Is most music now made for headphones

[coppice]

Back in the 60s, 70s and 80s I listened to music through both headphones and speakers. From the early 90s until a few days ago I used headphones almost all the time. Finally I have a nice pair of speakers again, and I have been listening to a variety of things over the last few days. Older recordings generally sound very good. However, I am finding that a quite a few recordings from the last few years, that I am used to hearing sound good through headphones, can sound really odd through these speakers. Especially things like drums. Is headphone reproduction now the only thing recording engineers focus on?

[Jeroen3]

I think you're experiencing the loudness war.

[nfmax]

Those of us with unilateral deafness  - about 5% of the population have it to some extent - hope not!

[coppice]

I think you're experiencing the loudness war.

Nope. That's a completely separate issue. The loudness war takes away all the dynamics, and adds clipping distortion. That sounds just as bad, whether you use headphones or speakers.

[DrG]

Remember these?

[fourfathom]

I think you're experiencing the loudness war.

Nope. That's a completely separate issue. The loudness war takes away all the dynamics, and adds clipping distortion. That sounds just as bad, whether you use headphones or speakers.

I agree that the loudness war results in reduced dynamic range (and the music suffers for this), but is actual clipping still a problem?  In the early days of CDs and other digital media, producers who were used to relying on the soft (and musically fairly benign) clipping of magnetic tape would overdrive the digitizers and get horrible hard-clipping.  But this problem was eventually recognized and processing techniques were developed to provide low-distortion compression or other dynamic-range reduction, so hard-clipping shouldn't be much of an issue now (I presume).

Dynamic range reduction is entirely appropriate when in a high-noise listening environment such as cars, or when using tiny headphones/earbuds in noisy places.  It's only in good listening environments that we can take advantage of high dynamic range material. otherwise the low-level passages are buried in the ambient noise.

I have no idea if the OP's observation is related to this.  I wouldn't be surprised if some sort of binaural spatial processing is currently being used in modern music, especially in genres that are typically "enjoyed" with earbuds.  I put "enjoyed" in quotes because I don't really enjoy much of the modern pop music -- not because of the sound quality, but because there is a pattern of reducing the melody to a very basic note relationships and a simplification of the song structure.  There are obviously exceptions, and thank god for that.

[borjam]

Unfortunately most music is mastered folllowing the orders of idiotic producers who want it to be louder than the rest. So it's a disgusting soup of distortion.

Never before in history recording equipment has been better, yet released recordings are awful. Listening to an old recording such as a Pink Floyd release on vinyl can be a really surprising experience, and not because vinyl is superior to CD (it's not!) but because releases on CD are utterly butchered.

[fourfathom]

[...] So it's a disgusting soup of distortion. [...]

But is this distortion due to the compression (which does not have to add harmonic distortion or similar artifacts), or is it because that's the "sound" the producer is deliberately looking for?  For example the use of auto-tune as an effect, and just in general, adds artifacts that really annoy me but are not the same thing as overload-related distortion.  I would appreciate seeing an analysis of modern recordings -- I don't know for sure the answer here.

I found it interesting that "underload" was also a problem with early CD digitization, which added to the dynamic range complaints.  With vinyl, very small signals would gently fade into the ever-present background noise (or hiss, with tapes).  With CDs, as the signal became weaker the effective number of bits (ENOB) would diminish, adding harmonic distortion to these weak signals.  At the lowest levels, you would have only one or two bits, essentially turning a sinewave into a squarewave.  Once this issue was recognized, they started adding carefully-shaped digital pseudonoise to the digitizer, which shifted the artifacts up in frequency to where they could be filtered out.  We've discussed a similar concept when analyzing the performance of low-resolution SDRs, where the normal atmospheric and thermal noise create the same improvement in ENOB.  Modern CDs and other digital formats sound better because of this.

[coppice]

I think you're experiencing the loudness war.

Nope. That's a completely separate issue. The loudness war takes away all the dynamics, and adds clipping distortion. That sounds just as bad, whether you use headphones or speakers.

I agree that the loudness war results in reduced dynamic range (and the music suffers for this), but is actual clipping still a problem?  In the early days of CDs and other digital media, producers who were used to relying on the soft (and musically fairly benign) clipping of magnetic tape would overdrive the digitizers and get horrible hard-clipping.  But this problem was eventually recognized and processing techniques were developed to provide low-distortion compression or other dynamic-range reduction, so hard-clipping shouldn't be much of an issue now (I presume).

There are a number of resources on line showing waveforms from some famous recordings that have been reissued as remasters, so there are both older and newer CDs available. In many cases the peak volume has been cranked up so there is a huge amount of clipping on the newer version, but very little on the older one.

Dynamic range reduction is entirely appropriate when in a high-noise listening environment such as cars, or when using tiny headphones/earbuds in noisy places.  It's only in good listening environments that we can take advantage of high dynamic range material. otherwise the low-level passages are buried in the ambient noise.

There is certainly a place for dynamic reduction in many situations. Some classical recording making a remote volume control highly valuable, as in anything but a very quiet room you need to turn up the quiet passages so you can hear them, and when it gets loud you need to turn it down so the family and neighbours don't complain. However, a lot of the dynamic range compressions on modern discs is just ridiculous. Song known for getting much of their mood and impact from their dynamics have ended up with nearly a flat volume throughout the song. There is a good argument for recordings carrying multiple sets of dynamics information, to be able to switch the playing to suit the environment.

I have no idea if the OP's observation is related to this.  I wouldn't be surprised if some sort of binaural spatial processing is currently being used in modern music, especially in genres that are typically "enjoyed" with earbuds.  I put "enjoyed" in quotes because I don't really enjoy much of the modern pop music -- not because of the sound quality, but because there is a pattern of reducing the melody to a very basic note relationships and a simplification of the song structure.  There are obviously exceptions, and thank god for that.

That's the kind of thing I think may be going on. With headphones I am hearing a drum kit spread out in front of me. With speakers I hear 2 point sources that don't blend. Older recordings give more of a sense of a spread out drum kit in both cases.

[Electro Fan]

Back in the 60s, 70s and 80s I listened to music through both headphones and speakers. From the early 90s until a few days ago I used headphones almost all the time. Finally I have a nice pair of speakers again, and I have been listening to a variety of things over the last few days. Older recordings generally sound very good. However, I am finding that a quite a few recordings from the last few years, that I am used to hearing sound good through headphones, can sound really odd through these speakers. Especially things like drums. Is headphone reproduction now the only thing recording engineers focus on?

It's probably not the only thing but no doubt it is an important thing.

When music is played by the original musicians (live or in a studio) they have an idea of what they want the music to sound like.  Then the recording engineers take over with their skill (including their objectives and preferences) with the recording technology to produce the final product.  Along the way, someone has to determine who the customers will be and where/how they are going to listen.  It's very possible that from the 1950s to 1990s that many customers listened with a hifi system in some room in their home.  Every system would be a little different and the rooms would vary in many respects (size and shape including relative dimensions of the walls, floor, and ceiling, surface materials, furnishings etc.)  I have long been of the opinion that in addition to a turntable/tone arm/cartridge or a disc and the pre amp, amp, and speakers that the room is a major contributor to what the listener will experience.  The room can make a very big difference in the ability to support and discern frequency response and a sense of spatial imaging.  Nonetheless, from the 50s' to 90's including the transition from analog to digital music producers were pretty sure that most of their music was either going to be heard on the radio (kind of a gateway drug) or a home stereo system.  So in an effort to sell more music music producers mixed up their final product with some idea of the target acoustical environment:  a room in a home.

Since the 1990s when hifi systems started precipitously losing ground to computers - especially in terms of young people's interest - and more recently with the phone becoming the center of people's musical library and mode of listening, music producers have had to figure out how to adapt to the acoustical environment of the target market.  In the old days it might have been possible to make music mixed well for two speakers in a living room or family room or bedroom and maybe also a pair of Koss headphones.  Now I'm betting a lot of music producer are saying "we hope we still have people with their hifi systems enjoying our music but we can't afford to miss the market for ear buds".  And as good as Apple ear buds are, frankly they won't sound as intriguing as a good high end hifi system - but you can't optimize for everything so something has to give.  A room (not to mention room to room variations) and the area in and around your ear are simply different acoustic environments.

The good news is that much of the best music ever produced was made in the 1960s and 1970s and the original versions will likely live forever, and if you like that music and you want the Beatles, the Rolling Stones, Led Zeppelin, and Jimi Hendrix to grace your home - just build a decent hifi system and you can get a sense of what it was all about.  Same with lots of jazz and other artists from the 50s and 60s.  So, yeah, I think what you are hearing is a refection of a change in how/where people listen to music.   

[Messtechniker]

Listening habits have changed from focussed to casual/background and so the mastering target has also changed from HIFI to kitchen/bedroom radio, car radio, cheap computer speakers and ear buds/phones. Since I am in the comfortable position of doing upfront recordings and if I like the performance, I do a second master specifically for my old refurbed big fat Tannoys, besides the main master for whatever low-fi target.

Lately I have had some visitors only used to this casual low-fi stuff, where I had to stop the listening session because they suffered heart problems. They complain that the music is much to much in their face. And this at slightly lower than normal levels. Me thinks an old recording of the Bolero or the Khachaturian Sabre Dance - without touching the volume control - would kill them. Nowadays I usually ask whether or not they feel up to it.

[Gyro]

Ah, somebody else with nice old big fat Tannoys!

I'm currently going through the reverse experience, having recently picked up a pair of '70s Wharfedale Isodynamic headphones at a local emporium (I'd sadly trashed my original pair many years ago). Luckily most of my material is older but even more modern releases seem to have an impossibly wide soundstage, where I would have maybe expected them to be 'narrowed down' a bit to suit headphone listening, maybe that's how people with earbuds expect the soundstage to be these days.

Surprisingly (or maybe unsurprisingly?), with decent amplification the Isodynamics sound very similar to the Tannoys, aside from the sound staging obviously. I don't see me wearing them, and the associated amplification, in the street anytime soon though! (sorry, I digress).

[coppice]

I'm currently going through the reverse experience, having recently picked up a pair of '70s Wharfedale Isodynamic headphones at a local emporium (I'd sadly trashed my original pair many years ago).

I remember those things as sounding very good, but not being the most comfortable things to wear for extended periods. I wonder why that principle has not been used more often. It seemed a sound alternative to electrostatic headphones.

[xmo]

I think Electro Fan said it well.  Producers and recording engineers want their product to sound good to the widest range of customers.

To me, newer music does sound good with some directionality and 'space' around the performance but earlier recordings have a much more expansive sound stage.

On the right system, some recordings are incredible. According to one reviewer: "the speakers create a 3-D sound stage that when you close your eyes there are no walls, no speakers just the instruments and singers"

[Gyro]

I'm currently going through the reverse experience, having recently picked up a pair of '70s Wharfedale Isodynamic headphones at a local emporium (I'd sadly trashed my original pair many years ago).

I remember those things as sounding very good, but not being the most comfortable things to wear for extended periods. I wonder why that principle has not been used more often. It seemed a sound alternative to electrostatic headphones.

They do still exist in an altered form, often referred to as 'Orthodynamc'. After the originals (ID1) with their large area corrugated Polyimide diaphragms, Rank developed with the ID2 / Leak 3000, which had a smaller circular diaphragm [Edit: with a centre termination contact, effectively a smaller excursion ring radiator Those were NAD/Yamaha] in an on-ear format. They didn't review as well and Rank sold the design to Fostex, who went on to produce several models, as did Yamaha, but none followed the original ID1 'pseudo electrostatic' large diaphragm format. There are also several tweeters that claim to use the isodynamic / orthodynamic principle (in fact Wharfedale used one for a time).

As far as comfort, the original plasticised 'faux leather' over-ear earpad and headband coatings broke down over time and either flaked (as my originals) or turned to a sticky mess (my newly found ones). Fortunately all of that dissolves away in IPA to reveal a nice velvety fabric underneath! Replacing the original foam with 1/2" memory foam makes for a very comfortable and cosseting experience.

P.S. There are actually a pair of ID1s in the V&A Museum, together with a Boothroyd -Stewart Lecson Audio AC1 Pre/power amp setup. http://collections.vam.ac.uk/item/O1194711/wharfedale-isodynamic-headphones-headphones-rank-radio-industrial/

[coppice]

As far as comfort, the original plasticised 'faux leather' over-ear earpad and headband coatings broke down over time and either flaked (as my originals) or turned to a sticky mess (my newly found ones). Fortunately all of that dissolves away in IPA to reveal a nice velvety fabric underneath! Replacing the original foam with 1/2" memory foam makes for a very comfortable and cosseting experience.

That was true innovation. In the early 70s most faux leather headphone pads lasted for years. Sometimes they stiffened, but they remained usable. Wharfedale was ahead of the curve in developing pads that need frequent replacement.  Now the only pads that last are the velvet ones, on things like my AKG K712s.

The isodynamic headphones were interesting against things like the Stax models of the time. They had some of the good qualities of the Stax sound, but could achieve greater volume. I found the most satisfying headphones of that period were the Pioneer electret ones I used to own. They had the good qualities of Stax, but capable of any volume your ears could stand.

[madsbarnkob]

Back in the 60s, 70s and 80s I listened to music through both headphones and speakers. From the early 90s until a few days ago I used headphones almost all the time. Finally I have a nice pair of speakers again, and I have been listening to a variety of things over the last few days. Older recordings generally sound very good. However, I am finding that a quite a few recordings from the last few years, that I am used to hearing sound good through headphones, can sound really odd through these speakers. Especially things like drums. Is headphone reproduction now the only thing recording engineers focus on?

It is a matter of audience.

The last 10 years (or more, getting older goes fast) the main stream music production is used on mobile devices. No-one really has a set of stereo speakers anymore. You might just find modern age BOOMBOXES now.

So music is mastered for the medium the audience has, head phones.

As the owner of a pair of JBL 4333, I am experiencing the same issues as you. With 15" Bass drivers it is perfectly clear when something was mastered on studio monitors or head phones.

[Gyro]

As far as comfort, the original plasticised 'faux leather' over-ear earpad and headband coatings broke down over time and either flaked (as my originals) or turned to a sticky mess (my newly found ones). Fortunately all of that dissolves away in IPA to reveal a nice velvety fabric underneath! Replacing the original foam with 1/2" memory foam makes for a very comfortable and cosseting experience.

Wharfedale was ahead of the curve in developing pads that need frequent replacement. 

 

I never manage to hear a pair of electrets, I do remember them becoming popular due to the lack of Stax style energiser units. I wonder how their diaphragms lasted long term.

[wraper]

I think you're experiencing the loudness war.

Nope. That's a completely separate issue. The loudness war takes away all the dynamics, and adds clipping distortion. That sounds just as bad, whether you use headphones or speakers.

Clipping is present only in worst cases. However you complained about drums which certainly sound crappy in such compressed music. There is no punch whatsoever. Loudness war is averaging everything to the same level so everything becomes un-distinguished mud.

That sounds just as bad, whether you use headphones or speakers.

Not necessarily. It highly depends on what it's played. Say on laptop or phone speakers such crap sounds the best. And even when headphones are used, phones tend to exaggerate low frequencies. While on decent system it sounds like trash because you suddenly can hear flaws in it.

[wraper]

I think Electro Fan said it well.  Producers and recording engineers want their product to sound good to the widest range of customers.

And that widest range is the trashiest range as well. I have a good system, and certainly can hear some pleasure destroying artifacts in many tracks. And the same tracks don't have those issues when I play them on some trash with mobile oriented class-D amp.

[borjam]

[...] So it's a disgusting soup of distortion. [...]

But is this distortion due to the compression (which does not have to add harmonic distortion or similar artifacts), or is it because that's the "sound" the producer is deliberately looking for?  For example the use of auto-tune as an effect, and just in general, adds artifacts that really annoy me but are not the same thing as overload-related distortion.  I would appreciate seeing an analysis of modern recordings -- I don't know for sure the answer here.

Compression is non linear by definition. Brick wall limiting can make itself really obvious.

The main problem is, they want it loud. So they push the levels to the limit, so loud that they even overflow the D/A converter.

Of course distortion is a legitimate effect to use in music creation. It's often used on purpose from subtle (microphones and preamplifiers with valve stages) to obvious (electric guitars and synthesizers). But the aim of the loudness war is not to offer a better sound or an intentionally distorted one, but being just louder.

A chilling "technical achievement" I heard is one of the latest releases by the Red Hot Chilli Peppers (monarchy of Roses). It seems to be mastered in a way that is intended to be played with the D/A distortion caused by overloading the D/A. Try reducing the volume digitally (ie, for example lowering the volume on iTunes) so that it doesn't overload the D/A and behold the result.

I found it interesting that "underload" was also a problem with early CD digitization, which added to the dynamic range complaints.  With vinyl, very small signals would gently fade into the ever-present background noise (or hiss, with tapes).  With CDs, as the signal became weaker the effective number of bits (ENOB) would diminish, adding harmonic distortion to these weak signals.  At the lowest levels, you would have only one or two bits, essentially turning a sinewave into a squarewave.  Once this issue was recognized, they started adding carefully-shaped digital pseudonoise to the digitizer, which shifted the artifacts up in frequency to where they could be filtered out.  We've discussed a similar concept when analyzing the performance of low-resolution SDRs, where the normal atmospheric and thermal noise create the same improvement in ENOB.  Modern CDs and other digital formats sound better because of this.

There is no free lunch, that's a universal law!

Dithering is a very good technique but of course you add a bit of noise. Noise that you can shape so that it's really hard to hear, so at the end of the day most of the noise you hear when playing music is not caused by the dithering process, but the analog stages themselves (microphones, preamps, mixers, equalizers, and your own playback equipment).

Anyway all of this technical progress has been thrown down the drain with the loudness war.

And no, nowadays producers don't want recordings to sound good on every device. They just want it loud!

But of course you can't generalize. There are properly mastered classical and jazz recordings out there. And many listeners will complain that the volume is incredibly low.

[fourfathom]

Compression is non linear by definition.

Well, yes, but this doesn't necessarily result in audible distortion.  If the compression is look-ahead or anticipatory (I made up those terms, don't know what the professionals call it), then gain can be slowly reduced before the dynamics peaks hit, then restored afterwards.  Depending on the adjustment rate this may lead to "breathing" or other noticeable artifacts, but not harmonic distortion within the audio frequency range.  More rapid compression will become limiting and distort the lower frequency components, and faster still becomes clipping, distorting everything.

But I'm being pedantic.  I agree that these dynamics processing techniques can really hurt the recorded sound.  But as has been said, the producers want to tailor the sound to suit the listening environment.  I recall in the 1960's and '70's producers would take the mix, put it on a cassette (or cartridge) tape and review it in their car while driving around town.  This obviously led to heavy compression -- a "living room" mix would be buried under the ambient noise.

[SiliconWizard]

Back in the 60s, 70s and 80s I listened to music through both headphones and speakers. From the early 90s until a few days ago I used headphones almost all the time. Finally I have a nice pair of speakers again, and I have been listening to a variety of things over the last few days. Older recordings generally sound very good. However, I am finding that a quite a few recordings from the last few years, that I am used to hearing sound good through headphones, can sound really odd through these speakers. Especially things like drums. Is headphone reproduction now the only thing recording engineers focus on?

Not all, but at least most "mainstream" music is mastered these days with a priority for listening devices that people use most statistically. This sucks, but makes sense.

So, that's basically headphones (and often cheap ones), and crappy speakers. Not your typical "HiFi" gear that has (unfortunately) almost completely disappeared in most homes.

You can still find some producers that release music (for some of their artists, far from all) with different mastering approaches depending on the medium - typically one for CDs, one for compressed audio (such as MP3), and sometimes one for high-res digital (usually 24-bit audio), the latter sounding much better on good gear (provided it's indeed mastered specifically for this and not just re-encoded crap made for CDs.)

[SilverSolder]

Basically, these days, you have to get the "good stuff" specifically for hi-fi use (e.g. hdtracks.com)

But Amazon is talking about making their streaming service high definition.  Not sure what that means in practice, but they claim better-than-CD quality.

I use soundcloud.com frequently, the streaming seems to be very high quality there.

[borjam]

Compression is non linear by definition.

Well, yes, but this doesn't necessarily result in audible distortion.  If the compression is look-ahead or anticipatory (I made up those terms, don't know what the professionals call it), then gain can be slowly reduced before the dynamics peaks hit, then restored afterwards.  Depending on the adjustment rate this may lead to "breathing" or other noticeable artifacts, but not harmonic distortion within the audio frequency range.  More rapid compression will become limiting and distort the lower frequency components, and faster still becomes clipping, distorting everything.

I was being a bit pedantic as well. The main source of undesirable harmonic distortion in recordings nowadays is D/A clipping because it's driven too hard.

Compression is not bad per se. It's a fantastic tool when used properly. Both as a way to reduce dynamics so that your recording can be listened to in realistic conditions and as a creative tool of course.

So what is realistic conditions? Try to listen to "Urban Bushmen" by the Art Ensemble of Chicago (which is an outstanding recording full of life) at the street. You will miss most of it because there are passages with soft, subtle percussion sounds. So I understand compression in pop and rock music.

But the loudness war is not about music you can hear on portable players, but making it sound deafening even when a singer is almost whispering. Insanity!



But I'm being pedantic.  I agree that these dynamics processing techniques can really hurt the recorded sound.  But as has been said, the producers want to tailor the sound to suit the listening environment.  I recall in the 1960's and '70's producers would take the mix, put it on a cassette (or cartridge) tape and review it in their car while driving around town.  This obviously led to heavy compression -- a "living room" mix would be buried under the ambient noise.
[/quote]