Creating examples of "Loudness Wars" effect

[TheSonicTruth]

Sorry, I was thinking of Brothers in Arms. I thought Sultans of Swing was on the same album as Money for Nothing, but it isn't. SA-CD.net is a good place to find multichannel stuff.

That betrays your age. No wonder you so vigorously defend modern 'hot' mastering. It probably defined the era of music you grew up listening to. Isn't it also great to have your phone in your pocket, instead of having to find one of those glass & metal boxes to go inside and make a call, and drop change in the phone?

 

[bigshot]

That betrays your age. No wonder you so vigorously defend modern 'hot' mastering. It probably defined the era of music you grew up listening to. Isn't it also great to have your phone in your pocket, instead of having to find one of those glass & metal boxes to go inside and make a call, and drop change in the phone?

I have no idea what point you're trying to make here. Is this ageism or something?

 

[Zapp_Fan]

Sorry, I can't agree with the above. You have no idea exactly what's going on in loudness processing during mastering. The game is, how loud can I make it and still sound ok? It's not just compressors, or just limiters, or clippers, it's a very complex combination of all of that, layered, stacked, spectrally split and recombined...really, really complex. Compressors play a vital role, but hardly are the only mechanism.

Sorry, I guess I was not clear. My point was that if people were for some reason not allowed to use compressors or limiters, they would still find many ways to increase perceived loudness and RMS. Actually I don't think we disagree on that point.

And I do have many years using those tools (minored in music technology in college, along with many years as a crappy, but technically inclined amateur musician) although not at a professional level. So I will defend myself by asserting that I have more than the average punter's clue about how these processes work, anyway.

Regarding your point about not blaming the engineers for crappy remasters: 100%. The only way to get this to change is to affect the revenues of the labels, and then complain so they know WHY their revenues are dropping. And even then, they are not exactly known for accepting the wisdom of the consumer...

 

[bigshot]

Regarding your point about not blaming the engineers for crappy remasters: 100%. The only way to get this to change is to affect the revenues of the labels, and then complain so they know WHY their revenues are dropping. And even then, they are not exactly known for accepting the wisdom of the consumer...

I don't think it's the labels' fault either. They're just giving their audience what it wants. People who buy or stream a song at a time, rather than playing an album all the way through, have different needs than old school listeners. If I was listening to popular music that way, I'd want it all evened out dynamically so I don't have one song too quiet and the next one too loud. Basically, technology has made people listen to music differently. They aren't sitting on a couch listening to a 20 minute record side all the way through any more. They're pulling out the songs they like and shuffling them while they're walking down a busy street. In that sense, it's closer to the way 45s in a jukebox were played... and 45s were highly compressed so they would all sound consistently loud over the din of a bar. The audience wants the music loud and they're listening in loud environments. What else can the labels do?

I had a musician friend ask me to do a quick master for him. I talked with him a little about what he wanted. He told me he wanted to preserve the dynamics... go easy on the compression. I said OK and I mastered it for my home speaker system. I gave him the tracks and he burned a CD and said he'd listen to it. The next day he called me and said that he was listening to it in his car and the CD I mastered for him sounded much quieter than his other CDs. You can't have it both ways! I imagine that mastering engineers get handed all kinds of contradictory demands. They have to balance it according to what they are being told and temper that according to what the audience needs. I'm sure there are a lot of compromises that need to be made to get something that works.

 

[TheSonicTruth]

I don't think it's the labels' fault either. They're just giving their audience what it wants. People who buy or stream a song at a time, rather than playing an album all the way through, have different needs than old school listeners. If I was listening to popular music that way, I'd want it all evened out dynamically so I don't have one song too quiet and the next one too loud. Basically, technology has made people listen to music differently. They aren't sitting on a couch listening to a 20 minute record side all the way through any more. They're pulling out the songs they like and shuffling them while they're walking down a busy street. In that sense, it's closer to the way 45s in a jukebox were played... and 45s were highly compressed so they would all sound consistently loud over the din of a bar. The audience wants the music loud and they're listening in loud environments. What else can the labels do?

I had a musician friend ask me to do a quick master for him. I talked with him a little about what he wanted. He told me he wanted to preserve the dynamics... go easy on the compression. I said OK and I mastered it for my home speaker system. I gave him the tracks and he burned a CD and said he'd listen to it. The next day he called me and said that he was listening to it in his car and the CD I mastered for him sounded much quieter than his other CDs. You can't have it both ways! I imagine that mastering engineers get handed all kinds of contradictory demands. They have to balance it according to what they are being told and temper that according to what the audience needs. I'm sure there are a lot of compromises that need to be made to get something that works.

Bigshot what you said here made more sense here than anything else I've heard you say on this subject!

So please don't blow and say something like, "the remaster is better", like some other engineers on here keep saying. It really is a case by case basis, but I still maintain, as in my profile pic: turning em up ain't remastering!

 

[bigshot]

Maybe now you're actually listening to what I say instead of trying to lead me into saying something I don't mean. I've been saying the same thing all along. You just have to ask the right questions. Not leading ones.

 

[pinnahertz]

Sorry, I guess I was not clear. My point was that if people were for some reason not allowed to use compressors or limiters, they would still find many ways to increase perceived loudness and RMS. Actually I don't think we disagree on that point.

Yes, we do disagree. Compressors and limiters are essential elements of a total loudness processing system. Take them away and all you can do results in gross distortion.

And I do have many years using those tools (minored in music technology in college, along with many years as a crappy, but technically inclined amateur musician) although not at a professional level. So I will defend myself by asserting that I have more than the average punter's clue about how these processes work, anyway.

Um...no comment.

Regarding your point about not blaming the engineers for crappy remasters: 100%. The only way to get this to change is to affect the revenues of the labels, and then complain so they know WHY their revenues are dropping. And even then, they are not exactly known for accepting the wisdom of the consumer...

Money talks, everything else walks. If people stopped buying loudly mastered recordings in favor of less processed versions, the practice would die a quick death. That won't happen primarily because people are not given a choice.

 

[pinnahertz]

I don't think it's the labels' fault either. They're just giving their audience what it wants. People who buy or stream a song at a time, rather than playing an album all the way through, have different needs than old school listeners. If I was listening to popular music that way, I'd want it all evened out dynamically so I don't have one song too quiet and the next one too loud. Basically, technology has made people listen to music differently. They aren't sitting on a couch listening to a 20 minute record side all the way through any more. They're pulling out the songs they like and shuffling them while they're walking down a busy street. In that sense, it's closer to the way 45s in a jukebox were played... and 45s were highly compressed so they would all sound consistently loud over the din of a bar. The audience wants the music loud and they're listening in loud environments. What else can the labels do?

Just a small point, but 45s were all mastered quite differently, so much so that juke boxes had compressors in them to even things out, so that the would all sound consistently loud over the din of a bar.

I had a musician friend ask me to do a quick master for him. I talked with him a little about what he wanted. He told me he wanted to preserve the dynamics... go easy on the compression. I said OK and I mastered it for my home speaker system. I gave him the tracks and he burned a CD and said he'd listen to it. The next day he called me and said that he was listening to it in his car and the CD I mastered for him sounded much quieter than his other CDs. You can't have it both ways! I imagine that mastering engineers get handed all kinds of contradictory demands. They have to balance it according to what they are being told and temper that according to what the audience needs. I'm sure there are a lot of compromises that need to be made to get something that works.

And there it is, in the flesh! The musician had, in his hand, the solution to his problem and he chose not to use it in favor of an alternate, but inferior method. The solution to a quieter recording was the volume control, and he refused to consider that as an option.

There's why the loudness war will have no end. Because it can't. The global reference is loud-processed, and who wants to loose that contest?

 

[gregorio]

[1] Yes, we do disagree. Compressors and limiters are essential elements of a total loudness processing system. Take them away and all you can do results in gross distortion.
[2] Money talks, everything else walks. If people stopped buying loudly mastered recordings in favor of less processed versions, the practice would die a quick death. That won't happen primarily because people are not given a choice.
[3] The solution to a quieter recording was the volume control, and he refused to consider that as an option.
[4] There's why the loudness war will have no end. Because it can't.

1. To be honest, it would be possible to achieve the same or pretty close to the same loudness without a compressor or limiter. You could just use volume automation to raise the level of everything except the peaks. It would be a hell of a lot of work to write in all that volume automation and of course you wouldn't get the desirable colouration and/or non-linear effects many compressors and limiters are prized for but you could achieve a very similar total loudness with no gross distortion.

2. Again, I don't agree. If people were given the choice they would NOT choose the less loudly mastered recordings. Some/Many of the current music genres are compositionally designed for extreme compression. Take away that loud mastering and the pieces just don't work, they'd be un-listenable. The problem is when pieces/songs not so well structurally designed for extreme compression are nevertheless required to attain the same loudness as those that are.

3. And with some justification. Most people listen to music, at least some of the time, if not all the time, in circumstances where changing the volume control is not practical, say when driving, travelling or working. And even when it is physically practical, it's often still not desirable, the volume control would only be changed when the listener realises the song/track is too loud or too quiet, which typically would be some way into the song and then you've damaged the listening experience as you've created a very off-putting level change in the middle of the song.

4. I think it is likely to end, as replay-gain and loudness normalisation becomes more ubiquitous but it's going to take a long time because: A. There is no sign of a loudness normalisation level becoming ubiquitous. B. The tools don't exist to instantly compare what happens to the track when it's loudness normalised and C. The composition/structure of songs will then change to reflect the loudness normalisation paradigm but that effectively means the evolution of some/many current genres into new genres.

G

 

[pinnahertz]

1. To be honest, it would be possible to achieve the same or pretty close to the same loudness without a compressor or limiter. You could just use volume automation to raise the level of everything except the peaks. It would be a hell of a lot of work to write in all that volume automation and of course you wouldn't get the desirable colouration and/or non-linear effects many compressors and limiters are prized for but you could achieve a very similar total loudness with no gross distortion.

2. Again, I don't agree. If people were given the choice they would NOT choose the less loudly mastered recordings. Some/Many of the current music genres are compositionally designed for extreme compression. Take away that loud mastering and the pieces just don't work, they'd be un-listenable. The problem is when pieces/songs not so well structurally designed for extreme compression are nevertheless required to attain the same loudness as those that are.

3. And with some justification. Most people listen to music, at least some of the time, if not all the time, in circumstances where changing the volume control is not practical, say when driving, travelling or working. And even when it is physically practical, it's often still not desirable, the volume control would only be changed when the listener realises the song/track is too loud or too quiet, which typically would be some way into the song and then you've damaged the listening experience as you've created a very off-putting level change in the middle of the song.

4. I think it is likely to end, as replay-gain and loudness normalisation becomes more ubiquitous but it's going to take a long time because: A. There is no sign of a loudness normalisation level becoming ubiquitous. B. The tools don't exist to instantly compare what happens to the track when it's loudness normalised and C. The composition/structure of songs will then change to reflect the loudness normalisation paradigm but that effectively means the evolution of some/many current genres into new genres.

G

1. My use of automation has been fairly limited, but it seems it would be difficult and extremely time consuming to create the equivalent of a fast peak limiter that way. Not impossible.

2. Agreed, much contemporary music is structured to work intentionally with loud mastering. I think there could be exceptions, and I think that using less globally, particularly some of the short time-constant stuff, may satisfy some of the desire for less processing while still maintaining listenable averages. In other words, I don't believe loudness processing is a binary decision.

3. As in 2., I agree, but I also think that making the volume of a single track listenable is different that competitive processing, and there's certainly a bit of compromise that could be made without damaging the listening experience.

4A. Agreed, but use is increasing by means of certain play-out devices and apps default settings. Most users never change defaults. Apple defaulted Soundcheck to ON when they started their streaming service via iTunes. Not ubiquitous, but going toward that enough that it might be good to consider the results.

4B. You can turn Soundcheck on and off within any iTunes or iOS device while playing and immediately hear what's going on in terms of gain change.

4C. Agreed. Hopefully demonstrating what replay-gain/Soundcheck does will have some impact. I'm not optimistic, though.

As you know, there's a big difference between what happens with a volume control, replay-gain, and what's going on with loudness-war mastering. Replay-gain is essentially a volume control pre-set based the result of a loudness algorithm scan of a track or album, and is a fixed gain adjustment. It does set apparent average volume, but does not change dynamics. Same with a plain old volume control. Loudness-processing changes a whole lot of things dynamically, and again, can be extremely helpful, but also abused and destructive. As long as the war is competitive (and that's the unfortunate part) there will be a drive toward abusing the tools. And "abuse" is entirely subjective! One mans abuse is another's gold record. I just don't see any of this changing, and I'm very sure most listeners don't care anyway.

On the other hand, those who are vocally against the loudness-war...many don't understand either the mechanisms or the goals. They look at a loudness evaluation app and it imposes a bias that emphasizes their opinion, even though the app/meter is not entirely accurate. The loudness database...not sure that's doing anyone any good. It's cited as proof over and over. But music itself has changed as the tools changed, and that's not what is reflected.

But there's the issue of adding additional processing to older recordings during re-release. That one seems unnecessary, or at least, over done. What do you think about that?

 

[Steve999]

I have a question. For real. This is taking things down to a much lower level but I do have a question. If I turn on soundcheck or replay gain in Apple Music of Foobar2000 (I am now running both side by side) am I altering audio my files? If so, is it just the metadata? If not, does the software keep a separate database? Or is it all done in real time? One of my worries is that if I use SoundCheck or ReplyGain it will alter my audio files, or at least the metadata in my audio files, and then when I run a backup it will take a million years because all of the files are changed.

4A. Agreed, but use is increasing by means of certain play-out devices and apps default settings. Most users never change defaults. Apple defaulted Soundcheck to ON when they started their streaming service via iTunes. Not ubiquitous, but going toward that enough that it might be good to consider the results.

4B. You can turn Soundcheck on and off within any iTunes or iOS device while playing and immediately hear what's going on in terms of gain change.

4C. Agreed. Hopefully demonstrating what replay-gain/Soundcheck does will have some impact. I'm not optimistic, though.

As you know, there's a big difference between what happens with a volume control, replay-gain, and what's going on with loudness-war mastering. Replay-gain is essentially a volume control pre-set based the result of a loudness algorithm scan of a track or album, and is a fixed gain adjustment. It does set apparent average volume, but does not change dynamics. Same with a plain old volume control. Loudness-processing changes a whole lot of things dynamically, and again, can be extremely helpful, but also abused and destructive. As long as the war is competitive (and that's the unfortunate part) there will be a drive toward abusing the tools. And "abuse" is entirely subjective! One mans abuse is another's gold record. I just don't see any of this changing, and I'm very sure most listeners don't care anyway.

 

[castleofargh]

I have a question. For real. This is taking things down to a much lower level but I do have a question. If I turn on soundcheck or replay gain in Apple Music of Foobar2000 (I am now running both side by side) am I altering audio my files? If so, is it just the metadata? If not, does the software keep a separate database? Or is it all done in real time? One of my worries is that if I use SoundCheck or ReplyGain it will alter my audio files, or at least the metadata in my audio files, and then when I run a backup it will take a million years because all of the files are changed.

the file has only the metadata changed, not the way the signal is encoded. but during playback, the signal sent to the DAC is attenuated(or boosted sometimes) digitally. so it's the same as if you lowered the volume with foobar's volume slider, except that it's a different setting for each song.

 

[Steve999]

Thanks. My worry is that if the metadata is changed when I apply SoundCheck or replygain the next time I run an incremental backup (which I do quite often) it will re-back-up all of my music files. Is this a legitimate concern? It sounds like it is. I suppose that would not be terrible but just a one-time inconvenience.

Also, another question just occurred to me: do replaygain and SoundCheck play nice with each other if you apply them to the same music file?

And another: Does SoundCheck work with Apple TV?

the file has only the metadata changed, not the way the signal is encoded. but during playback, the signal sent to the DAC is attenuated(or boosted sometimes) digitally. so it's the same as if you lowered the volume with foobar's volume slider, except that it's a different setting for each song.

 

[castleofargh]

a metadata is one value, if you run the scan again it will give the same value(unless you change some settings). if the data wasn't there it will be added after a scan. if it already was there, for foobar warns you that you already have it. all of this is harmless even if you replace a tag value by another a billion times, all you will have in the end is that metadata tag with the latest value and absolutely no other change.

I'm apple illiterate because I'm a hater, so I can't help about that.

some old DAPs could bug when they were presented with metadata tags they didn't know, but it's rare nowadays and I wouldn't expect this to happen on any half ok computer app.

 

[bfreedma]

Thanks. My worry is that if the metadata is changed when I apply SoundCheck or replygain the next time I run an incremental backup (which I do quite often) it will re-back-up all of my music files. Is this a legitimate concern? It sounds like it is. I suppose that would not be terrible but just a one-time inconvenience.

Also, another question just occurred to me: do replaygain and SoundCheck play nice with each other if you apply them to the same music file?

And another: Does SoundCheck work with Apple TV?

The chances are that you will do a full one time backup of the files after you add metadata. Without getting into too much minutia, it depends on how you’ve configured the backup software. Most consumer backup products default to file level backup - when the file is flagged as changed, the entire file is backed up. If you’re backing up at the block level rather than file level and your software supports it, the backup will only run against the storage blocks that changed within the file resulting in a smaller amount of data being backed up post metadata insertion and change.

Real world, I wouldn’t worry unless you’re backing up to a cloud service with significantly above market pricing for data transmission and storage rates.

How many TBs of music are you working with and what backup solution are you using? OS, Software, and storage model (local or cloud).