[nick_charles]

  Scanned my CD rips and got a max of DR28 for this beaut (track is 11:00 long):

 
Conversion to both 128k and V0 lame mp3 doesn't change the DR rating.

 
Cool , what track is it ?

 

[RRod]

   
Cool , what track is it ?

 
Piece called «Ji II» off of this disc.

 

[Aaromtaar]

The DR rating will be preserved even then, done this test recently with Nik Bartsch's Mobile: Continuum, 2016 release from ECM Records. For mobile consumption I usually encode AAC (Appple, 256 kbps) but tried also lower rates. See the results below.
 
 
  

 

[reginalb]

  The DR rating will be preserved even then, done this test recently with Nik Bartsch's Mobile: Continuum, 2016 release from ECM Records. For mobile consumption I usually encode AAC (Appple, 256 kbps) but tried also lower rates. See the results below.
 
 
  

 
I guess this makes sense, given that the compression algorithm is using masking, it's going to be taking out silent sounds near the peaks, but leave in the longer silent sections, which isn't going to necessarily lower the dynamic range, even if you're looking at such a bludgeoningly low bit-rate. Dynamic range is just one measure of a playback file. I'd say it's a fairly important one today, but certainly not the only figure worth examining. 

 

[Aaromtaar]

Precisely, the DR algorithm relevant area of the spectrum, peak/RMS remains unaltered. Initially I was somewhat reluctant, considering the mp3 encoding algorithm hi-cut characteristics; totally agree with you on the importance of DR evaluation, but there's certainly more than meets the ear/brain.

 

[gregorio]

  Dynamic range is just one measure of a playback file. I'd say it's a fairly important one today

  totally agree with you on the importance of DR evaluation

 
Why?

 

[reginalb]

   
Why?

 
Well, for a few reasons. I have found when I like one version of an album more than another, it's usually one that gets a better score when I then run it through this test. The albums that have, for me, the most lasting appeal, those that I can really unwind to, are albums that again, score better. Now, this is of course subjective and could be influenced by all sorts of biases, I get that. But there is also at least some research on the topic suggesting that people prefer less compressed music, for example: 
 
http://scitation.aip.org/content/asa/journal/jasa/132/2/10.1121/1.4730881
 
http://www.robtoulson.rt60.co.uk/rt_docs/IN2010_BC-RT-JP_DynamicRange.pdf
 
These results mirror my personal experience. So I do try to find more dynamic masterings, unfortunately, sometimes that leads to things like Adele 25. I have the most dynamic version (the vinyl) but my digitized copy of that vinyl still has audible surface noise when I use very sensitive in-ears, despite my spending an inordinate amount of time trying to clean up the recording. 
 
So, at the end of the day, I look to the measurement because it's important to me, I like more dynamic recordings. I think the measurements are important because there is evidence that I am not alone in preferring more dynamic recordings. 

 

[castleofargh]

 the loudness war is IMO the worst thing to hit the industry after guys making artistic decisions based on money prospect only. but there is more to a good sound than the DR value. even when focusing only on dynamic. a voice is by nature very dynamic, but if you leave it untouched to get "the real sound like the artist blablablah", it often sucks big time. to me that's not relaxing or very enjoyable for long. if it's fun as a once in while thing, I wouldn't want highly dynamic music all day long at home, or even worst, in noisy environments with my IEMs.
so while dynamic brickwalling should be punishable by law, I believe that a lot of materials are much better with some degree of dynamic compression. I guess our preference for masters with higher DR values comes from how we're often comparing compressed stuff with brickwalled stuff. making us hate compression more than we should.

 

[reginalb]

 the loudness war is IMO the worst thing to hit the industry after guys making artistic decisions based on money prospect only. but there is more to a good sound than the DR value. even when focusing only on dynamic. a voice is by nature very dynamic, but if you leave it untouched to get "the real sound like the artist blablablah", it often sucks big time. to me that's not relaxing or very enjoyable for long. if it's fun as a once in while thing, I wouldn't want highly dynamic music all day long at home, or even worst, in noisy environments with my IEMs.
so while dynamic brickwalling should be punishable by law, I believe that a lot of materials are much better with some degree of dynamic compression. I guess our preference for masters with higher DR values comes from how we're often comparing compressed stuff with brickwalled stuff. making us hate compression more than we should.

I would liken it to HDR in photography. Done well, high dynamic range photos are awesome, and can create a stunning image. But photographers hate it. Not because it's a bad tool, but because so many people, when going through toning of the image, end up applying so much saturation and contrast that the resulting photo makes your eyes bleed. So photographers lament HDR when it's really bad toning that they don't like (though you usually see that toning in HDR photos).

Compression exists as a tool because good things can be done with it, it's just been abused.

 

[gregorio]

Quote:

  [1] The albums that have, for me, the most lasting appeal, those that I can really unwind to, are albums that again, score better. [2] Now, this is of course subjective and could be influenced by all sorts of biases, I get that. [3] But there is also at least some research on the topic suggesting that people prefer less compressed music... [4] I have the most dynamic version (the vinyl) ...
[5] I think the measurements are important because there is evidence that I am not alone in preferring more dynamic recordings. 

 
1. In most cases and for most people the opposite is true.
2. There is a high possibility of this being a factor in your case (see #4).
3. The research does not suggest that! It suggests that under a specific set of circumstances people prefer less compressed music.
4. The algorithm used by the DR database has been proven to give false readings for vinyl rips, typically registering several points higher DR than is actually the case.
5. This is the area where things get tricky. Although you may not be alone, you are in general, in a tiny minority.
 

  [1] the loudness war is IMO the worst thing to hit the industry after guys making artistic decisions based on money prospect only. ... [2] I wouldn't want highly dynamic music all day long at home, or even worst, in noisy environments with my IEMs. ... [3] so while dynamic brickwalling should be punishable by law, [4] I believe that a lot of materials are much better with some degree of dynamic compression.

 
1. To be honest, I personally think there are far worse. For example; incredibly cheap/accessible recording technology and the incredibly low average prices consumers now pay for music have significantly driven down the amount of time, money and expertise invested in creating music products.
2. This goes to the heart of the issue. A high dynamic range is preferable only in relatively rare circumstances; when critically listening with a good quality sound reproduction system in a relatively quiet listening environment. While these circumstances are common amongst head-fi members, they are far rarer amongst the general public.
3. In actual fact the exact opposite of your statement is already true! In TV broadcast in some countries (the USA for example), not brickwall limiting the signal is punishable by law (CALM Act, 2010). I suspect there maybe a misuse/misunderstanding of the term "brickwall" here?
4. Again, this is another "heart of the issue" statement. In practise, some degree of compression is virtually always desirable and indeed, there are virtually no commercial audio products released which do not have compression applied. The issue is not as simple as many view it to be! Given that compression is effectively a requirement, the question is, how much compression is the right amount? The reason the issue is not simple is because the correct answer is, "it depends". It depends on the genre of the music, where, how, when and with what the consumer will be listening and how the music was performed and how it was recorded/sourced. Most commonly, the answers to these questions are only a basic starting point for the amount of compression applied and there are further levels of subtly within these answers. For example, orchestral symphonic music is generally the most dynamic genre but exactly how much dynamic range depends on the composer and the symphony (in addition to other factors). Mozart symphonies for example would generally have less dynamic range than say a symphony by Mahler or Stravinsky. Some genres, electronica in general and EDM in particular for example, depend on a large amount of applied compression. Indeed, an amount of compression which would be considered well into "loudness war" territory if applied to many other genres but if not applied to EDM would result in a loss of some of the defining characteristics of what makes EDM, EDM!
 

Compression exists as a tool because good things can be done with it, it's just been abused.

 
True. The difficulty is in deciding when it's been abused, what constitutes "abuse". Listening to orchestral music while driving, one would need a great deal of compression to actually render it listenable, an amount of compression which would certainly constitute "abuse" when listening to that same recording on a decent hifi system at home. The DR database is a useful tool (although there are better IMHO) for giving some indication of dynamic range, what I strongly disagree with is it's concept of massively over-simplifying the issue by using the blanket statement that higher DR is better. Is a Mahler symphony automatically way better than the average EDM track? Is a Mahler symphony automatically better than a Mozart symphony? Is a recording of a Mahler symphony automatically better than a different recording of exactly the same Mahler symphony, just because it has a higher DR score? The answer to these and a number of similar questions is, "not necessarily, it depends, the exact opposite could be true!" and in this respect the DR score is of relatively little importance and potentially, extremely misleading!
 
G

 

[reginalb]

  Quote:
 
1. In most cases and for most people the opposite is true.
2. There is a high possibility of this being a factor in your case (see #4).
3. The research does not suggest that! It suggests that under a specific set of circumstances people prefer less compressed music.
4. The algorithm used by the DR database has been proven to give false readings for vinyl rips, typically registering several points higher DR than is actually the case.
5. This is the area where things get tricky. Although you may not be alone, you are in general, in a tiny minority.
 
 
1. To be honest, I personally think there are far worse. For example; incredibly cheap/accessible recording technology and the incredibly low average prices consumers now pay for music have significantly driven down the amount of time, money and expertise invested in creating music products.
2. This goes to the heart of the issue. A high dynamic range is preferable only in relatively rare circumstances; when critically listening with a good quality sound reproduction system in a relatively quiet listening environment. While these circumstances are common amongst head-fi members, they are far rarer amongst the general public.
3. In actual fact the exact opposite of your statement is already true! In TV broadcast in some countries (the USA for example), not brickwall limiting the signal is punishable by law (CALM Act, 2010). I suspect there maybe a misuse/misunderstanding of the term "brickwall" here?
4. Again, this is another "heart of the issue" statement. In practise, some degree of compression is virtually always desirable and indeed, there are virtually no commercial audio products released which do not have compression applied. The issue is not as simple as many view it to be! Given that compression is effectively a requirement, the question is, how much compression is the right amount? The reason the issue is not simple is because the correct answer is, "it depends". It depends on the genre of the music, where, how, when and with what the consumer will be listening and how the music was performed and how it was recorded/sourced. Most commonly, the answers to these questions are only a basic starting point for the amount of compression applied and there are further levels of subtly within these answers. For example, orchestral symphonic music is generally the most dynamic genre but exactly how much dynamic range depends on the composer and the symphony (in addition to other factors). Mozart symphonies for example would generally have less dynamic range than say a symphony by Mahler or Stravinsky. Some genres, electronica in general and EDM in particular for example, depend on a large amount of applied compression. Indeed, an amount of compression which would be considered well into "loudness war" territory if applied to many other genres but if not applied to EDM would result in a loss of some of the defining characteristics of what makes EDM, EDM!
 
 
True. The difficulty is in deciding when it's been abused, what constitutes "abuse". Listening to orchestral music while driving, one would need a great deal of compression to actually render it listenable, an amount of compression which would certainly constitute "abuse" when listening to that same recording on a decent hifi system at home. The DR database is a useful tool (although there are better IMHO) for giving some indication of dynamic range, what I strongly disagree with is it's concept of massively over-simplifying the issue by using the blanket statement that higher DR is better. Is a Mahler symphony automatically way better than the average EDM track? Is a Mahler symphony automatically better than a Mozart symphony? Is a recording of a Mahler symphony automatically better than a different recording of exactly the same Mahler symphony, just because it has a higher DR score? The answer to these and a number of similar questions is, "not necessarily, it depends, the exact opposite could be true!" and in this respect the DR score is of relatively little importance and potentially, extremely misleading!
 
G

 
Firstly, I think that you have missed that we have already discussed in the thread that the type of music surely counts. You're right that a lot of it has been for things like the car, but I want my music to sound good at home, not in the car. Sure, there are some people that don't, but it would be trivially easy at this point to have the limiting done in the playback system. 
 
I would like to see some studies that show what you're saying, because I've gone through several, and I've never seen a well designed study that suggests that people prefer less dynamic music.
 
Also, the law that you cite does not make it punishable by law to not brickwall, certainly a brickwall limiter could help with compliance, but there is nothing in the law that inherently requires a brickwall limit, nor would a brickwall be necessary to achieve volumes that meet the requirements set by the law. There is also a similar federal law, and all they say is that commercials can't be louder than the programming. You don't need (though, again, it could be helpful) a brickwall for this. And it certainly isn't aimed at fidelity of music, it's aimed at advertisers that had blaring volumes on their commercials. 

 

[gregorio]

  [1] You're right that a lot of it has been for things like the car, but I want my music to sound good at home, not in the car.
[2] Sure, there are some people that don't, but it would be trivially easy at this point to have the limiting done in the playback system. 
[3] I would like to see some studies that show what you're saying, because I've gone through several, and I've never seen a well designed study that suggests that people prefer less dynamic music.
[4] Also, the law that you cite does not make it punishable by law to not brickwall, certainly a brickwall limiter could help with compliance, but there is nothing in the law that inherently requires a brickwall limit, nor would a brickwall be necessary to achieve volumes that meet the requirements set by the law. There is also a similar federal law, and all they say is that commercials can't be louder than the programming. You don't need (though, again, it could be helpful) a brickwall for this. And it certainly isn't aimed at fidelity of music, it's aimed at advertisers that had blaring volumes on their commercials. 

 
1. Good at home for whom? Many people listen to music at home as background music, while they perform other tasks or entertain. In which case, a more compressed, a low dynamic range is preferable. You seem to be missing my point, music isn't mixed/mastered according to what you personally want, it's mixed/mastered for a market and the market for those who are listening critically on good hifi systems in quiet environments is extremely small. A recording with a higher DR is not necessarily better even for this small minority, there are other far more important factors, even more so given that the DR score is spurious from a vinyl rip.
2. Really? How much limiting should the playback system apply? Would it be able to auto detect the genre and apply more or less? Would it be able to know my artistic intentions as a music producer of how much limiting to apply?
3. I don't know of any, I can't think of a good reason to run such a study. If music has too great a dynamic range for the audiences' playback system/environment they are either going to be listening to inaudible music half the time or be constantly/frequently adjusting the volume control, in effect applying their own, manual compression. Either scenario would obviously not be preferable to a master with less dynamic range. I've seen studies before which set out to provide evidence of something which everyone already considers "obvious" so I don't know why they haven't done that with DR.
4. I'm not sure what law you're referring to, I'm talking about the federal law (CALM Act). It's not aimed specifically at music fidelity, it is aimed at broadcast TV audio. It is as you say, intended to stop commercials being louder than the programming and by necessity that of course means that the loudness of the programmes has to be known/limited. The peak limit requirement is -2dBTP in the USA, this could in theory be achieved by manually "riding the gain" to stay within the legal limit but is impractical, in practise it's done automatically, by a brickwall limiter. Failure to adhere to the -2dBTP brickwall and integrated loudness limits is punishable under the CALM Act.
 
G

 

[reginalb]

   
1. Good at home for whom? Many people listen to music at home as background music, while they perform other tasks or entertain. In which case, a more compressed, a low dynamic range is preferable. You seem to be missing my point, music isn't mixed/mastered according to what you personally want, it's mixed/mastered for a market and the market for those who are listening critically on good hifi systems in quiet environments is extremely small. A recording with a higher DR is not necessarily better even for this small minority, there are other far more important factors, even more so given that the DR score is spurious from a vinyl rip.
2. Really? How much limiting should the playback system apply? Would it be able to auto detect the genre and apply more or less? Would it be able to know my artistic intentions as a music producer of how much limiting to apply?
3. I don't know of any, I can't think of a good reason to run such a study. If music has too great a dynamic range for the audiences' playback system/environment they are either going to be listening to inaudible music half the time or be constantly/frequently adjusting the volume control, in effect applying their own, manual compression. Either scenario would obviously not be preferable to a master with less dynamic range. I've seen studies before which set out to provide evidence of something which everyone already considers "obvious" so I don't know why they haven't done that with DR.
4. I'm not sure what law you're referring to, I'm talking about the federal law (CALM Act). It's not aimed specifically at music fidelity, it is aimed at broadcast TV audio. It is as you say, intended to stop commercials being louder than the programming and by necessity that of course means that the loudness of the programmes has to be known/limited. The peak limit requirement is -2dBTP in the USA, this could in theory be achieved by manually "riding the gain" to stay within the legal limit but is impractical, in practise it's done automatically, by a brickwall limiter. Failure to adhere to the -2dBTP brickwall and integrated loudness limits is punishable under the CALM Act.
 
G

 
Regarding 3 and 1, there have been studies, and they have determined the exact opposite of what you're asserting. There aren't tons of studies, though, so I thought maybe you knew of some others. I linked to a couple earlier. I think you're just wrong on this one, most people, once exposed to different options, prefer more dynamic recordings. Again, there haven't been tons of studies on this, but that's what they've found. Though you're certainly right that it's going to be genre specific. 
 
Regarding 4 we're largely in agreement, I would just re-assert that it's just not applicable to music. 
 
Regarding 2, lots of playback systems actually already do it, just like in regards to 4, where it's actually usually done with software on the way out the door, because it's the broadcasters that are required to ensure compliance, not the people recording the ads. And there was already software in a lot of TVs to do it before it became a mandate for broadcasters.
 
Regarding what you mentioned about vinyl and the dynamic range database, thank you, I didn't realize that. I did Google around a bit to see what I could see on the points you're making, and watched Ian Sheppard's video about vinyl and the dynamic range meter used for the DRDB, pretty interesting stuff.

 

[VNandor]

 Regarding 3 and 1, there have been studies, and they have determined the exact opposite of what you're asserting. There aren't tons of studies, though, so I thought maybe you knew of some others. I linked to a couple earlier. I think you're just wrong on this one, most people, once exposed to different options, prefer more dynamic recordings. Again, there haven't been tons of studies on this, but that's what they've found. Though you're certainly right that it's going to be genre specific. 

That kind of contradicts the whole idea of the loudness war. I bet recordings get compressed because that's what most people like (demand) and not because the sound engineers are like "why not compress it". Don't get me wrong, I agree that people could prefer more dynamic recordings compared to badly compressed ones however people also tend to prefer a louder sound. That could cause a demand for compression. Now we might have ridiculously strong amplifiers that could blow up our ears and our headphones in seconds but most people probably don't have it. They either turn the volume up so much the sound starts to distort or they keep the volume down and it just won't be loud enough. (Still could be loud enough to damage the hearing in the long term but that's an other topic.) So what's the "solution"? Louder signals because most people don't go out to buy some amplifier.

 

[reginalb]

  That kind of contradicts the whole idea of the loudness war. I bet recordings get compressed because that's what most people like (demand) and not because the sound engineers are like "why not compress it". Don't get me wrong, I agree that people could prefer more dynamic recordings compared to badly compressed ones however people also tend to prefer a louder sound. That could cause a demand for compression. Now we might have ridiculously strong amplifiers that could blow up our ears and our headphones in seconds but most people probably don't have it. They either turn the volume up so much the sound starts to distort or they keep the volume down and it just won't be loud enough. (Still could be loud enough to damage the hearing in the long term but that's an other topic.) So what's the "solution"? Louder signals because most people don't go out to buy some amplifier.

 
Eh, I don't know. The average user with an iPhone and buds have a pairing that can get plenty loud. That average use also doesn't typically have some obscure, hard to drive headphone or speakers.