a 2 week vacation tomorrow and may not be able to respond quickly to further discussions.
 
The notion that this defines what we perceive is superficially logical, but misses that important point that there is a brain between the ears that processes the BINAURAL signals, separating the direct sound from reflections and among those, the early from the late - and so on.
Perhaps the following is an interesting approach. Not to replace anything, but to supplement the spinorama measurements.
It is in stereo, though, and the limits of listening with generic HRTF (+non-calibrated HP) and without head tracking still apply.
 
Don't bother, I just scanned the McGill paper and they are interested in spatial bass, not timbrally accurate bass. They are different, and there is a thread on this site discussing it at length. They simply found that below about 50 Hz decorrelated (spatial) bass was harder to discern in their small room - and this seems entirely plausible. Small rooms also present bigger problems for timbral accuracy. Ideally we want both timbral accuracy and some delivery of spatial bass. It has yet to be done in small rooms as far as I know.
Hi Floyd,
Agree, as @Thomas Lund Lund has stated, much more research to be done! Regardless I'll look at that chapter ;-). Btw, I know of no law preventing EQ of "spatial bass".
Cardioid loudspeakers offer the advantage of reducing the boundary effect of the wall behind the loudspeaker. It also alters the energy transferred to room modes. The former is probably widely advantageous.
The decay times are determined by the amplitude and Q of the room modes. The loudspeakers simply provide energy to these modes, some more than others, some are directionally sensitive, monopoles not so. If the energy is reduced, so is the ringing. If EQ is used the ringing also is reduced. It seems like physics. What am I missing?
Perhaps I misunderstood your previous highlighted comments. I thought you were agreeing with what these AES papers found:
Low-frequency polar pattern control for improved in-room response - Juha Backman
Controlling the mode excitation of rooms by using multiple low frequency cardioids in multichannel systems - Convention Paper 6404
(again apologies for AES being completely broken, no links available). From the latter:
The use of a cardioid to radiate low frequencies in small rooms offers clear advantages over a monopole. First, a more balanced modal excitation is achieved; second, the decay times for singular (sparsely) modes in the room are noticeably shorter. The reduced amount of acoustic energy (-4.8dB) that is coupled into the room’s modal system reduces the decay times of the sparsely modal frequency range. Utilising this approach the acoustic design of the room requires less acoustic treatment in the modal range, thus it represents the more cost effective solution.
If not, what would the advantage be?

BTW, I am leaving on a 2 week vacation tomorrow and may not be able to respond quickly to further discussions.
Enjoy!
 
If not, what would the advantage be?
This has to be quick :). The advantage applies only if one is not using subwoofers. In other words the playback sound level is determined by the mid & high frequencies, and if the bass energy coupling is less with cardioid woofing, the room mode excitation will of course be less than a monopole woofer. But the location of the cardioid woofer in this case is determined by stereo needs, not where room mode peaks and nulls happen to be. The results are not guaranteed.
If one is using subwoofers the reduced coupling factor of cardioids is no advantage - just reduce the output from monopole subs and save money. Right?
 
How small is "small?"
There are no rigid rules that I know of, but my criterion is based on room resonances falling within the musical frequency range. One thing I love about live classical concerts the profoundly neutral smoothly extended bass in the large halls. By this standard all domestic spaces are small rooms.
 
If one is using subwoofers the reduced coupling factor of cardioids is no advantage - just reduce the output from monopole subs and save money. Right?
I can't agree with this, cardioids are not simply reduced output monopoles, for the same LP magnitude response, they offer lower decays times, as the above studies confirmed.
Please enjoy your vacation, maybe can resume this wonderful discussion upon your return.
Take care Floyd.
 
I can't agree with this, cardioids are not simply reduced output monopoles,
I'm on my way out . . . Monopoles are pressure sources maximally coupling energy at pressure peaks in standing waves. Dipoles are velocity sources and maximally couple their energy in pressure nulls, and when oriented 90 deg from the null line. One can place a dipole in a null and turn the resonance amplitude -and its decay - up and down by rotating it. Cardioids are a combination of both.
 
Monopoles are pressure sources maximally coupling energy at pressure peaks in standing waves. Dipoles are velocity sources and maximally couple their energy in pressure nulls, and when oriented 90 deg from the null line. One can place a dipole in a null and turn the resonance amplitude -and its decay - up and down by rotating it. Cardioids are a combination of both.
We are in agreement there, Safe travels
 
Back
Top Bottom