Worst measuring loudspeaker?

[sigbergaudio]

Well, Toole says it is the primary piece of information for frequencies above the bass range, in that you have to get it right first, and if you don't get it right, you can basically forget about trying to get top tier results. That's not 'pretty thin', that's 'pretty major', so IMHO your claim that 'we know' it is 'pretty thin' is both presumptive and wrong.

What we do know is that on-axis anechoic is not sufficient in itself, and if you don't also get off-axis behaviour right, then you also won't get top tier results. So the off-axis anechoic is a secondary but necessary piece of information.

Then we have the summed in-room response, that I would call a third-tier piece of information (maybe that's the one you should have said 'we know is pretty thin'), in that it tells us more about the room treatment than the speakers themselves. Furthermore people are prone to misuse it or exaggerate its significance. Hint.

Since I on a relatively regular basis listen to speakers that has a decent but not perfect on-axis response and find them to sound pretty good, and also listen to speakers that has a near perfect on-axis response and find them to sound pretty dull, I am not sure I agree. Time and time again I see people on this forum expecting to understand how a speaker will sound based on the on-axis response, while they simply cannot (unless it's objectively horrible of course).

So if we see a very poor on-axis response, we can assume it likely won't sound very good.
If we see a good on-axis response, we still don't know if it will sound very good.
A decent on-axis response, we also don't know.

 

[tove]

Well, Toole says it is the primary piece of information for frequencies above the bass range, in that you have to get it right first, and if you don't get it right, you can basically forget about trying to get top tier results. That's not 'pretty thin', that's 'pretty major', so IMHO your claim that 'we know' it is 'pretty thin' is both presumptive and wrong.

Toole is not the end-all. I agree with Toole that it is primary in the sense that it is the first thing to look at, for multiple reasons. But at least in a theoretical sense, I strongly disagree that on-axis must be a determinant of top tier results. The only truly important thing, to my knowledge, is that the off-axis response follows proportionally to the on-axis response. This would mean that the speaker takes well and predictably to equalization.

I agree with @sigbergaudio as far as on-axis being a pretty flimsy way of judging overall performance. Seems to me that a lot of its perceived importance comes from probabilities and one's experiences rather than an actual deduced logical reason. We know from experience that wildly non-linear responses are more likely to come from an incompetent and bad build, but as far as I know it is not a necessity. Do please educate me if you know better.

Then we have the summed in-room response, that I would call a third-tier piece of information (maybe that's the one you should have said 'we know is pretty thin'), in that it tells us more about the room treatment than the speakers themselves. Furthermore people are prone to misuse it or exaggerate its significance. Hint.

Well, here I would have to disagree very strongly. I know Toole believes in "listening through the room", but this is not gospel by a long shot. You are of course correct it tells us very little about the speakers, but it tells us a whole lot about the system as a whole, certainly much more than an anechoic on-axis does.

So if we see a very poor on-axis response, we can assume it likely won't sound very good.
If we see a good on-axis response, we still don't know if it will sound very good.
A decent on-axis response, we also don't know.

I didn't see your response before I posted mine. I agree with this completely, at the very least in a theoretical sense. Real world probabilities are probably more in the good on-axis response camp, more than decent on-axis camp, though.

 

[Sal1950]

Since I on a relatively regular basis listen to speakers that has a decent but not perfect on-axis response and find them to sound pretty good, and also listen to speakers that has a near perfect on-axis response and find them to sound pretty dull, I am not sure I agree. Time and time again I see people on this forum expecting to understand how a speaker will sound based on the on-axis response, while they simply cannot (unless it's objectively horrible of course).

This is about the speaker ever being able to produce an accurate reproduction of its input, not your personal preferences in tonal coloration.
If a speaker can't do that, all bets are off in it ever sounding natural. Toole is right.

 

[sigbergaudio]

This is about the speaker ever being able to produce an accurate reproduction of its input, not your personal preferences in tonal coloration.
If a speaker can't do that, all bets are off in it ever sounding natural. Toole is right.

Well, I like to think my preference is one of natural tonality, but I don't have much to back that up with beyond the sound signature of our speakers of course.

 

[MKR]

Well, I like to think my preference is one of natural tonality, but I don't have much to back that up with beyond the sound signature of our speakers of course.

Natural tonality? What does that even mean? Seems to me flat FR to start with would give you best chance for “natural tonality” (though again, not sure what that means)

 

[Sal1950]

Natural tonality? What does that even mean? Seems to me flat FR to start with would give you best chance for “natural tonality” (though again, not sure what that means)

Yep, Natural tonality = flat. Everyone knows that.

 

[sigbergaudio]

Natural tonality? What does that even mean? Seems to me flat FR to start with would give you best chance for “natural tonality” (though again, not sure what that means)

We can call it neutral tonality if you prefer. That it sounds natural aka similar to what you'd expect it to sound like in real life, to the extent you are listening to something with processing that isn't so hard that is impossible to determine.

 

[MattHooper]

We can call it neutral tonality if you prefer. That it sounds natural aka similar to what you'd expect it to sound like in real life, to the extent you are listening to something with processing that isn't so hard that is impossible to determine.

As hinted at in your response, the naturalness will of course depend quite a bit on the recording. If you have a recording emphasizes sibilance in an unnatural way, you’d have to adjust the frequency balance of that speaker out of neutral in order to have the voice sound more natural.

I’ve heard voices and instruments sound
“ natural” on both neutral and more coloured systems.

For instance, I’ve heard certain less-than-neutral systems that have just enough emphasis in the lower midrange to make voices sound (to me) more natural or more like the real thing. By that I mean, lots of vocal tracks tend to give me the “ chopped off head floating in space” effect. It’s sort of all throat or head resonance. Whereas when I’m in the presence of somebody speaking in a room, there is often significantly more chest resonance attached to their voice: that their head is attached to a resonating body. Sometimes I’m truly surprised it just how much chest resonance there is to some peoples voice.

The system I’m talking about seemed to emphasize the lower mids and perhaps upper bass just enough to avoid an unnatural emphasis, or boxiness, and instead I had the impression that many singers were appearing before me in a more fleshed out form, complete with some chest resonance, more like a real person in front of me producing that sound. (and that also played out in reproducing many instruments in a way that sounded more substantial and physical to me and hence a bit more natural and realistic, then some other more neutral speakers).

A neutral speaker is going to avoid the type of resonances that tend to strike our ears as unnatural. So that’s one leg up… especially but not limited to “ naturally recorded voices and instruments.”

But it’s a compromise either way. The variations in recordings make it so. And I think it’s possible that a Soundsystem may have some trends that somebody may focus on and enjoy, and which might strike them as “ overall more natural sounding.” (for instance maybe a slight deemphasis in the highs, which may have the trend of making various a unnatural recorded artifacts less prominent)

 

[tove]

Whereas when I’m in the presence of somebody speaking in a room, there is often significantly more chest resonance attached to their voice: that their head is attached to a resonating body. Sometimes I’m truly surprised it just how much chest resonance there is to some peoples voice

This is incredibly interesting and a nice catch. I'll be listening for this. I would however with some certainty claim that this is about the recording technique, and not the speakers.

With regards to the natural/neutral thing, I don't think it's very useful to bicker about the word we use. In the end, natural will be the same as the response of the microphone and/or mix/master system. And these tend toward neutral because it is more repeatable and a natural consequence of many technical designs. Then again, neutral is not objective good, nor is it perfectly defined in everyone's mind. It's all down to likelihoods really and neutral and natural are bound to be roughly the same on average.

 

[MattHooper]

This is incredibly interesting and a nice catch. I'll be listening for this. I would however with some certainty claim that this is about the recording technique, and not the speakers.

I’ve long been obsessed by the comparison of real versus reproduced sound. Always closing my eyes listening to real life sound sources in order to survey the type of qualities I hear, and ask myself “ what is it that is distinguishing this real sound from what I usually hear through sound systems?”

When people are talking around me, I sometimes close my eyes to concentrate on just the sonic characteristics I’m hearing.
Not that I’m doing this constantly of course.
But I’ve done it often enough to have my wife sometimes say “ why are you closing your eyes when I’m talking?” (I will leave any husband wife jokes there…)

One of the aspects that always strikes me in real voices is the very obvious organic quality - the sound is being produced by dampened wet flesh not some mechanical or electronic object. There is a softness, ripeness, roundness, organic quality, as opposed to the more electronic sounding voices I tend to hear and reproduced sound, voices often sound both more see-through but also hardened at the edges and sort of squeezed unnaturally tight. (some of which I attribute to the inadequacy of stereo phantom imaging, and some of which I attribute to the type of colorations that can occur through the recording and production processes.)

As I’ve said, I perceive my old CJ tube amps as having some softening, rounding, enriching effect, which to me often makes vocals sound more human. Purely anecdotally, my little Spendor s3/5s have always blown me away and how they reproduce the human voice. There is a sense of density yet round softness it seems to capture the gestalt of real voices better than many other speakers I’ve owned. When I listen directly to my wife’s voice speaking, and then listen to a female vocalist through the Spendors, there is less of an obvious distinction - much of the human quality seems there.

Yet the last time I put a Bryston amp in my system replacing the tube amplifiers, I was somewhat shocked to hear some degree of this human/organic quality diminished.
Voices sounded a little more squeezed hard and artificial. Not at all, due to some hardness imparted by the amplifier.
I take the amplifier as simply neutrally passing on the recorded signal, so this was just a more neutral presentation of the recordings. But I was a reminder of the role my tube amps seemed to have been playing in nudging the Spendors to such natural sound with voices. All anecdotal of course, but I just have to please my own perception, and that’s where my perception has ended up. It’s one of the reasons I don’t automatically associate “ neutral” with “ more natural.”

Another reason would come from working in POST PRODUCTION sound, where the dialogue guys are constantly altering dialogue recordings (hence not simply neutrally reproducing those recordings ) in order to make them sound more natural.

Likewise, in my sound design/editing duties I am constantly putting together totally unnatural combinations of sounds, sometimes in the service of producing something that sounds more natural.

So for me, there is a very serious gap between the concept of Fidelity and neutrality, versus what actually ends up sounding more natural or realistic.

Cheers.

 

[Jack B]

I’ve long been obsessed by the comparison of real versus reproduced sound. Always closing my eyes listening to real life sound sources in order to survey the type of qualities I hear, and ask myself “ what is it that is distinguishing this real sound from what I usually hear through sound systems?”

When people are talking around me, I sometimes close my eyes to concentrate on just the sonic characteristics I’m hearing.
Not that I’m doing this constantly of course.
But I’ve done it often enough to have my wife sometimes say “ why are you closing your eyes when I’m talking?” (I will leave any husband wife jokes there…)

One of the aspects that always strikes me in real voices is the very obvious organic quality - the sound is being produced by dampened wet flesh not some mechanical or electronic object. There is a softness, ripeness, roundness, organic quality, as opposed to the more electronic sounding voices I tend to hear and reproduced sound, voices often sound both more see-through but also hardened at the edges and sort of squeezed unnaturally tight. (some of which I attribute to the inadequacy of stereo phantom imaging, and some of which I attribute to the type of colorations that can occur through the recording and production processes.)

As I’ve said, I perceive my old CJ tube amps as having some softening, rounding, enriching effect, which to me often makes vocals sound more human. Purely anecdotally, my little Spendor s3/5s have always blown me away and how they reproduce the human voice. There is a sense of density yet round softness it seems to capture the gestalt of real voices better than many other speakers I’ve owned. When I listen directly to my wife’s voice speaking, and then listen to a female vocalist through the Spendors, there is less of an obvious distinction - much of the human quality seems there.

Yet the last time I put a Bryston amp in my system replacing the tube amplifiers, I was somewhat shocked to hear some degree of this human/organic quality diminished.
Voices sounded a little more squeezed hard and artificial. Not at all, due to some hardness imparted by the amplifier.
I take the amplifier as simply neutrally passing on the recorded signal, so this was just a more neutral presentation of the recordings. But I was a reminder of the role my tube amps seemed to have been playing in nudging the Spendors to such natural sound with voices. All anecdotal of course, but I just have to please my own perception, and that’s where my perception has ended up. It’s one of the reasons I don’t automatically associate “ neutral” with “ more natural.”

Another reason would come from working in POST PRODUCTION sound, where the dialogue guys are constantly altering dialogue recordings (hence not simply neutrally reproducing those recordings ) in order to make them sound more natural.

Likewise, in my sound design/editing duties I am constantly putting together totally unnatural combinations of sounds, sometimes in the service of producing something that sounds more natural.

So for me, there is a very serious gap between the concept of Fidelity and neutrality, versus what actually ends up sounding more natural or realistic.

Cheers.

Hi. I am very impressed with your most recent posting...you come across as genuinely thoughtful and logical and without an axe to grind. Your detailed attention to the sound of a human voice in a real space, vs the (all too often) sound of a reproduced human voice, should be something that most of us can relate to, if we listen carefully.

That said, I tend to think that your amplifier preference in this regard may well be compensating for defects in the recording chain and/or in the speakers/room. An amplifier as a tone control, which many tube amps are, whether in isolation or connected to a real-live speaker. Dr. Toole's Circle of Confusion comes to mind. How many recordings are mastered using speakers that are not flat or have weird dispersion characteristics, or are mastered by old farts using their long-failed hearing instruments (like me?) And of course, microphones do not hear the same way people do. Defects (even subtle ones) will be baked into recordings. Seems inevitable, given the many variables.

Still, correcting defects in one component of the reproduction chain by intentionally introducing them via another component does not seem to be optimum. Not saying that is what you are doing, but consider the possibility.

It's a wonder that anything recorded sounds real to any degree...and different people have differing sensitivities to the various imperfections that abound in our hobby. Ya picks yer poison.

 

[Newman]

We can call it neutral tonality if you prefer. That it sounds natural aka similar to what you'd expect it to sound like in real life, to the extent you are listening to something with processing that isn't so hard that is impossible to determine.

How about accurate tonality? Or is that a trigger word for you? Because it's actually the correct adjective in this case.

After all, the musicians themselves often choose to put the vocals through a modifier that gives them the tone or sound they want for the particular piece. What we presumably want, at home, is to accurately hear the vocals as modified.

 

[tove]

How about accurate tonality? Or is that a trigger word for you? Because it's actually the correct adjective in this case.

After all, the musicians themselves often choose to put the vocals through a modifier that gives them the tone or sound they want for the particular piece. What we presumably want, at home, is to accurately hear the vocals as modified.

Accurate? You get us the curve music studios are supposed to use, and some proof that a significant number of them actually use it in a calibrated monitoring system.

Accurate is not a word I would use. It's really more about probability. The distribution of likely tonality curves in systems used for mixing/mastering.

With a nicely calibrated neutral tonality, you'll be finding masters that are either too light or too heavy in some areas. You'll know you're on the razors edge (close to the distribution peak) when many good value productions sound off in different directions.

Now, this level of criticality isn't necessarily what you might want. I'd call is the most critical, sure. But one might like to bias the tonality some to minimize the problems. We would then be moving away from neutral, sure.

 

[MattHooper]

How about accurate tonality? Or is that a trigger word for you? Because it's actually the correct adjective in this case.

No because in the realm of audio “ accurate” does not equate to “ natural” sounding. (mostly due to nature of recordings.)

The other part of your post acknowledges this:

After all, the musicians themselves often choose to put the vocals through a modifier that gives them the tone or sound they want for the particular piece. What we presumably want, at home, is to accurately hear the vocals as modified.

In other words, if you reproduce those modified vocals accurately they will not sound “ natural.”

That’s why it’s worthwhile to have those two different modifiers, and deploy them usefully.

 

[Shorty]

‘Objective tonality’?

 

[Fidji]

We have another candidate to talk about - maybe not worst, but for sure not that great either

Totem Acoustic Element Fire V2 loudspeaker

Totem Acoustic was founded in 1987, in Montreal, Canada, by a former high school math teacher named Vince Bruzzese. The company's first product, the Model 1 loudspeaker, impressed me so much I bought a pair. My old pal George Stanwick, who worked at Stereo Exchange, spotted me walking up...

www.stereophile.com

with hilarious response from manufacturer, I think they do it properly - not arguing about some measurement details - but dismiss them completely ;-) .

• While we thank Stereophile for its efforts and of course Herb for this review, we rebuke traditional measurements, as no one is listening in an anechoic chamber and the need for excessive toe-in creates a narrow, nonideal, harder-to-enjoy listening arrangement.

 

[MattHooper]

we rebuke traditional measurements,

Rather convenient….

 

[kharan]

This certainly an interesting approach.

John Atkinson does understand traditional loudspeakers and their goals. But this Totem Fire V2 is simply not a traditional loudspeaker.

Torrent Crossoverless Drivers are a huge step up in speed, transient articulation, and clarity without having control loss on the lower two octaves of music. This eliminates tone and clarity issues which invariably lead to harmonic inconsistencies higher up the frequency register. Without any reservations, the Fire V2 does not need either a crossover or specific time aligning via tilt- or low-pass filters, as our full analysis of FFT slices in the band that covers Torrent-to-tweeter integration is more than exemplary. Remember, this is the 21st century.

—Totem Acoustic

 

[Newman]

Accurate? You get us the curve music studios are supposed to use, and some proof that a significant number of them actually use it in a calibrated monitoring system.

I don't know what you mean. It doesn't matter what 'curve' (?) studios use, or are supposed to use. No matter what they use, the playback system needs to keep it flat and reproduce it. Accurately.

Accurate is not a word I would use. It's really more about probability. The distribution of likely tonality curves in systems used for mixing/mastering.

I still don't get your point. Surely you are not suggesting the playback system ideally decodes the tonal modifiers used in individual productions or tracks? We want to hear whatever it was that the musicians and/or engineers in the studio heard, when they played to themselves the studio master (not the CD master or vinyl master or Spotify master), and said to themselves, that's it. Complete with every sound modifier they did or didn't use on each instrument or vocal. Accurately.

With a nicely calibrated neutral tonality, you'll be finding masters that are either too light or too heavy in some areas. You'll know you're on the razors edge (close to the distribution peak) when many good value productions sound off in different directions.

Now, this level of criticality isn't necessarily what you might want. I'd call is the most critical, sure. But one might like to bias the tonality some to minimize the problems. We would then be moving away from neutral, sure.

Yeah, no, this (popular) idea of the playback system being a non-accurate likability enhancer is not supported by the evidence that is collected when controlled listening tests are employed. Sure, when you are confronted with a poor-sounding mastering, it's a good idea to have a tone control in the system easily to hand, so you can make a tolerability adjustment, which could be in any direction, and that's why you don't want to buy special speakers or amps with fixed inaccurate reproduction: it might be just right for one bad mastering, but just wrong for the next dozen bad masterings...and never quite right for all the really good masterings. Oops!

Basically, the notion of a superior tuned playback system (with a fixed 'tone quality') is a marketing myth. Thanks Linn! Thanks Naim, Audio Research, Audio Note, even Marantz (to mention only a few)! Thanks a bunch!

There are a few areas where the evidence from controlled listening tests does support making playback adjustments to the accurate reproduction model, but it is strongly recommended to employ an accurate playback system as a starting point. At least that way we have an accurate reference as a default and know what we are doing when making changes.

• One area I already mentioned is the adjustable tone control for bad masterings.
• Another area is the choice to add a little reverb, because testing has shown that sound engineers have a chronic, probably work-entrained, sensitivity to reverb and tend to make the recordings 'dry'. This means you and I won't perceive as much reverb as they did when they thought it sounded right, so adding a little in playback is a good option to have. One, pathetic, approach to solving this is to have our home playback room a little on the reverberant or 'lively' side, but this introduces small-room reverberation which is not the euphonic sort. There are some psychoacoustically-developed algorithms out there that are much more the right idea, usually found residing in the bowels of AVRs, but one would want it to be adjustable in strength and not overdone.
• The third and final area that comes to mind is the level of bass output. Testing has shown that this really is an area where preference varies per individual as a matter of taste. Not the bass extension, which needs to be deep, nor the measured bass FR curve shape, which needs to be smooth and level as possible but without equalising deep troughs, but just the level between 30-130 Hz should be adjusted to taste. The research-based starting point is a soft boost of 2-3 dB, which roughly simulates small-room gain.

cheers

 

[MattHooper]

Thanks Jack for your very nice reply!

That said, I tend to think that your amplifier preference in this regard may well be compensating for defects in the recording chain and/or in the speakers/room.

Yes, some of that is certainly true (if we grant the amplifiers are altering the sound in the way I perceive). At least in terms of the type of artefacts common to recordings.

I always try to audition my speakers extensively before purchasing, and always preferably with solid state amplification, so that the speakers are given a chance to sound as neutral as possible, and I don’t have to worry that any colorations might be due to the amplification.

So that’s how I always evaluate loudspeakers.

I preferred my speakers over many others, including Revel, Kii Audio. Nothing stood out as to what I find to be a defect. But there isn’t a single speaker I’ve had in my room (including some really neutral speakers) that I didn’t prefer more on my tube amplifiers, for the type of characteristics I described.

(While I’m not using room correction, my room was redesigned with the input of a professional accusation, with carefully deployed materials and acoustic properties. It is a terrific sounding room, and every speaker I have placed in the room has sounded wonderful.

Defects (even subtle ones) will be baked into recordings. Seems inevitable, given the many variables.

Yup. A neutral well-designed loudspeaker will be free of additional unnatural resonances, but of course neutral doesn’t guarantee natural sound or high sound quality: that’s going to be recording dependent.

Still, correcting defects in one component of the reproduction chain by intentionally introducing them via another component does not seem to be optimum. Not saying that is what you are doing, but consider the possibility.

I wouldn’t say that I’m correcting for a defect in my speakers or anything else. I still sometimes use solid state amplification on my channel system, and the speakers still sound terrific. I just like it even more with the tube amplification. (and I am making some compromises in the sound with the tube amplification, among them bass is not as tight… though I still like the bass quality with the tubes).

My own goal is simply that I achieve a sensuousness of sound quality that drives me in of itself, while also enhancing the music music listening experience. Whether a neutral component or a less than neutral component achieves that for me, I don’t care per se.

However, I am usually drawn to loudspeakers that are fairly neutral at least, because obvious colorations are distracting to me.
However, if a system can have some colorations that are cannily introduced, which I find enjoyable without those colorations being obvious or jumping out to me as colorations, then I am fine with that too.

It’s the end result that matters to me.

As far as sub optimal, I listen to a range of recordings that are highly variable and recording quality. My ideal is that a system strikes the balance between allowing all those recordings to be pleasurable… so that I don’t feel like I have to go reaching for EQ or something like that to adjust per recording… but also that the distinct characteristics of different recordings are presented clearly.
Because I consider the different nature of recordings themselves to be of interest and part of the fun. So just enough coloration perhaps to nudge the sound in a direction that I find more pleasant and pleasurable, not enough to homogenize.

And for me, that’s exactly the balance I have struck. Recordings allow the space around my loudspeakers to shape shift dramatically, while I find the tube amps gently nudged the sound towards a bit more organic and less mechanical character, to my ears.

And since that is achieved, I don’t feel any need to introduce other equipment like EQ (I had one for 20 years and got rid of it because I didn’t need it), and I get the additional jollies and that I just love the look of tube amplifiers, the retro nature of the technology, how they work - I can see the musical signal glowing in the tubes! - and their connection with a long history in audio. All sorts of aspects that are not supplied by an EQ or some bit of software. So for me, the result is optimal in suiting my goals.

Cheers.