What Contributes to "Smoothness" In a Speaker's Sound?

[ksulliva01]

Perhaps some engineers can weigh in on this question that has always bugged me. Comparing two speakers recently that both share pretty flat and respectable freq. responses, I noticed one pair has a decidedly smoother overall sound character. For lack of any better terms, "buttery," "rich," "polite," all come to mind. It's not high frequency rolloff as the speaker is very detailed and extends past 20K. The other speaker in question sounded a tad "harsher" with a kind of forwardness to the presentation that can sound good or bad depending on recording.

Speaker 1 (the smooth one) did measure on this site with ~2db dip between 2-4K. I guess this is essentially the BBC dip. Is this what's causing that warmth and politeness? It doesn't seem to be enough to add up. Or is it this dip in combination with room reflection due to a different dispersion pattern + other factors??

 

[Jim Taylor]

Perhaps some engineers can weigh in on this question that has always bugged me. Comparing two speakers recently that both share pretty flat and respectable freq. responses, I noticed one pair has a decidedly smoother overall sound character. For lack of any better terms, "buttery," "rich," "polite," all come to mind. It's not high frequency rolloff as the speaker is very detailed and extends past 20K. The other speaker in question sounded a tad "harsher" with a kind of forwardness to the presentation that can sound good or bad depending on recording.

Speaker 1 (the smooth one) did measure on this site with ~2db dip between 2-4K. I guess this is essentially the BBC dip. Is this what's causing that warmth and politeness? It doesn't seem to be enough to add up. Or is it this dip in combination with room reflection due to a different dispersion pattern + other factors??

If you figure out how to describe it logically and define it scientifically, perhaps we can figure out how to test for it properly and measure it accurately.

Jim

 

[DanielT]

Perhaps some engineers can weigh in on this question that has always bugged me. Comparing two speakers recently that both share pretty flat and respectable freq. responses, I noticed one pair has a decidedly smoother overall sound character. For lack of any better terms, "buttery," "rich," "polite," all come to mind. It's not high frequency rolloff as the speaker is very detailed and extends past 20K. The other speaker in question sounded a tad "harsher" with a kind of forwardness to the presentation that can sound good or bad depending on recording.

Speaker 1 (the smooth one) did measure on this site with ~2db dip between 2-4K. I guess this is essentially the BBC dip. Is this what's causing that warmth and politeness? It doesn't seem to be enough to add up. Or is it this dip in combination with room reflection due to a different dispersion pattern + other factors??

.. + other factors??
Yes, distortion.

If you'll excuse me, I'll skip a discussion of your words like buttery," "rich," "polite," because I don't know how you interpret that in sound terms. So generally about good sound instead.

In terms of determining good sound according to me and what I've seen others say, in descending scale of importance, to start with a good FR both on and off axes being the most important then dispersion and lastly distortion.

BUT the more SPL speakers are driven up to, the more distortion increases. So the question can be how good power handling a speaker can have. In addition, a low powered amplifier can in such a case be driven into clipping and that distortion usually sounds worse to our ears than speaker distortion.
The more distortion, the worse the sound, of course. At what level this distortion becomes annoying to even really annoying depends. The simplest is that if you turn up the sound really high on an old junk car stereo, or a clock radio something where, even at low SPL, it becomes really annoying with distortion. Just to get a feel for it all that is.

Dispersion that you were onto, absolutely. A matter of taste and what you like. A narrow directional sound can be experienced as "exact" but then also "narrow". Loudspeakers with wide sound dispersion can sound diffuse. Or lets say "smudged" but with a big sound (yes I realize that's a subjective description).
Then it depends on the type of room they play in, how much furniture, carpets and so on there is.

Speakers with poor FR, or large swings up and in FR. Then neither dispersion nor level of distortion matters that much. You can clearly hear these big swings up and down and you can be damned annoyed by that. Please note that this is MY personal experience. I'm guessing many others would say more or less the same but I'm not sure.

 

[kemmler3D]

~2db dip between 2-4K. I guess this is essentially the BBC dip. Is this what's causing that warmth and politeness? It doesn't seem to be enough to add up. Or is it this dip in combination with room reflection due to a different dispersion pattern + other factors??

IME a 2dB dip in that region could definitely lead someone to describe a speaker that way. It doesn't sound like a lot, but when you are comparing small differences and really thinking about it that way, the subjective impression can shift quite a bit based on a few dB.

I once spent a solid week voicing headphones by ear and +/- 2dB is enough to produce that kind of impression.

I would say +/- 2dB is where differences start to become obvious, but are still not distinct enough that you can easily put your finger on it.

It could also be dispersion, if the tweeter is beaming more or less it will tend to change the tonality in-room a bit.

 

[Keith_W]

There is also a question whether those speakers remain linear with increasing volume. Lots of people forget to test for this. One frequency response at 70dB, another frequency response at 80dB.

 

[JustJones]

What 2 speakers were you comparing? Might give more to go on knowing the types of speakers.

 

[Salt]

If you figure out how to describe it logically and define it scientifically, perhaps we can figure out how to test for it properly and measure it accurately.

Jim

Semantically perfectly

 

[ksulliva01]

Wow I forgot how poorly subjective adjectives are received here, sorry about that, I knew "buttery" was taking it too far

Let me try and rephrase my question.

Two different speakers, both 2-way, ported bookshelves, each with +/- 3db anechoic freq. response between 60hz and 20Khz., have notably different tonal characteristics when subjectively evaluated in my room, with most other factors accounted for. Same amp, same distance between L/R, same distance and toe-in angle to listening spot. Listening levels are matched with a db meter and are moderate (below 75 db) from a 150wpc class D amp, so the amp is not distorting much here, and both speakers I believe have respectable distortion characteristics.

The tonal differences that I'm referring to as either "smooth" or "polite" if I had to guess, would be correlated to the in-room perceived energy levels between roughly 2K and 4K. This relief (or emphasis) in the upper midrange seems to be the difference I'm trying to get at. Is this not the famous British "polite" speaker treatment that originated in a certain not-to-be-named mini monitor? From doing a fair bit of audio mixing, I understand that +/- 2 or 3 db across a certain frequency range like this can without a doubt cause a notable difference in the overall tone of a voice or instrument. However that doesn't seem to fully explain what I hear in these two sets of speakers, and my inkling was that a wider dispersion and the resulting reflections in my room is causing speaker 2 (the not "smooth" one) to gather energy in the upper mids which lends a more aggressive overall tonal characteristic.

Speaker set #1 are the Elac DBR-62s. These are the ones that sound less aggressive in the upper mids. Speaker set #2 are the CSS Citon 1TD. The Elacs have that tiny waveguide on the tweeter, the CSS tweeters are essentially flat with front baffle.

 

[AnalogSteph]

IME a 2dB dip in that region could definitely lead someone to describe a speaker that way. It doesn't sound like a lot, but when you are comparing small differences and really thinking about it that way, the subjective impression can shift quite a bit based on a few dB.

Indeed. Another factor to consider is literal smoothness of the frequency response... we can hear peaks much more easily than dips, and unevenness tends to sound rough. One thing that can be hard to spot in measurements but definitely result in subjective harshness is poorly suppressed woofer breakup modes... the rule of thumb is that they should be at least 20 dB down, which can be quite hard to achieve in things like metal cone woofers (they break up high but tend to have rather nasty mountain ranges in return).

Low Distortion Filter for PTT6.5X04-NAA

purifi-audio.com

Ironically woofer breakup issues may be falsely attributed to the tweeter instead (somewhat understandably - it is treble after all).

Speaker set #1 are the Elac DBR-62s. These are the ones that sound less aggressive in the upper mids. Speaker set #2 are the CSS Citon 1TD. The Elacs have that tiny waveguide on the tweeter, the CSS tweeters are essentially flat with front baffle.

The Critons seem to have the characteristic mid-treble widening associated with a flat-baffle tweeter.

CSS Criton 1TD-X Kit Speaker Review

This is a review and detailed measurements of the CSS Criton 1TD-X Kit speaker. I requested a sample from the company and shipped a pair to me already built in nice veneer: I believe the kit costs US $770 including an MDF flatpack. Typical of DIY kits, you get the value add of very heavy...

www.audiosciencereview.com

The Elacs have better-controlled directivity, not 100% perfect but definitely closer.

Elac Debut Reference DBR-62 Speaker Review

This is a review and detailed measurements of the Elac Debut Reference DBR-62 bookshelf speaker. It was kindly purchased new and drop shipped to me for testing. The DBR-62 costs US $600 a pair from multiple sources/dealers. The DBR-62 comes in black and walnut finish or the distressed Oak and...

www.audiosciencereview.com

These would be expected to be more relaxed in diffuse field.

 

[Chrispy]

Wow I forgot how poorly subjective adjectives are received here, sorry about that, I knew "buttery" was taking it too far

Let me try and rephrase my question.

Two different speakers, both 2-way, ported bookshelves, each with +/- 3db anechoic freq. response between 60hz and 20Khz., have notably different tonal characteristics when subjectively evaluated in my room, with most other factors accounted for. Same amp, same distance between L/R, same distance and toe-in angle to listening spot. Listening levels are matched with a db meter and are moderate (below 75 db) from a 150wpc class D amp, so the amp is not distorting much here, and both speakers I believe have respectable distortion characteristics.

The tonal differences that I'm referring to as either "smooth" or "polite" if I had to guess, would be correlated to the in-room perceived energy levels between roughly 2K and 4K. This relief (or emphasis) in the upper midrange seems to be the difference I'm trying to get at. Is this not the famous British "polite" speaker treatment that originated in a certain not-to-be-named mini monitor? From doing a fair bit of audio mixing, I understand that +/- 2 or 3 db across a certain frequency range like this can without a doubt cause a notable difference in the overall tone of a voice or instrument. However that doesn't seem to fully explain what I hear in these two sets of speakers, and my inkling was that a wider dispersion and the resulting reflections in my room is causing speaker 2 (the not "smooth" one) to gather energy in the upper mids which lends a more aggressive overall tonal characteristic.

Speaker set #1 are the Elac DBR-62s. These are the ones that sound less aggressive in the upper mids. Speaker set #2 are the CSS Citon 1TD. The Elacs have that tiny waveguide on the tweeter, the CSS tweeters are essentially flat with front baffle.

How about starting off with a measurement of each in your environment?

 

[kemmler3D]

my inkling was that a wider dispersion and the resulting reflections in my room is causing speaker 2 (the not "smooth" one) to gather energy in the upper mids which lends a more aggressive overall tonal characteristic.

Speaker set #1 are the Elac DBR-62s. These are the ones that sound less aggressive in the upper mids. Speaker set #2 are the CSS Citon 1TD. The Elacs have that tiny waveguide on the tweeter, the CSS tweeters are essentially flat with front baffle.

I'd say that explanation is totally plausible. Also, the speaker without waveguide will have less even directivity and also more diffraction - so you are getting like, little pockets of more direct treble at certain frequencies, this could maybe come out as "harshness" in a way that doesn't clearly show in an FR measurement.

 

[ksulliva01]

How about starting off with a measurement of each in your environment?

I attempted some measurements using REW of the Critons awhile back to see if there were any obvious peaks. Nothing looked too bad, honestly almost looks too good which leads me to believe I may have performed the test incorrectly. I also want to say for the record these Critons sound great and are not harsh so to speak, only in relation to the Elacs would I have really noticed the difference. I have some time alone this week, so maybe I'll read up more on REW and test each pair back to back. Would using a cardioid condenser skew the results at all? It's all I have on hand although I do have calibration files for it.

 

[MattHooper]

Wow I forgot how poorly subjective adjectives are received here, sorry about that, I knew "buttery" was taking it too far

No worries. Not everybody here is allergic to or dismissive of such adjectives. I think “buttery” is a perfectly good subjective description. The word evokes to me a high-end that is exceedingly smooth, rather than sharp or coarse, free of grain or edginess, easy on the ears and inviting to listen to… which can impart sort of feeling of luxuriousness to the sound.

In fact, I find my system to have such a character. I’m using Joseph audio perspective 2 speakers, which on their own have an extremely pure grain free sound. And then I’m driving them with Conrad Johnson tube amps. The sound strikes me as it once vivid and extended open and airy, Zero cents of “ Roll off in the highs” so that high frequency, detail like cymbals, chimes the top end of strings and woodwinds etc can pop out of the mix with vivid realism.

And yet the highs are completely incredibly smooth, free of any sense of grain edge or distortion, With an overall relaxed quality that has allowed me to crank this system louder than any other system I’ve owned. (I have tinnitus and hyperacusis so I can be quite sensitive to aggressive and coarse highs).

Once I set the system up originally, I had my pal come over, who is an audio reviewer, and not wanting to prejudice him I just had him sit down and listen (and play whatever he wanted). When he finished, I asked him to describe the sound of the system.
Everything he said was absolutely bang on in terms of what I hear. Detail after detail, the specific character of the bass, the imaging and everything else, was like he was pulling the sonic images from my head. And then he said one of the most prominent things he noticed was the upper frequencies, how they sounded simultaneously extended and brilliant, well totally and utterly smooth and relaxing to listen to. He literally use the words “ it’s like the equivalent of listening to melted butter.” he said he could’ve sat there all day listening because the combination of vividness and relaxed quality
was so ingratiating and easy to listen to.

That’s one reason why I really enjoy exchanging subjective impressions with other audiophiles.

Anyway, that’s to say, I get your language, even if it’s not much of a thing around here

 

[AnalogSteph]

A lot of things look flat when you're using a whopping 180 dB vertical scale. The standard for CEA2034 is a mere 50 dB.

Anyway, you can see a slight upwards tilt in the 2-4 kHz area that could explain things. Was this measurement done with a traditional omni?

I have done measurements with a cardioid before but I imagine it would make MMM a major pain depending on how good its directivity is... a good even-dispersion SDC may work, a classic LDC with 32 mm capsule not so much.

 

[RickS]

There is clearly a difference in how these speakers were voiced. While basically agree with @AnalogSteph’s earlier post, could use eq to lessen the on-axis FR difference and then you would be left with one less potential cause. If not the on-axis response, then the overall power response is the most likely explanation imo. As mentioned, the difference in directivity is very apparent. Not too surprising as the Elac is a more contemporary design.

 

[kemmler3D]

The Criton also has a healthy 3dB bump in the top octave... certainly wouldn't make it sound more buttery or laid back...

And yeah @Rick Sykora is right, if you look at the PIR there's quite a swing from one speaker to the other in the mids and presence region. I would not say we'd expect these speakers to sound "the same" based on measurements, maybe vaguely similar.

 

[jhaider]

Wow I forgot how poorly subjective adjectives are received here, sorry about that, I knew "buttery" was taking it too far

Let me try and rephrase my question.

Two different speakers, both 2-way, ported bookshelves, each with +/- 3db anechoic freq. response between 60hz and 20Khz., have notably different tonal characteristics when subjectively evaluated in my room, with most other factors accounted for. Same amp, same distance between L/R, same distance and toe-in angle to listening spot. Listening levels are matched with a db meter and are moderate (below 75 db) from a 150wpc class D amp, so the amp is not distorting much here, and both speakers I believe have respectable distortion characteristics.

The tonal differences that I'm referring to as either "smooth" or "polite" if I had to guess, would be correlated to the in-room perceived energy levels between roughly 2K and 4K. This relief (or emphasis) in the upper midrange seems to be the difference I'm trying to get at. Is this not the famous British "polite" speaker treatment that originated in a certain not-to-be-named mini monitor? From doing a fair bit of audio mixing, I understand that +/- 2 or 3 db across a certain frequency range like this can without a doubt cause a notable difference in the overall tone of a voice or instrument. However that doesn't seem to fully explain what I hear in these two sets of speakers, and my inkling was that a wider dispersion and the resulting reflections in my room is causing speaker 2 (the not "smooth" one) to gather energy in the upper mids which lends a more aggressive overall tonal characteristic.

Speaker set #1 are the Elac DBR-62s. These are the ones that sound less aggressive in the upper mids. Speaker set #2 are the CSS Citon 1TD. The Elacs have that tiny waveguide on the tweeter, the CSS tweeters are essentially flat with front baffle.

The upper mids are where the dispersion disruption between woofer and tweeter-on-a-flat-waveguide lies. So IMO that’s the obvious source of coloration.

 

[Chromatischism]

Irregular off-axis radiation pattern not matching direct sound = harsher, especially if there's spikes in the treble

 

[Multicore]

Measure the spectrum of both with pink nose in room. From that, see if you can develop an EQ to make one sound like the other. If you can, you may have the answer, or a strong clue. Do it in mono.

Given room effects and difference in radiation patterns it may not work out but it's more likely to inform than just talking about it.

 

[Bugal1998]

Wow I forgot how poorly subjective adjectives are received here, sorry about that, I knew "buttery" was taking it too far

As @MattHooper said, not everyone has such an allergic reaction to subjective descriptions.

Yes, the utility of subjective descriptions as reliable conveyors of common experience is higly questionable, but until we can directly transfer experiences from one brain to another it will remain part of the audio community's experience.

In the absence of objective data an experiential dialog needs to start somewhere, and I see no issue with that starting point being a subjective impression. Even once objective data is available there's still a need to convey how that data may translate into a subjective experience. Sure, it would be ideal if each person could see the data and listen for themselves, but that's often not possible.

The problem with subjective descriptions--as I see it--is when reviewers and salespeople use flowery terms to mislead consumers and peddle snakeoil with no accountability. No one should be ok with that.