[TimVG]

Multiple capsules don't do anything else apart from giving you the same results as a single capsules, averaged over a slightly larger area to smooth out minor interference problems. They can however, introduce new issues since the microphone array introduces interference of its own as the waves get shorter.

Window gating is one way to filter out reflections, at the cost of resolution below a certain threshold as window gets narrower. In any case, it's impossible to produce a directivity index in any room with both loudspeaker and microphone capsule being fixed in position, it just doesn't work that way.

There are many solutions to work around loudspeakers with less than stellar directivity. These solutions cost money however, sometimes a lot of it. It's easier just purchasing good loudspeakers that behave themselves. You can still tailor your room to taste, which is not the same as trying to solve problems that could (should) have been taken care of in R&D.

 

[12B4A]

I've been wondering if any of the slew of circular beam forming mics PCBs for voice recognition (example) and/or acoustic cameras (example) could be leveraged at the hobbyist level for room characterization and system analysis.

 

[Kal Rubinson]

This is exactly what Trinnov does, when capturing data for their Room Optimizer. 4 capsules however, since they need to work in 3 dimensions.

Here's one you can play with: https://www.ebay.com/itm/New-Trinno...ov-Optimizer-Microphone-For-Sherwood-Newcastle-R-972-Home-Theater-/251767915954

 

[craig john]

Here's one you can play with: https://www.ebay.com/itm/New-Trinno...ov-Optimizer-Microphone-For-Sherwood-Newcastle-R-972-Home-Theater-/251767915954

It's hard to tell... are those all the same height? To calculate directionality in 3 dimensions one would need different height mic's, no?


Edit: Looking closer the middle mic looks taller.

 

[Scotth3886]

Time for a music break?

https://tidal.com/album/77674838

 

[Kal Rubinson]

It's hard to tell... are those all the same height? To calculate directionality in 3 dimensions one would need different height mic's, no?
Edit: Looking closer the middle mic looks taller.

Yes, the center-rear one is taller. These are surplus from an old Sherwood AVR with a version of Trinnov. I reviewed it here: https://www.stereophile.com/images/archivesart/510mitr.mic.jpg

 

[Floyd Toole]

Mr Toole, if lateral reflections of sufficient threshold can cause image shift and therefore ASW, are there such things as ASH (Apparent Source Height) and ASD (Apparent Source Depth) from ceiling and rear wall reflections, and if so, how might the timbral difference of these reflections relative to the on-axis response affect their perception?

Young-Ho

The width cue is easily perceived because of the placement of the ears. Binaural hearing is highly effective in the horizontal plane, which is why so much importance is placed on lateral reflections in concert halls - and homes.

Elevation cues are spectral, above about 7 kHz, and our localization precision is very poor because it requires some familiarity with the sound source spectrum. With the normal directivity of tweeters I would not expect ceiling reflections to add much or anything to a sense of elevation.

The sense of depth is also very imprecise, but most of the evidence points to a proportion between the direct sound and delayed reflections and reverberation. This information is in recordings and the times involved are so large compared to those created by small room reflections that they are not likely to be a substantial factor. Play with a good stereo upmixer to appreciate what augmented lateral reflections can do for a sense of space and envelopment, including depth. What does seem to be an important factor in delivering a sense of depth is to start with loudspeakers free of resonances. These tend to draw attention to the loudspeakers, not allowing the depth cues in recordings to be effectively communicated. This also applies to the surround loudspeakers - they need to be neutral as well.

 

[youngho]

The width cue is easily perceived because of the placement of the ears. Binaural hearing is highly effective in the horizontal plane, which is why so much importance is placed on lateral reflections in concert halls - and homes.

Elevation cues are spectral, above about 7 kHz, and our localization precision is very poor because it requires some familiarity with the sound source spectrum. With the normal directivity of tweeters I would not expect ceiling reflections to add much or anything to a sense of elevation.

The sense of depth is also very imprecise, but most of the evidence points to a proportion between the direct sound and delayed reflections and reverberation. This information is in recordings and the times involved are so large compared to those created by small room reflections that they are not likely to be a substantial factor. Play with a good stereo upmixer to appreciate what augmented lateral reflections can do for a sense of space and envelopment, including depth. What does seem to be an important factor in delivering a sense of depth is to start with loudspeakers free of resonances. These tend to draw attention to the loudspeakers, not allowing the depth cues in recordings to be effectively communicated. This also applies to the surround loudspeakers - they need to be neutral as well.

Thank you! I have been working my way through the third edition of your book and had been re-reading about the precedence effect. I had forgotten that the detection thresholds were relatively similar for vertical and lateral reflections, so I was wondering if this had any significant image shift effect, especially since the vertical reflections could be earlier, where the thresholds were lower, than the lateral ones in medium and larger rooms.

Similarly, there was a recent discussion on diyAudio about depth of soundstaging, with many discussants seeming to opine that more space behind the loudspeakers promoted this perception. The detection thresholds seemed significantly higher for near-coincident reflections compared with vertical or lateral ones, so I thought this seemed less likely than the ceiling ones but I thought I would ask. As I recall, there were some in the diyAudio thread who commented on cognitive factors (being able to see the loudspeakers) contributing to the perception of depth.

Can you comment on loudspeaker enclosure diffraction and effects on imaging, soundstaging, or perception of depth?

Thanks so much,

Young-Ho

 

[Jsin_N]

Points already addressed at least 10 times in this thread.

I'm not sure what cone material has anything to do with subjective blind listening tests. At the end of the day, it's about the sound emanating from the speakers and the sound only. That's what they're testing.

Colorations due to cabinet resonance is a huge factor. Dr. Toole has already commented on this and edge diffraction in this thread.

Also, he's given the reason why speakers are chosen with a 80-something or 90 something percentage confidence using Tukey Kramer correlation. Can't remember.

Gooddoc posted up a useful PDF on a company outside of Harman that uses spinoramas in their development and how to interpret it. You should give that a read.

Everything you're asking is answered here already.

 

[Floyd Toole]

Lastly, research is always on thin ice if the funding of the research can be questioned. Any well controlled studied that is peer-reviewed would be subjected to such scrutiny. I still think the results are less conclusive as what most people think. Given all the variables involved, I don't find it difficult to see why Revels are not the only speaker selling in containers and I would not all put it down to factors other than sonics.

All of this has been discussed in this and other threads, but here is a brief review. Most of the research leading to understanding the importance and meaning of anechoic measurements as correlated with double-blind subjective evaluations was done during my 26 year tenure as a research scientist at a Canadian government funded establishment: the National Research Council, in Ottawa. During those years, starting in 1985 I published several peer reviewed-papers in the Journal of the Audio Engineering Society. The work was widely appreciated by loudspeaker professionals and, because of my demonstrated expertise, in 1991 I was approached by Harman to be Corporate VP Engineering and incorporate the scientific knowledge and method into their product development programs. I set up a corporate research group to continue the quest for knowledge and it still exists after my retirement in 2007. All research results were and are publicly presented and published. The research group does not design products.

Revel has chosen to apply the knowledge to their products. Anybody can - it is in the public domain.

The "spinorama" measurement method is now in a standard created by CTA, a group populated by Harman competitors: ANSI/CTA-2034-A (2015). “Standard Method of Measurement for In-Home Loudspeakers”, Consumer Technology Association, Technology and Standards Dept., www.CTA.tech.

A few other manufacturers publish data in that or related formats. Professional measurement equipment exists that outputs data in the spinorama format. Others use it or derivatives of it but choose not to publish the fact or the results. Several international professional audio and acoustics groups populated by Harman competitors have given me personal awards for my contributions to the industry.

"Thin ice"? I don't think so . . .

If you really want to know the facts, have a read through:
https://amazon.com/Sound-Reproducti...-dp-113892136X/dp/113892136X/ref=mt_paperback?_encoding=UTF8&me=&qid=1554936336

My mug shot and bio are there too

 

[Karl Maga]

Lastly, research is always on thin ice if the funding of the research can be questioned. Any well controlled studied that is peer-reviewed would be subjected to such scrutiny. I still think the results are less conclusive as what most people think. Given all the variables involved, I don't find it difficult to see why Revels are not the only speaker selling in containers and I would not all put it down to factors other than sonics.

This post would’ve been more appropriately made a few days ago, on April 1st.

 

[Stephan Mire]

Points already addressed at least 10 times in this thread.

I'm not sure what cone material has anything to do with subjective blind listening tests. At the end of the day, it's about the sound emanating from the speakers and the sound only. That's what they're testing.

Colorations due to cabinet resonance is a huge factor. Dr. Toole has already commented on this and edge diffraction in this thread.

Also, he's given the reason why speakers are chosen with a 80-something or 90 something percentage confidence using Tukey Kramer correlation. Can't remember.

Gooddoc posted up a useful PDF on a company outside of Harman that uses spinoramas in their development and how to interpret it. You should give that a read.

Everything you're asking is answered here already.

Sorry, I don't think my objections have been adequately answered. Saying "it was answered already" doesn't boost any confidence. If my objections have already been answered 10 times over then it should be no problem for you to put my mind at ease and answer them for me.

 

[Stephan Mire]

As much as I have respect for the work done by Harman, the findings cannot so easily be generalised in my opinion as what many believe. I have some scepticism of the speakers they have used as comparisons. Some of them, as I recall, can be best be described as having lumpy frequency responses with obvious idiosyncratic characteristics. Did they investigate how people would compare a system that sums flat with good or acceptable dispersion but with different crossover points, etc? Have they controlled for different cone material, let's say paper, PP or modern materials such as ceramic or beryllium? Factor in different crossover topologies and cabinet colourations, or the absence thereof.

Here are 2 snapshots of some from one of their presentations.

It compares 4 speakers and then they conclude on how these speakers are rated, give confidence levels etc. Very credible but… Questions I have for example are, do listeners rated speaker 3 lower than 1 and 2 because of the dip in the midrange of speaker 2 or the bumpy upper bass response. Did they probe these differences and did they report the results? I think using averages in any research is statistically not the most powerful tool to understand any question. (The average family has 1.5 children). These tools, which is only the preliminary descriptive approach in a study cannot detect the possible underlining structure of all the factors involved in the phenomenon. especially intervening variables.

What if there are factors or cluster that can only be identified with multivariate statistically tools, ie there might be clusters in the ratings, ie a certain group of people who do prefer a speaker different to what is indicated in the averages. What is the inter-rater correlation vs agreement?

Lastly, research is always on thin ice if the funding of the research can be questioned. Any well controlled studied that is peer-reviewed would be subjected to such scrutiny. I still think the results are less conclusive as what most people think. Given all the variables involved, I don't find it difficult to see why Revels are not the only speaker selling in containers and I would not all put it down to factors other than sonics.

I perused the thread but I don't see any of these objections answered with any clarity. If Dr Floyd Toole could please address them it would be appreciated.

 

[GIEGAR]

Can anybody advise me how to choose wharfedale speakers? I am looking to set up a family karaoke sound system

Karaoke is murderous on typical domestic speakers. You should be looking at pro PA monitors (example), not Wharfdales and the like.

 

[Jsin_N]

Sorry, I don't think my objections have been adequately answered. Saying "it was answered already" doesn't boost any confidence. If my objections have already been answered 10 times over then it should be no problem for you to put my mind at ease and answer them for me.

I feel no obligation to put your mind at ease. Read. It's all there. I'm not doing your work for you.

 

[12B4A]

Sorry, I don't think my objections have been adequately answered. Saying "it was answered already" doesn't boost any confidence. If my objections have already been answered 10 times over then it should be no problem for you to put my mind at ease and answer them for me.

I perused the thread but I don't see any of these objections answered with any clarity. If Dr Floyd Toole could please address them it would be appreciated.

https://www.avsforum.com/forum/89-s...speakers/3038828-how-choose-loudspeaker-what-science-shows-88.html#post57883800

If one is skeptical and curious, they'll research. If one just wants to introduce doubt, they'll ask a slew of begging questions on boards and demand instant answers.

 

[Stephan Mire]

https://www.avsforum.com/forum/89-s...speakers/3038828-how-choose-loudspeaker-what-science-shows-88.html#post57883800

If one is skeptical and curious, they'll research. If one just wants to introduce doubt, they'll ask a slew of begging questions on boards and demand instant answers.

I am sceptical and I asked legitimate questions regarding the graphs I posted on page 88 which I don't see addressed and I did read much of the thread. Saying "go do the research" is not helpful.

 

[TimVG]

On what page??? Must I go back to page 1 and sift through any number of pages to find the information that specifically addresses my concerns?

1)Did listeners rate speaker 3 lower than 1 and 2 because of the dip in the midrange of speaker 2 or the bumpy upper bass response?

Floyde? Sean?

They rated it lower because of the issues already listed. Less flat listening window (bumps and depressions) and directivity issues. The proof lies with number 1 and 2, two loudspeakers that don't feature these problems - and consequently were rated higher. Number 4 has these issues, and a whole bunch more, and subsequently was rated lower.

It can be quite difficult for people to believe that a simply graph can give us such a wealth of information, yet it does.

 

[Stephan Mire]

They rated it lower because of the issues already listed. Less flat listening window (bumps and depressions) and directivity issues. The proof lies with number 1 and 2, two loudspeakers that don't feature these problems - and consequently were rated higher. Number 4 has these issues, and a whole bunch more, and subsequently was rated lower.

It can be quite difficult for people to believe that a simply graph can give us such a wealth of information, yet it does.

Okay, but you don't have to be snotty about it. Have they controlled for different cone material, let's say paper, PP or modern materials such as ceramic or beryllium? Factor in different crossover topologies and cabinet colourations, or the absence thereof? I don't see this in the research.

 

[TimVG]

Okay, but you don't have to be snotty about it. Have they controlled for different cone material, let's say paper, PP or modern materials such as ceramic or beryllium? Factor in different crossover topologies and cabinet colourations, or the absence thereof? I don't see this in the research.

If it measures the same it will sound the same - there is however much more to a driver than the material of the membrane. If there are issues, such as resonances either from the driver or from the enclosure - these will show up in the spinorama plot. But I believe you are not interpreting the research correctly: the research sets out guidelines as to what constitutes a good loudspeaker in terms of objective performance aka the end result. The research is not a loudspeaker design manual in itself.