Headphones or speakers?

[drogulus]

      I think some of the Stereophile discs were recorded with a Blumlein pair supplemented with outriggers (probably for bass response).

     

[Rod Corkin]

(a) Recording engineers listening through headphones are attending to attributes other than "imaging" (the apparent distribution of sound in space).
(b) Dynamic range of a CD = 96dB.  Dynamic range of an orchestra or rock band > 115dB.

Who is talking about an orchestra or rock band?? I'm taking largely about string quartets, harpsichords etc. You get a range from barely audible to speaker distorting levels at the same amp volume setting on some of the CDs I've got. I can't say I've ever heard anything like this live from a quartet. But even which orchestral music the instrumental balance is often not realistic. Really sometimes these guys haven't got a clue, the sound is completely fake.

[DavidRoss]

Who is talking about an orchestra or rock band?? I'm taking largely about string quartets, harpsichords etc. You get a range from barely audible to speaker distorting levels at the same amp volume setting on some of the CDs I've got. I can't say I've ever heard anything like this live from a quartet. But even which orchestral music the instrumental balance is often not realistic. Really sometimes these guys haven't got a clue, the sound is completely fake.

Ever heard the expression, "the pot calling the kettle black?"

Maximum sound pressure levels of various sources, including musical instruments

[sound67]

You get a range from barely audible to speaker distorting levels at the same amp volume setting on some of the CDs I've got.

There are some classical recordings (one I readily remember is Järvi's La Mer/Roussel 4th etc on Chandos) where you'll find it difficult to adjust the volume only once during playback because of the huge dynamic range. Actually I don't think this is "fake" at all. On the contrary, this dynamic range may just be too close to reality for the speakers to cope. But I never ever experienced this with string quartets or harpsichords.

Usually, limiters and compressors are used to keep the dynamic range within the limits that most speakers can handle. Whether that's an entirely good thing is open to question (especially in this day and age of "living-room-friendly" miniature speakers - so aptly called Brüllwürfel in German), since that is clearly a distortion of reality.

Thomas

[Rod Corkin]

There are some classical recordings (one I readily remember is Järvi's La Mer/Roussel 4th etc on Chandos) where you'll find it difficult to adjust the volume only once during playback because of the huge dynamic range. Actually I don't think this is "fake" at all. On the contrary, this dynamic range may just be too close to reality for the speakers to cope. But I never ever experienced this with string quartets or harpsichords.

Usually, limiters and compressors are used to keep the dynamic range within the limits that most speakers can handle. Whether that's an entirely good thing is open to question (especially in this day and age of "living-room-friendly" miniature speakers - so aptly called Brüllwürfel in German), since that is clearly a distortion of reality.

Thomas

Well I have looked around on other forums and the issue of unrealistic dynamic range has been mentioned by others, I'm not just making this up. But there are worse things - mix issues (eg solo vocalists too prominent in the mix) and inconsistent recording levels between movements (I presume the result or different recording sessions) are other things I find annoying even on the most modern recordings.

[71 dB]

That observation--correct in principle though not in particulars--is not relevant to the issue.

Sorry, what is the issue? The title is "Headphones or speakers" and people are talking about dynamic ranges and decibels. I made a correct observation. Why is it not relevant? I am confused. 

[DavidRoss]

Sorry, what is the issue? The title is "Headphones or speakers" and people are talking about dynamic ranges and decibels. I made a correct observation. Why is it not relevant? I am confused. 

(a) A raucous crowd fueled by booze, narcotics, and testosterone is not usually present at recording sessions (except, of course, when Celine Dion is in the studio).
(b) The recording engineer's challenge in capturing low-level information masked by a high noise floor does not change the dynamic range of the music itself.

At least you know you're confused, which is great, since knowing you don't know makes learning possible.  Alas, poor Corkin! 

Though he has the particulars backwards, it is true that more attention is getting focused on the issue of unrealistic dynamic range in digital recordings these days.  I don't believe that severe compression making ultra "hot" recordings has much infected the classical music industry (though wouldn't be surprised to see it with those "crossover" records aimed at a general audience, i.e. Renee Fleming Sings Radiohead), but it is an increasingly common practice with pop music that has gotten so out of hand that the range on a typical pop record these days is something like 6dB.  See, for instance, this site, for a clear explanation of what's being called "the loudness war." 

I suspect that among the main reasons the masses seem to prefer music recorded at unrealistically high levels with severe dynamic compression are (a) unfamiliarity with real music, having been exposed almost exclusively to recordings (and of amplified, not "acoustic" music) and (b) listening to (or hearing, at least!) music against a noisy background environment, i.e. in the car or at the gym or on a busy street or playground.  The more nearly realistic range of properly recorded classical music in such environments makes it almost impossible to hear because one must constantly turn the volume down and up to accommodate loud and quiet passages in the music.  (Ever try listening to a BIS CD in your car?  Impossible!).  Even at home, with inadequate playback gear, if one turns up gain to hear very soft pianissimos, the following tutti fortissimos are likely to drive the amplifier into clipping and/or cause speaker breakup or congestion.

I've never heard a recording with different levels between movements.  Perhaps Mr. Corkin is equally confused on this point and thinks that the loudness difference between the fortissimo tutti ending of movement one and the delicate pianissimo solo flute beginning of movement two is due to the recording and not to the music itself.  However, he is certainly right about some of the other choices made by recording engineers (or, more accurately, made by their business masters but executed by the engineers!)--particularly the practice of "spotlighting" a featured performer by boosting her levels relative to the orchestra or ensemble accompanying her. 

This practice seems particularly widespread in recordings featuring one of the current crop of "sweet young thing" violinists or sopranos, the decision driven by marketing, not technical considerations.  Even this, however, is not necessarily unrealistic, since what the listener hears in real life depends largely upon his position in the acoustic space relative to the performers.  For instance, a couple of years ago we attended a series of violin concerto performances, for which we chose seats in center front.  The loudness of the soloists relative to the rest of the orchestra in that spot differed significantly from the balance one hears, say, two-thirds of the way back, or in the dress circle.

[Bogey]

Great post David.  Just out of curiousity, and your recent spins of vinyl, how do you compare the two sounds (vinyl v. cd) when it comes to different music genres and keeping your above comments in mind?  Hope you caught my earlier post on this.

Also, can we toss in the notion of the boom of portable players making this "loudness" ingrediant a factor?

[scarpia]

(b) Dynamic range of a CD = 96dB.  Dynamic range of an orchestra or rock band > 115dB.

Really? You forget the background noise. The audience of a rock concert makes easily 70 dB noise. 115 dB - 70 dB = 45 dB is the usable dynamic range.

That observation--correct in principle though not in particulars--is not relevant to the issue.

Mr. Ross, it was your observation which was irrelevant.  The number you cite (apparent from the link you later posted) is not "dynamic range" but "loudness"  Dynamic range is the ratio between the intensity of the loudest and softest passages in a sound source.  Loudness is the ratio of the intensity to a standard intensity (supposed to be the lowest intensity that is perceptible).  The source for your 115dB is quoting intensity, how loud the rock band is with respect to the standard sound level.   There is no limitation on how intense a sound can be recorded on a CD, you just turn the gain down.  There is a limitation on dynamic range, since the softest sound that a CD can reproduce is 96 dB below the loudest sound.  71dB is correct that a rock band doesn't require much dynamic range, since it is uniformly loud and does not vary in intensity much (as evidenced by the fact that it is typically heard against a large amount of background noise.
 

[DavidRoss]

Mr. Ross, it was your observation which was irrelevant.  The number you cite (apparent from the link you later posted) is not "dynamic range" but "loudness"  Dynamic range is the ratio between the intensity of the loudest and softest passages in a sound source.  Loudness is the ratio of the intensity to a standard intensity (supposed to be the lowest intensity that is perceptible).  The source for your 115dB is quoting intensity, how loud the rock band is with respect to the standard sound level.   There is no limitation on how intense a sound can be recorded on a CD, you just turn the gain down.  There is a limitation on dynamic range, since the softest sound that a CD can reproduce is 96 dB below the loudest sound.  71dB is correct that a rock band doesn't require much dynamic range, since it is uniformly loud and does not vary in intensity much (as evidenced by the fact that it is typically heard against a large amount of background noise.

Dynamic range is the difference, expressed in decibels, between peak sound pressure level (i.e. the loudest tutti passage) and minimum spl (ideally 0 dB in the case of a rest sufficiently long for reverberation to cease in a perfectly quiet environment).  Per the source I cited, the maximum spl for an orchestra is ~137 dB, and for a rock band it's ~150 dB.  The dynamic range of either is the same (you have heard of rests, haven't you?).  What 71dB refers to as the "usable dynamic range"--i.e. the difference between peak spl and the noise floor—is somewhat less: under typical recording conditions the noise floor is, say, ~30 dB, making a usable dynamic range of >100 dB.

The condition you describe involving a rock band playing at a near uniform spl within a very limited dynamic range is hardly representative of music in general and has virtually no relevance in discussing the ability of CDs to adequately represent the dynamic range inherent in all kinds of music...and it has no relevance whatsoever to the issue raised by Mr. Corkin and addressed in my foregoing posts.

[DavidRoss]

...how do you compare the two sounds (vinyl v. cd) when it comes to different music genres and keeping your above comments in mind?  Hope you caught my earlier post on this.

Also, can we toss in the notion of the boom of portable players making this "loudness" ingrediant a factor?

To me, Bill, vinyl playback just generally sounds more like real music--slightly richer, fuller, a bit more relaxed and more liquid--in comparison to CD, somewhat like the comparison between tubes and transistors.  What one hears will, of course, vary tremendously depending upon playback equipment and conditions.  CD mastering and Digital to Analog Conversion have come a long way since the '80s, when the original CDs were lifeless and so shrill they were painful to hear.  Upsampling, oversampling, and high-resolution have significantly improved digital sound as well.

Insofar as the difference among genres, I would say that a much greater percentage of classical recordings are made with a concern for musical fidelity, with some labels specializing in such "audiophile" virtue, whereas the overwhelming majority of pop recordings are engineered as hot and highly compressed as possible.  And, yes, I think it's no coincidence that the death of high fidelity has gone hand in glove with the spread of "mp3s" and portable digital music players.

[sound67]

To me, Bill, vinyl playback just generally sounds more like real music--slightly richer, fuller, a bit more relaxed and more liquid--in comparison to CD, somewhat like the comparison between tubes and transistors. 

Both I think are myths. The LP has a reduced dynamic, as have tubes in comparison to transistors, not to mention those lovely pops. Most of the tall tales about either hark back to the CD mastering, or transistor-amp technology,  . The tube issue e.g. is raised here: http://www.theaudiocritic.com/downloads/article_1.pdf

Digital technology is cold and analytic, while analogue is warm and cozy. Yeah, right. I tried several tube amps (and tube CDPs) myself. Only badly designed amps sound warm and cozy. 

The real problem is that people no longer want properly sized speakers in their living rooms. All these mini-cube 5.1 / 6.1 / 7.1 sets, independent of the price, are just incapable of delivering a true hifi-sound over the whole dynamic spectrum.

Thomas

[Bogey]

To me, Bill, vinyl playback just generally sounds more like real music--slightly richer, fuller, a bit more relaxed and more liquid--

One sales pitch I heard recently for some high end equipment was that "with this set-up, your cds will sound like vinyl."   

[sound67]

Yeah. I'm waiting for the day they come up with something like "Buy this internet radio, and it will sound like a Volksempfänger" 

Someties, it does.

Thomas

[DavidRoss]

One sales pitch I heard recently for some high end equipment was that "with this set-up, your cds will sound like vinyl."   

Might be some good stuff, eh?  "New and improved" ain't necessarily so.  My ears aren't particularly golden--especially in middle age!--but I'd rather trust my own hearing than the pet theory of some arrogant jerk with an ax to grind.  Interesting that vinyl is making a comeback.  Of course, according to those who consider themselves our betters, only idiots would prefer the virtues of analog over those of digital.  If they enjoy wrapping themselves in their comforting blankets of smug superiority, let them; we "idiots" will just have to content ourselves with enjoying superior sound.

Incidentally, I listened to Lloyd-Jones's Naxos CD of Alwyn's Lyra Angelica last night for the first time in a year or so--astonishingly good sound!

[sound67]

but I'd rather trust my own hearing than the pet theory of some arrogant jerk with an ax to grind.  Interesting that vinyl is making a comeback.  Of course, according to those who consider themselves our betters, only idiots would prefer the virtues of analog over those of digital.  If they enjoy wrapping themselves in their comforting blankets of smug superiority, let them; we "idiots" will just have to content ourselves with enjoying superior sound.

They are the ones that are arrogant? 

Determining the accuracy of audio reproduction is a completely objective process. Evaluating aesthetic satisfaction derived from audio reproduction is, on the other hand, totally subjective. Critics who confuse or conflate the two are doing the consumer a great disservice and are responsible for most of the grotesque misinformation that blights today's audio journalism. We believe in measuring and we believe in listening but we don't believe in measuring with our ears.

(from the Audio Critic hp)

Incidentally, I listened to Lloyd-Jones's Naxos CD of Alwyn's Lyra Angelica last night for the first time in a year or so--astonishingly good sound!

And it's not even on vinyl.

Thomas

[DavidRoss]

They are the ones that are arrogant? 

Yes.  See the following quote (which you conveniently reproduced above):

Determining the accuracy of audio reproduction is a completely objective process. Evaluating aesthetic satisfaction derived from audio reproduction is, on the other hand, totally subjective. Critics who confuse or conflate the two are doing the consumer a great disservice and are responsible for most of the grotesque misinformation that blights today's audio journalism. We believe in measuring and we believe in listening but we don't believe in measuring with our ears.

(from the Audio Critic hp)

Having only part of the story is just ignorance.  Having only part of the story but insisting that you know it all is arrogance.  Having only part of the story, insisting that you know it all, and attacking anyone who disputes your omniscience as either an idiot or an unscrupulous huckster (or both) indicates a far more serious personality disorder than mere arrogance.

And it's not even on vinyl.

That's correct.  The recording engineer has far more to do with preserving or distorting sound quality than the recording's storage medium.  If the source is good, then your downstream playback equipment has a shot at reproducing good sound.  If the source is crappy, then good equalization or digital-to-analog conversion, clean signal amplification, and even the best, most neutral, full-range loudspeakers will deliver crappy sound.

[71 dB]

(a) A raucous crowd fueled by booze, narcotics, and testosterone is not usually present at recording sessions (except, of course, when Celine Dion is in the studio).
(b) The recording engineer's challenge in capturing low-level information masked by a high noise floor does not change the dynamic range of the music itself.

At least you know you're confused, which is great, since knowing you don't know makes learning possible.  Alas, poor Corkin!

I know what I'm talking about but I don't know what you are talking about. It's clear that in studio the background noise level is very low (or the studio sucks) but I was talking about live concert were fans make noise and also levels of music are much higher than in studio.

Though he has the particulars backwards, it is true that more attention is getting focused on the issue of unrealistic dynamic range in digital recordings these days.  I don't believe that severe compression making ultra "hot" recordings has much infected the classical music industry (though wouldn't be surprised to see it with those "crossover" records aimed at a general audience, i.e. Renee Fleming Sings Radiohead), but it is an increasingly common practice with pop music that has gotten so out of hand that the range on a typical pop record these days is something like 6dB.  See, for instance, this site, for a clear explanation of what's being called "the loudness war."

I know the loudness war, I am an acoustics engineer for heaven's sake!

I suspect that among the main reasons the masses seem to prefer music recorded at unrealistically high levels with severe dynamic compression are (a) unfamiliarity with real music, having been exposed almost exclusively to recordings (and of amplified, not "acoustic" music) and (b) listening to (or hearing, at least!) music against a noisy background environment, i.e. in the car or at the gym or on a busy street or playground.  The more nearly realistic range of properly recorded classical music in such environments makes it almost impossible to hear because one must constantly turn the volume down and up to accommodate loud and quiet passages in the music.  (Ever try listening to a BIS CD in your car?  Impossible!).  Even at home, with inadequate playback gear, if one turns up gain to hear very soft pianissimos, the following tutti fortissimos are likely to drive the amplifier into clipping and/or cause speaker breakup or congestion.

Agreed.

Portable players should have build-in dynamic compression that can be turned off. I also recommend dramatic reduction of channel separation below 500 Hz in headphone listening. Perhaps even the headphones themselves could have this feature? CDs should have "original" dynamics, of course.

[DavidRoss]

I know what I'm talking about but I don't know what you are talking about. It's clear that in studio the background noise level is very low (or the studio sucks) but I was talking about live concert were fans make noise and also levels of music are much higher than in studio.

Yes.  We were discussing the ability of CDs to reproduce realistic dynamic range of music.  You observed that there is a relatively narrow effective dynamic range of live rock music played against an environmental noise floor of ~70dB.  I responded that your observation was not relevant to the matter under discussion:  recorded music, with a very low background noise level rather than the loud background noise level caused by thousands of noisy fans at a rock concert.

[Mark]

I must say, chaps, that I've been thoroughly enjoying this unfolding discussion. And FWIW, I agree about the 'vinyl nostalgia' that blinds some to the virtues of CD sound. Yes, early CDs may have been engineered in too clinical a way that meant some of the 'warmth' so often referred to by vinyl fanboys was lost. But can this still be said to be true? No, it can't. There are now so many top-end players and amplifiers able to deliver superb highs and lows that really, this whole debate should be something of a non-starter.