[Strangelove424]

  It may look insane from hi-end audio perspective, but we are on minority side and not most economically attractive. DR 6 db song will sound better than DR 15 db on average boombox or car because there is less chance that it will clip cheap amp.
 
Adding on-fly compression is probably to so straightforward, will cost something anyway and most people won't even notice the difference.

 
If a sound system has limitations, it's not the responsibility of an entire field of art to compensate for that. That's for the audio system designer to account for, not the guy in a studio making music. And I don't think it will necessarily sound better with low DR, even in a car. I drive an average Honda with stock stereo, and even listening to music there (usually streaming though my phone) I stick to well mastered music, and still think it sounds better with more dynamics. Loud pop music and radio, on the other hand, sounds like pure garbage, even on a lowly system. I can't imagine how tons of dynamic range compression and digital signal clipping are healthy for cheap car stereos either. I wouldn't believe for a second they compress the dynamic range for the health of our audio systems. They do it to make their songs stick out more, in hopes they'll sell more copies. It's about money. 

 

[tarnumf]

Well, current state of audio industry suggests that it's more profitable to manipulate audio material, rather than improve audio systems. Plus for general public "re-mastered" is equal to "new and improved".
 
In their defense, DR levels are not always making music sound bad. It's rather new generation of mastering engineers who have little idea what they are doing and applying it over the top. Same goes in photography too - raise saturation and contrast in picture and more people will like it, raise it even more, to unnatural levels and even more people will like it.
 
And as real-life illustration: I have couple tracks from STS Digital Dynamic Experience Volume 1 and while their DR ranges are pretty average, 8-10 db, they sound pretty nice. For example "Stevie Ray Vaughan - Roughest place" has DR 11 on that disc and same song on MFSL version of original album has DR 16. Isn't that the irony?

 

[Strangelove424]

I’d like to see a shift from the high-resolution or “remastered” (as in cranked and squashed) re-release marketing gimmicks to something with real meat. Like an assortment of classic rock’n’roll mastered by classical music specialists, for example. I’ve heard some amazing releases (especially in multi-channel) that convinced me a brand new master from a really good engineer can be a huge selling point, and actually worth the often steep asking price. 

There’s lots of reasons a master might sound bad, dynamic range compression being one of many. It certainly doesn’t help though, and can often ruin a perfectly good track. As everything in life, balance and taste are key. I’ll admit that 10db of range can still sound okay, depending on the type of material. But at a DR of 3 or 4 db, some of these albums are getting dangerously close to the threshold of audibility. That essentially means zero perceptible change in level.
 
For the most part, dynamic range is still a major indicator of enjoyment for me. I want to hear instruments (voices included) do their thing. I’d like the natural dynamic range they produce to be captured as best as possibly given within the capabilities of their medium, which even in CD format is incredibly high. 

 

[linglingjr]

Are any of you guys audio engineers?  What do they even teach you when you drop the cash to do that in university?  
 
I agree the whole "re-mastered" hoopla works to bring in more cash just like re releasing on new technology (vinyl, tapes, CDs) any idea why 0 DR is so popular?  I remember reading something about making it all as loud as possible in clubs/shows where they blast it at 2000dB but I have no idea.  

 

[Dark_wizzie]

I think the average person doesn't even pay attention to their music. I've went on Youtube to check out a few tracks and some of them have horrible clipping (dunno how the uploader managed to screw it up), but nobody complained. I'm not talking audiophile golden-ear level stuff, I'm talking very obvious clipping here. Makes me wonder what the hell is wrong with people. If those people can't detect clipping when it's crammed into their skull, they won't give a crap about DR. And so... almost nobody cares about DR in the world. Nobody is pressured to make any changes.

 

[esteboune]

Strangelove,
 
thanks a lot for this great ressources.
 
i'm slowly educating myself on the music enjoyment path (can't say audiophile yet!!!)
 
As many of you here, i invest quite a lot of money in decent gear to enjoy my favorite music...
 
And i also invest in music, and i definitely made mistakes....HDtracks, 24/192, re-mastered version (biggest mistake!)
 
I wrongly associated bit rate and bit depth to music quality.
 
Quite recently, I discovered that the way that the album was mastered will impact much more that the compression...Way more...
 
And man, I did learn my lesson.
NO more remastered version!
 
quick exemple: Dire Straits, Money for nothing. Original release (top) VS. Remastered release 2005 (bottom)
 

 
Even though the remastered version is a 24/192 vs 16/44, the orignal is much more alive, the DR is bigger, as the enjoyment...

 

[RRod]

It's always buyer beware out there. It would of course be trivial for these HD companies to put out information and samples that were actually useful (waveforms, spectrograms, dynamism/loudness measurements, un-adultered sound samples, comparison samples), but this is of course not in their best interest. Instead the buyer has to do his own digging: asking on forums like these how specific masters sound, getting general feelings for which artists and companies release legit material, etc.

 

[Husky]

I've just come across this old thread and would like to recommend a tool to analyze compression and clipping called ClippingAnalyzer (in MS Excel): http://www.ber-sd.com/dl_clipping_eng.html

Here is the result for "Money for Nothing" from the original CD. There is no clipping and almost no compression (1%).

The first remaster from 1996 has clipping and compression (32%):

 

[gregorio]

Hmm, this is a bit of an old thread so hopefully the OP (and other contributors) has a bit better understanding of compression, limiting, clipping, the loudness war and mastering in general. Just in case though:
I understand the desire of audiophiles to find problems/weaknesses in their systems and thereby try to improve their listening experience. I applaud the fact that what the system is actually playing is the focus of this thread, which far too often is simply ignored and leads to some of the most ridiculous and bizarre of audiophile claims. If more consumers considered (and valued!) production and mastering then it would literally revolutionise the music industry, both the creation of music and all the equipment used to reproduce it!! On the other side of the coin, the danger (as clearly demonstrated by this and other threads) is in looking for simple answers to apparently simple questions and in the process to simplify the issue to such a point that it becomes virtually as ridiculous as the audiophile claims which are ridiculous because they ignore the issue! There is clearly some fundamental misunderstandings in this thread about compression and mastering in general and therefore some truly bizarre statements. For example:

[1] I some how managed to try out an Oasis album without realizing they're credited for some of the worst brick wall (I like sausage waveform more) mastering of ALL TIME.
Picked some random song from their discography:
[2] It really isn't listenable ...
[3] ... but why does anyone master their stuff like this?

1. Yes, arguably "Morning Glory" is this single most influential album in the history of the loudness war.
2. Huh, you know "Morning Glory" is one of the best selling albums of all time? Are you saying that 22 million people bought an album which they couldn't listen to?
3. In the case of "Morning Glory" I can think of about 350 million reasons! Don't you think that's a decent incentive for other artists and labels?

[1] They are absolutely breaking the rules of their own craft, violating the most basic standards of audio broadcast and distribution.
[2] I don't think most sound engineers think it sounds good, but there might be a producer sitting behind him who only has the sales numbers and his own promotion in mind, and unfortunately that commercial marketing entity sitting there has a stronger voice than the artist, the engineer, or good taste itself.
[3] Who rapes Beethoven like this? That's just sick.

1. And what punishment did they receive for such violation of the most basic standards of audio broadcast and distribution? Somewhere in the region of $1.2 billion in record sales alone!
2. In this case, the guy sitting behind the engineer was Noel Gallagher!
3. I've seen Beethoven "raped" much more than that! Who would do such a thing? Someone who wants to be able to hear all the notes in Fur Elise rather than just the odd bit of melody.

Here is the result for "Money for Nothing" from the original CD. There is no clipping and almost no compression (1%).

It's not clear what exactly that program it trying to measure or how it's trying to measure it. There is no way of measuring compression and "Brothers in Arms" was actually quite heavily compressed for it's day, certainly way more than 1% (although it's meaningless percentage anyway). The clipping doesn't really make any sense on the remastered version either and neither (again) does the compression.
----------------------------------------------------------
I don't know if anyone really wants to resurrect this thread, if so I'm happy to explain my points above but be warned, there are no simple black and white answers here because the question is in reality a rabbit hole and nowhere near the simple question it might appear to be.

G

 

[bigshot]

You're absolutely right. Form follows function.

I recently got the "super deluxe" version of the eponymous Fleetwood Mac album. It comes with an LP (ballast I'll never play), several CDs of demos (maybe I'll play once), and a DVD with a 5.1 mix (the main reason I bought it). It was quite expensive so I looked for reviews to see if the 5.1 mix merited the price. People online were saying that it was mastered too hot and had clipping. They showed charts that looked compressed, but they didn't show actual clipped sections. Other people said that it sounded great- the best the album had ever sounded. There was no consensus, so I took a chance and ordered it from Amazon, knowing I could return it if I didn't like it. When I put it on my system, it was a notch or two louder than most of my other albums, but it sounded clear and I couldn't hear any clipping at all. What I heard were songs that had been recorded and mastered to serve the dynamic requirements of AM radio airplay. The fade outs were perfectly clean and even all the way out (some of them on this album are quite long- as long as a full minute.) And the peaks were punchy and in your face as they should be, especially the snare drum hits. No problem at all.

I went back to one of the people who claimed to have heard clipping and asked him what part of the album he heard it in and what the clipping sounded like. He said it was throughout the whole album and it sounded like a buzziness in the bass and a limited frequency response in the high end. Uh huh... I pointed out that clipping doesn't sound anything like that and he immediately shifted to saying that it was probably because it was encoded in a high bitrate Dolby Digital DVD format rather than one of those overkill 24/192 blu-ray formats. Uh huh again... He clearly heard a slightly louder volume level and immediately leapt to the clipping assumption, and saw that it was on a DVD instead of a blu-ray and let expectation bias convince him that it didn't sound as good because it wasn't a blu-ray. He never used his ears.

People misuse the "loudness wars" arguments all the time. They compare dynamics of a Mahler symphony to a Rolling Stones single as if that is a fair comparison. More dynamics isn't necessarily better- the *proper* dynamics are. To know what the proper dynamics are, you have to use your ears. A lot of people who post reviews on the internet are afraid to connect their ears to their brain and go out on a limb and actually make a value judgement for themselves. They feel safer pointing to charts and diagrams produced by apps that don't exactly define what it is they're measuring. If you ask them questions, they don't even know what clipping or compression sounds like.

It's a lot easier for me to just listen carefully, analyze what I hear, and come up with an opinion on it.

 

[Strangelove424]

Hmm, this is a bit of an old thread so hopefully the OP (and other contributors) has a bit better understanding of compression, limiting, clipping, the loudness war and mastering in general. Just in case though:
I understand the desire of audiophiles to find problems/weaknesses in their systems and thereby try to improve their listening experience. I applaud the fact that what the system is actually playing is the focus of this thread, which far too often is simply ignored and leads to some of the most ridiculous and bizarre of audiophile claims. If more consumers considered (and valued!) production and mastering then it would literally revolutionise the music industry, both the creation of music and all the equipment used to reproduce it!! On the other side of the coin, the danger (as clearly demonstrated by this and other threads) is in looking for simple answers to apparently simple questions and in the process to simplify the issue to such a point that it becomes virtually as ridiculous as the audiophile claims which are ridiculous because they ignore the issue! There is clearly some fundamental misunderstandings in this thread about compression and mastering in general and therefore some truly bizarre statements.

I don't know if anyone really wants to resurrect this thread, if so I'm happy to explain my points above but be warned, there are no simple black and white answers here because the question is in reality a rabbit hole and nowhere near the simple question it might appear to be.

I had forgotten about this thread, and was glad today to see somebody take an interest in it again! The goal of it initially was to take the mystery and obfuscation out of mastering arguments or purchasing decisions, and give people the tools to pinpoint differences in masters they heard. (It was this hearing part I realized could be problematic, but I will address this soon.) The goal here was also to take undue emphasis away from HD music, where people I believe were noticing differences in mastering and trying to put the emphasis back where it belonged, on the human being making artistic/commercial decisions. And as you mention, it is nice to see a thread focused on music, that spring well of the humanities that seems to bring us all together, rather than the gear which seems often to drive forums apart with engineering supremacy arguments. So I think we all agree about the benefits the thread could bring, and also the problems. It occurred to me that the numbers don't tell everything. It's easy to construct a song from the mastering engineer's point of view using measurement and math, and to understand the song (and things like DR) in their entirety. It is much more difficult to deconstruct a song from the other side, and reverse engineer it with measurements to understand how the song works. For instance, the DR measurement tool has some glaring problems. A Google search will show many controversial arguments against its measuring algorithm, and the way that it averages level across the length of the song. This DR number can be very helpful, but it also became too all-important of a number, and has questionable correlations to a song's actual DR (whatever actual DR really means, because when you get into averaging, it's a Pandora's box of what to average). This is a situation where the numbers can't say everything, and by relying too much upon them, they become a crutch for understanding. Instead of listening to the song, people start to rely just on numbers. And that is a very dangerous phenomenon. I'm down to open this thread back up, but encourage people to always reference their measurements in context of how it sounds.

 

[Strangelove424]

You're absolutely right. Form follows function.

I recently got the "super deluxe" version of the eponymous Fleetwood Mac album. It comes with an LP (ballast I'll never play), several CDs of demos (maybe I'll play once), and a DVD with a 5.1 mix (the main reason I bought it). It was quite expensive so I looked for reviews to see if the 5.1 mix merited the price. People online were saying that it was mastered too hot and had clipping. They showed charts that looked compressed, but they didn't show actual clipped sections. Other people said that it sounded great- the best the album had ever sounded. There was no consensus, so I took a chance and ordered it from Amazon, knowing I could return it if I didn't like it. When I put it on my system, it was a notch or two louder than most of my other albums, but it sounded clear and I couldn't hear any clipping at all. What I heard were songs that had been recorded and mastered to serve the dynamic requirements of AM radio airplay. The fade outs were perfectly clean and even all the way out (some of them on this album are quite long- as long as a full minute.) And the peaks were punchy and in your face as they should be, especially the snare drum hits. No problem at all.

I went back to one of the people who claimed to have heard clipping and asked him what part of the album he heard it in and what the clipping sounded like. He said it was throughout the whole album and it sounded like a buzziness in the bass and a limited frequency response in the high end. Uh huh... I pointed out that clipping doesn't sound anything like that and he immediately shifted to saying that it was probably because it was encoded in a high bitrate Dolby Digital DVD format rather than one of those overkill 24/192 blu-ray formats. Uh huh again... He clearly heard a slightly louder volume level and immediately leapt to the clipping assumption, and saw that it was on a DVD instead of a blu-ray and let expectation bias convince him that it didn't sound as good because it wasn't a blu-ray. He never used his ears.

People misuse the "loudness wars" arguments all the time. They compare dynamics of a Mahler symphony to a Rolling Stones single as if that is a fair comparison. More dynamics isn't necessarily better- the *proper* dynamics are. To know what the proper dynamics are, you have to use your ears. A lot of people who post reviews on the internet are afraid to connect their ears to their brain and go out on a limb and actually make a value judgement for themselves. They feel safer pointing to charts and diagrams produced by apps that don't exactly define what it is they're measuring. If you ask them questions, they don't even know what clipping or compression sounds like.

It's a lot easier for me to just listen carefully, analyze what I hear, and come up with an opinion on it.

I can't speak directly to your 5.1 DVD, but I was extremely disappointed with a modern remaster of Fleetwood Mac's Rumors on Spotify. Not sure I heard clipping, but it was definitely harsh to my ears and brought out the worse in the micing of the singers. After finding only modern remasters on streaming, I went back to all my old Fleetwood Mac CDs. Never bothered capturing the Spotify version to DAW and analyzing dynamic range because it sounded like junk to me, and even if the numbers said otherwise I still wouldn't listen to it. Gotta listen to your ears first. But in some situation I think analysis could be helpful. I also think looking at the waveform might be a ton more helpful than looking at a DR#. A conductor doesn't hold up a single number for each piece, I don't think DR can be analyzed that way either tbh.

 

[bigshot]

I don't have that remaster because they didn't go to the trouble of doing a 5.1 mix on that. I have the old CD and it sounds fine. I had Fleetwood Mac on the old half speed mastered MFSL LP and that sounded great. The CD wasn't quite as good, but this new remaster tops the LP. Tusk is in a whole different league. That album always sounded terrible from the very beginning. The LP was pretty much unlistenable. The CD wasn't much better. I really think the sound quality and the cost of buying a 2 record set killed the band's momentum. But the remaster is a revelation. It's like a totally different album. The 5.1 mix sounds amazing. Tons of arrangement details that were just mush on previous releases. I actually like that album the best now.

With remasters it all depends. Some are massive improvements and some are duds. That's why I try to always look for reviews about the quality of the mastering before buying. But the quality of the reviews could stand improvement too! For the most part, I just stick with my old CDs. I only upgrade if there is a new 5.1 mix.

By the way, if you look at that Dire Straits waveform, it looks like there is just one really loud part that is pushing the overall level of the whole album down. I'm not familiar enough with the album to know what part that is, but it could be that it was mixed too hot originally, or maybe they were burning it in for LPs that could take it at the beginnings of sides. I would have to hear it to tell whether that particular part suffers for being compressed a bit. But looking at the chart, it appears that aside from sporadic narrow spikes, the rest of the album has plenty of room to normalize up without affecting the dynamics if you leave that one peak out of the equation. It may be that only that one brief loud part and a few tiny spikes are compressed and the majority of the album is exactly the same on both versions.

 

[Strangelove424]

5.1 mixes tend to be better. I think they assume the equipment will be decent and they usually let the mastering artist have a little more room to play. The problem with identifying remaster versions is that it is often unclear which remaster one has heard. I am particularly not fond of the Super Deluxe remasters, but as far as I know they date back to 2004! I know my 90s Fleetwood CDs sound very good. I have argued before that the music industry should be forced to label albums with the mastering engineer's name and each version of a song should have its own serial number.

I have not heard both versions of Dire Straits, but I probably am used to listening to something close to the first version from my old CDs. The second waveform appears to spend a whole lot more time near peak. Whereas the first version peaks briefly near peak at the intro (and only momentarily) the second version stays near that peak longer, and then revisits it many many times throughout the song. The second version also has far louder quiet parts, which must make the peaks less impactful.

 

[RRod]

By the way, if you look at that Dire Straits waveform, it looks like there is just one really loud part that is pushing the overall level of the whole album down. I'm not familiar enough with the album to know what part that is, but it could be that it was mixed too hot originally, or maybe they were burning it in for LPs that could take it at the beginnings of sides. I would have to hear it to tell whether that particular part suffers for being compressed a bit. But looking at the chart, it appears that aside from sporadic narrow spikes, the rest of the album has plenty of room to normalize up without affecting the dynamics if you leave that one peak out of the equation. It may be that only that one brief loud part and a few tiny spikes are compressed and the majority of the album is exactly the same on both versions.

It's a good point. People try to pretend that 'all peaks are sacred', but there are realities of hearing such as integration time that allow for wiggle room. The worst offender I've found is the somewhat infamous Hartke disc used in the BAS test. The SACD is mastered at about -35dB RMS, and I swear 20dB of that is just for this one 'ping' on a glockenspiel. Been wanting to set up a blind test on that one for some time to see how much can be taken off the top.