A Broad Discussion of Speakers with Major Audio Luminaries

[Thomas Lund]

A universal qualia, AE, possibly a whole dimension of hearing, is unaccounted for in audio reproduction guidelines.

After having spent decades reasoning against excessive dynamics distortion at the source ("Loudness Wars"), the situation with music in surround or 3D is rather the opposite: Excessive spatial/AE distortion during distribution, and by the listening room.

Michael Bishop from Telarc brought the topic up some 20 years ago, where he regarded measly satellite-and-sub systems an obstacle for music in surround gaining momentum. Based on recent AE studies, Michael was probably right, so we should try and avoid the same mistake in 3D.

Before dismissing the potential of music in surround or 3D, music lovers therefore ought to experience what stereo+ formats actually are capable of. Places satisfying the right-hand side of the chain, attached, unfortunately are sparse. I haven’t attended a home audio show even trying.

In case you make it to a pro audio conference with adequate 3D listening available, remember to be a little patient. Benefits of AE are not felt immediately but need 3-5 minutes to develop. This is consistent with affective touch where non-neural pathways (e.g. based on monoamines released to the tissues and/or to the bloodstream) complement our limited conscious bandwidth.

 

[Mark617]

Many thanks for this illuminating reply.

Anything less from major audio luminaries would be unacceptable.

 

[ADU]

Dr. Toole keeps referring to resonances as some all encompassing reason for why speakers are unable to reproduce what is input. Does anyone know what he means by resonance?

I would normally let Dr. Toole answer this for himself, but he appears to have left the thread. A resonance is an oscillation when an object is excited at, or close to its natural frequency. In your listening room, things like glass windows, doors, speaker cabinets, etc. can all resonate. He has said elsewhere (can't find it right now) that by far the largest source of resonance is the speaker driver. He decried high end audio where "heroic" measures are taken to reduce cabinet resonance but speaker driver resonance is relatively ignored.

Right. I think what Dr. Toole is referring to are resonances that stand out in some way, that are above the threshold of audibility. Which could affect the speaker's neutrality in some way.

I am not an expert in speaker design, but maybe the resonances in the cabinet or enclosure will show up as fairly narrow band peaks in both the on and off-axis responses on a spinorama graph. See his YouTube lecture for CIRMMT for more detail on this.

I have also noticed some broader band imbalances in the off-axis/sound power responses of some "neutral" speakers that are relatively flat on-axis, which could add some coloration as well. I assume these are related more to the drivers and cross-overs used on the speakers.

The graph below is an average of 10 speakers' SP responses.

 

[ADU]

I'm looking for some new speakers myself btw. And was wondering what, if anything, measurements and science can tell us about the differences in driver and enclosure design. And which are better.

I've been hearing some less positive things lately about the ported bass-reflex and passive radiator designs, and their effects on bass quality, for example. And was wondering if there was anything to this, and if I should be looking more at the sealed closed-box designs. Perhaps in combination with a sub or subs.

Here's one study that attempted to understand some of the reasons why sealed speakers like the Sony Yamaha NS10M were preferred by many audio engineers, which includes various measurements.

https://dt7v1i9vyp3mf.cloudfront.net/assetlibrary/n/ns10m.pdf

The PDF is from an article by Sound on Sound btw. Thoughts?

 

[levimax]

I am looking for some new speakers myself btw. And was wondering what, if anything, measurements and science can tell us about the differences in driver and enclosure design. And which are better.

I've been hearing some less positive things lately about the smaller ported bass-reflex and passive radiator designs, and their effects on bass quality. And was wondering if there was anything to this, and if I should be looking more at the sealed closed-box designs. Perhaps in combination with a sub or subs.

Here's one study, for example, that attempted to understand some of the reasons why sealed speakers like the Sony NS10M were preferred by audio engineers.

https://dt7v1i9vyp3mf.cloudfront.net/assetlibrary/n/ns10m.pdf

The PDF is from an article by Sound on Sound on btw. Thoughts?

Ported or sealed can both work fine if well designed. I prefer sealed as there is less to go wrong and they tend to be easier to integrate with subs because their roll off is slower and smoother but you give up some bass extension and efficiency. I would not base a purchase decision on sealed or ported and would read through the speaker reviews on this site for ideas and perspective. I am very happy with a pair of Neuman KH310 I recently purchased which are sealed and reviewed by Amir on this site.

 

[thewas]

The PDF is from an article by Sound on Sound btw. Thoughts?

See https://www.audiosciencereview.com/...s/transient-response.18745/page-5#post-763993
If you obviously have no deep bass you have also no issues there, everything is a compromise and in the end the audibility matters and the NS10 has other detrimental issues there which are long discussed in the corresponding threads.

 

[Floyd Toole]

I'm looking for some new speakers myself btw. And was wondering what, if anything, measurements and science can tell us about the differences in driver and enclosure design. And which are better.

I've been hearing some less positive things lately about the ported bass-reflex and passive radiator designs, and their effects on bass quality, for example. And was wondering if there was anything to this, and if I should be looking more at the sealed closed-box designs. Perhaps in combination with a sub or subs.

Here's one study that attempted to understand some of the reasons why sealed speakers like the Sony NS10M were preferred by many audio engineers, which includes various measurements.

https://dt7v1i9vyp3mf.cloudfront.net/assetlibrary/n/ns10m.pdf

The PDF is from an article by Sound on Sound btw. Thoughts?

BTW, it was Yamaha, not Sony, that made the NS-10M.

I know the authors of that paper, and we have had "discussions" about their perspective - which I don't agree with. The NS-10M is discussed in detail in the 3rd edition. The Auratone came first, as a representative of crappy loudspeakers in table radios and TVs of the 60s. That had legitimacy. The NS-10M came into the scene later - designed for the consumer market and incorrectly designed to radiate constant sound power, which for such a 2-way system resulted in a very non-flat on-axis response. The designers visited me at the NRCC in Canada, having read some of my papers, and they left regretting their decision, a design goal not to be repeated. However, the pros picked up on the timbral signature of the NS-10M as it closely duplicated that of the old Auratone, but with added bass. The UREI 811 series was an even larger speaker with a similar spectral balance. All of this is in the book with measurements demonstrating what I say. Many facilities use EQ (usually of misleading steady-state room curves) in setting up their systems, to there is another layer of uncertainty to what is actually heard.

As for the "magical" bass attributed to the NS-10M, like any loudspeaker with reduced low-frequency output, the bass sounds tighter. In fact, the time domain response of woofers is predictable from the frequency response because they are minimum-phase systems. One can literally "design" the spectral and time response of woofers with an equalizer. The measurements shown in the paper were done in an anechoic chamber that was not corrected at low frequencies, so some of the chamber resonances show up in the curves. They should have measured outdoors. The data are corrupted in the frequency range where important conclusions are being drawn. And, adjacent boundary (console) effects and room resonances are ignored, which in reality they cannot be.

If any recording engineer wants to "amplify" some part of the spectrum while mixing it is not necessary to use a faulty loudspeaker. Use the best, most neutral, resonance free loudspeaker possible and equalize it to imitate the direct sound that is desired. Timbre is dominated by direct sound, which is absolutely the case in "near field" listening. One can listen to many alternative "sounds" by touching icons, and return to neutrality at will. No need to clutter the control room landscape with many little loudspeakers.

There are no standards for bad sound, so trying to find a loudspeaker that will allow mixes to "translate" to all possible playback devices is a hopeless task, especially now that headphone listening is prevalent - and even some inexpensive ones can sound superb. Perhaps, it is worth considering that the customer who selects bad sounding playback devices should be reminded of the fact by what they hear, and be motivated to buy something better. Why cater to the lowest common denominator?

I think I may be grumpy this morning . . .

 

[MAB]

BTW, it was Yamaha, not Sony, that made the NS-10M.

I know the authors of that paper, and we have had "discussions" about their perspective - which I don't agree with. The NS-10M is discussed in detail in the 3rd edition. The Auratone came first, as a representative of crappy loudspeakers in table radios and TVs of the 60s. That had legitimacy. The NS-10M came into the scene later - designed for the consumer market and incorrectly designed to radiate constant sound power, so for such a 2-way system meant a very non-flat on-axis response. The designers visited me at the NRCC in Canada, having read some of my papers, and they left regretting their decision, a design goal not to be repeated. However, the pros picked up on the timbral signature of the NS-10M as it closely duplicated that of the old Auratone, but with added bass. The UREI 811 series was an even larger speaker with a similar spectral balance. All of this is in the book with measurements demonstrating what I say. Many facilities use EQ (usually of misleading steady-state room curves) in setting up their systems, to there is another layer of uncertainty to what is actually heard.

As for the "magical" bass attributed to the NS-10M, like any loudspeaker with reduced low-frequency output, the bass sounds tighter. In fact, the time domain response of woofers is predictable from the frequency response because they are minimum-phase systems. One can literally "design" the spectral and time response of woofers with an equalizer. The measurements shown in the paper were done in an anechoic chamber that was not corrected at low frequencies, so some of the chamber resonances show up in the curves. They should have measured outdoors. The data are corrupted in the frequency range where important conclusions are being drawn. And, adjacent boundary (console) effects and room resonances are ignored, which in reality they cannot be.

If any recording engineer wants to "amplify" some part of the spectrum while mixing it is not necessary to use a faulty loudspeaker. Use the best, most neutral, resonance free loudspeaker possible and equalize it to imitate the direct sound that is desired. Timbre is dominated by direct sound, which is absolutely the case in "near field" listening. One can listen to many alternative "sounds" by touching icons, and return to neutrality at will. No need to clutter the control room landscape with many little loudspeakers.

There are no standards for bad sound, so trying to find a loudspeaker that will allow mixes to "translate" to all possible playback devices is a hopeless task, especially now that headphone listening is prevalent - and even some inexpensive ones can sound superb. Perhaps, it is worth considering that the customer who selects bad sounding playback devices should be reminded of the fact by what they hear, and be motivated to buy something better. Why cater to the lowest common denominator?

I think I may be grumpy this morning . . .

You didn't come across as grumpy.

We hear tall tales about these Yamaha speakers often. Your book, and your postings here are some of few clear comments on why these speakers sound as they do, and help to debunk the mythologies about both the overall timbral balance and the bass response.

Thank you.

 

[Keith_W]

The NS-10M came into the scene later - designed for the consumer market and incorrectly designed to radiate constant sound power, so for such a 2-way system meant a very non-flat on-axis response.

We love it when you are grumpy

Anyway, if their design goal was to radiate constant sound power, it doesn't look as if they were very successful. (From spinorama.org).

 

[Timcognito]

 

[JPA]

... Why cater to the lowest common denominator?

I've often wondered the same thing. Mixing/engineering a recording so that it sounds decent on car radios, tabletop radios, etc. ignores the fact that radio listeners probably won't buy the recording, so why cater to them? And if they listen to music on crappy systems they probably don't care much about the sound quality anyway.

 

[Floyd Toole]

We love it when you are grumpy

View attachment 466769

Anyway, if their design goal was to radiate constant sound power, it doesn't look as if they were very successful. (From spinorama.org).

Thanks for being understanding. I just get tired of hearing the same old "stuff"over and over again. People don't read. This is all old news.

The curve you show is for the NS-10M Pro - the "improved" version for recording engineers that would play louder - but sound worse (unless you are a fan ).

Now that the topic is open again, here are some additional facts for the record; all are from the 3rd edition of my book, so it all is public knowledge.

Below is Figure 12.10 from the book. It shows a comparison of two loudspeakers with similar transducer sizes and therefore similar directivity indexes. One (the NS-10M original consumer product) and a JBL Pro that was designed to radiate flattish direct sound. You can see that they did a good job of hitting a flat sound power target over most of the frequency range.

The following Figure 12.11 shows the "improved" NS-10M Pro, which is significantly different, and worse sounding. Also shown is a comparison with the 1960s reference Auratone 5C, and (b) shows that it seems that Yamaha went the extra mile to make the Pro version sound even more like the crappy 1960s reference standard for "consumer audio". (c) shows that the designers changed their target performance, despite the praise from some professional enthusiasts - and even showed credible data in a spec sheet. It's a strange world.

The following Figure 12.9 shows a comparison between an Auratone 5C and a "time aligned" version of the old Altec 604-8G with a new horn: the UREI 811B, a large main monitor intended for in-wall/soffit mounting. Neither is anything to be proud of, but at least the Auratone was deliberately selected because it represented a common class of product at a point in time - now long gone. I see that someone is now making replicas of NS-10Ms for those feeling a sense of deprivation. I wonder if the woofers have white cones - it matters you know. In the double-blind subjective evaluations conducted at the NRCC for the Canadian Broadcasting Corporation, discussed in the book, the UREI rated the lowest of all candidate loudspeakers, in the judgement of professional recording engineers and producers who did not know what they were listening to. Knowing what we do now, that is no surprise.

The atrocious off-axis performance of the UREI may have contributed to control room designs of the period that used masses of absorption in the front half of the room. Best not to hear that dreadful sound at all.

That's it for now. Hope it all has added some historical perspective. Like much of history, it is interesting, but not to be repeated.

I think my grumpiness has turned into sarcasm . . .

 

[Floyd Toole]

It is also worth noting that, according to the measurements, the excess brightness that many users found in the NS-10M was caused by the woofer, yet the curative Kleenex or toilet paper was always placed over the tweeter. People didn't believe in measurements.

 

[respice finem]

To paraphrase a famous "meme" - you're wiping it wrong

 

[Mart68]

I see that someone is now making replicas of NS-10Ms for those feeling a sense of deprivation. I wonder if the woofers have white cones -

I have some and yes, they do (have white cones and sound quite bad).

 

[Palmer]

It is also worth noting that, according to the measurements, the excess brightness that many users found in the NS-10M was caused by the woofer, yet the curative Kleenex or toilet paper was always place over the tweeter. People didn't believe in measurements.

As one of those older guys who had to work with whatever was in the place we walked into, I can honestly say that the speakers didn't matter so much - the craft, if you had it, mattered more. Nobody measured, by the way. Some of us got Grammys, a lot of us didn't. To the people in the now - have fun, good luck - speak to the older folks when you have a chance, not everybody writes and we are passing away at some point.

Cheers

 

[bmc0]

As one of those older guys who had to work with whatever was in the place we walked into, I can honestly say that the speakers didn't matter so much - the craft, if you had it, mattered more.

Deeply flawed tools would surely make it more difficult to practice one's craft well, I'd think.

 

[goat76]

There are no standards for bad sound, so trying to find a loudspeaker that will allow mixes to "translate" to all possible playback devices is a hopeless task, especially now that headphone listening is prevalent - and even some inexpensive ones can sound superb. Perhaps, it is worth considering that the customer who selects bad sounding playback devices should be reminded of the fact by what they hear, and be motivated to buy something better. Why cater to the lowest common denominator?

I think I may be grumpy this morning . . .

I'm sorry to point this out, but if you believe that the NS10 was used in studios to somehow represent an "average bad" loudspeaker at the home of an average music consumer, you got it all wrong.

The only aspect of that idea that is true is the limited bass capacity, which allows for clear hearing of all elements in the music mix, ensuring they are not drowned out or missing energy in the midrange and thereby making the mix work well in frequency-limited systems as well. Making sure a mix translates well is NOT about "catering to the lowest common denominator"; a well-balanced midrange will also make a mix sound better in a good full-range system. That's what translation is all about; it goes both ways, otherwise it's not a good translation.

The best example I can give you is the sound of a kick drum. It can give the listener the impression that the kick drum is well and naturally heard in the mix while listening in a full-range system, but can be pretty much absent when listening to the same mix in the bass-limited system. If the mixing engineer makes sure the overtones of the kick drum are naturally heard higher up in the frequency range, the kick drum is not only heard better in bass bass-limited system, it will most likely also sound even better in a full-range system as well.

The above can quite easily be done by just limiting the bass frequency range of a full-range speaker with a high-pass filter, especially nowadays, when most audio productions are made in a digital work environment. But when the NS-10 started to gain popularity in studios, such an approach where likely not as easily done, or maybe not even thought of; the NS-10 just worked like that out of the box.

Other things mixing engineers found the NS-10 to be good at, compared to many other studio speakers of the time, were how easy it was to determine if the amount and the tails of the reverb sounded right, and if the level and time aspects of the compression sounded right. The reason for that may have been the pretty low level of distortion and the good performance in the time domain, compared to many other studio speakers of the time.



I don't care much about the NS-10s myself, but they obviously worked well as one of many studio tools for many mixing engineers; otherwise, they wouldn't use them, as their work is highly dependent on the result they got. It's not like it is a common thing that a music mix is done on just one studio monitor, and then "off it goes" to the music consumer without going through other speaker systems throughout the production. I have never heard an audio production and thought: Oh, that mix must have been done on a pair of NS-10s because of the way it sounds. Have you?

 

[goat76]

I've often wondered the same thing. Mixing/engineering a recording so that it sounds decent on car radios, tabletop radios, etc. ignores the fact that radio listeners probably won't buy the recording, so why cater to them? And if they listen to music on crappy systems they probably don't care much about the sound quality anyway.

Yes, it's a shame the way mastering has, for a long time, been used to maximize the loudness to the point of almost completely killing all the dynamics in many audio productions, but that has nothing to do with the use of the NS-10s and similar monitors, which are used in the production stage of mixing the music.

 

[Palmer]

Deeply flawed tools would surely make it more difficult to practice one's craft well, I'd think.

The opposite is true, believe me. Ran into Bo Diddley when he was doing smaller gigs, and he would blow your socks off - whichever guitar you'd give him.
These days everybody can look up just about anything on the interweb, promptly thinking it's not that hard - and have at it. Mixed results, mostly - pun intended.
Experience comes from failing, and trying again. As to deeply flawed, yeah, sure... If you say so.

Cheers