"Loudness War" and the Dynamic Range (DR) Database - some observations

[The Dude]

I must say though I don't quit understand all of this, that most of it is good and interesting to read.

 

I listen to quite a bit of metal, but I don't think any of it as High Fidelity. Some metal sounds really good recording-wise, quite a bit doesn't. Metal isn't supposed to sound nice, but that also depends on the genre.

 

I have found some to be good some to be great and some to be horrible.  The writer of the article brought up Death Magnetic, I though on CD it was a horrible recording, but when listened to on vinyl it sounded pretty good.  Type O Negative produced some pretty good recordings, i don't know what it is about them, the dynamic range doesn't seem to be there but the mixing appears to be really good.   Pantera one of my favorite bands sounds really good, though there are things about the productions I don't care for, but again the music is so good i don't pay much attention to them.  Trent Reznor of NIN is one that refuses to compress music, or so he said a couple years ago, those are some great recordings by him.

 

Understanding what Pete said, last year I received Rush's first album remastered on 200 gram vinyl and was not impressed,

[Chris A]

I think most hard rock and metal has the guitars emphasized somewhere in that HF range you are referring to, to give the impression the guitars are ripping your face off. Decreasing the bass would seem to accentuate the HF of already hot top end?

 

Yes.

 

I've found that the overall tonal balance of rock recordings (and all its derivatives, including metal, etc.) is probably the most sensitive genre for low bass, midrange and high frequency balance than, for instance orchestral or bebop (classic) jazz.  I'm sometimes surprised by the level of EQ creativity with the latter two genres, and the recordings still sound very similar to their live counterparts.  But if you mess with the bass on a rock recording by as little as 2-3 dB, it doesn't sound anywhere like it did.

 

I've often thought that the stories that I read on this forum and others about how people found "magical combinations of gear" (including Khorns, tubes, etc, that really sounded spectacular.  I'd bet that if you ran a REW sweep through those setups, that the frequency response would look like many of my unmastering EQ curves for rock recordings.  This may also explain why some folks never listen to anything else: they've inadvertently found a somewhat general reverse EQ curve with their setups for most reasonable-sounding rock, but other genres simply sound awful because their characteristic unmastering EQ curves don't look like the rock ones.

 

I've found that once you EQ everything back to flat (including restoring the bass octaves, and rebalancing the exaggerated highs, etc) for all your music, it all sounds good, regardless of genre.  There are no "house curves" used on my setup: it's as flat as I can make it.

 

A thought that's certainly something to ponder: standardized music mastering practices and it's potential effect on the music industry.

 

Chris

Edited December 19, 2015 by Chris A

[mike stehr]

Type O...Last time I seen Peter was with Hatebreed. The guy was a Monster...Bass in one hand, quart of Jägermeister in the other. Great band to hear live...

[mike stehr]

A thought that's certainly something to ponder: standardized music mastering practices and it's potential effect on the music industry.

 

It makes sense to me.

[The Dude]

Peter was with Hatebreed

 

More like deceased, sorry for the bad news

[mike stehr]

Yup. Died too young.

[Chris A]

Just an example of the current mass market culture of "Loudness" in this aspiring mastering novice's blog entry, written a mere 5 months ago.  Even the comments are interesting on this one:

 

https://www.reddit.com/r/edmproduction/comments/3cmj03/im_giving_up_on_the_loudness_war/

 

It seems that no one questions the need for loudness and greater loudness, ultimately.  However, the last time I noticed music buyers weren't buying based on how loud the tracks were...especially with auto-gain functionality that's now built-in and is defaulted to "on" in the personal music player marketplace--notably iPods, et al. 

 

This seems to be a holdover cultural imperative from an age that his been prophesied to die a quick death of late...again and again.

 

Chris

[The Dude]

I have dabbled in this thread a bit along with comparing tracks at the DR database.  As I try to understand it in a simple matter, here is what I have.  Recording companies have cranked up the loudness on the mastering, which makes the dynamic range seem better.  But it doesn't as some of us have better sound systems than most.  So like Chris has stated he has often jumped up to turn the music down.  However when doing so, the dynamics of the music are gone, do to the fact that they are compressed so they could be turned up during the mastering.  Is this about right. 

 

From the last article posted, the produce states going back to focusing on the music and not trying to compete with the major producers.  This in a sense would give us better quality which will seem more dynamic at lower volumes.  Being able then to keep the volume where we like it and still having the detail in the music.  Is the quality equal to DR? If so, poor dr would be poorer quality at lower levels.  Is it that big of deal for us with efficient sound systems? Wouldn't we be able to hear the quality at lower levels than what most people would?

[Chris A]

Actually, dynamic range and "loudness" are on either end of a logarithmic scale: to be louder, you MUST have lower dynamics, etc.

 

It's well known that people prefer music that is merely louder in A-B tests, as little as 0.1 dB difference in loudness.  People will statistically pick the louder track as "sounding better" when comparing tracks.  However, when the tracks are leveled in loudness carefully other factors (such as higher fidelity) become apparent.

 

Because there is the "auto-gain" function on your iPod, and most people will leave that feature "on", it will automatically level the loudness of each track to some pre-determined average loudness level.

 

Dynamics sound better, up to a point.  As you go far past 15 dB crest factor (i.e., average-to-peak, a DR rating of "15" or greater) the music will start to become somewhat disruptive to listen to at casual levels.  You'll have to turn it up to near concert volume in order to hear the music clearly, but the loudest sections of the music will prevent people in the room from being able to hold conversations during those loud sections. That's kind of a nutshell view of dynamic range.

 

At low DR ratings (for me, that's anything below "11"), the music starts to become the same loudness all the time - like background noise from a machine.

 

Chris

[mustang guy]

It's been a long time since this thread got activity. I thought it deserved a bump as it is an excellent discussion.

[willland]

 It's been a long time since this thread got activity. I thought it deserved a bump as it is an excellent discussion.

 

IMO, just listen to some Steely Dan from any medium(MP3, CD, LP, Pandora, etc.) if you want dynamics.

 

All you need to do is twist the volume knob to low to medium levels to experience it.

 

Bill

[Chris A]

I still unmaster all CD/stereo music that comes in the door.  I've got about 50% of my stereo music collection unmastered now.  I can tell you that it's pretty obvious when playing a music track that hasn't been unmastered--it sounds thin, strident and uninteresting. After unmastering, it sounds full, rich, balanced, much better to listen to at low SPL and high SPL, and it causes you to stop and listen.

 

I really do understand why a lot of people try to blame the CD format itself for bland, uninteresting music. But it's not the CD format...it's the mastering, which is largely reversible for music tracks that haven't been over-compressed. 

 

It's also a good skill to develop as a music lover and an audio enthusiast--since it is more than capable in saving lots of money spent on frivolous attempts to improve the music through buying new and expensive gear, and multiple versions/copies of music in the quest to find better sounding recordings.  Once you unmaster a recording with good starting dynamic range--I find that the need to replace those music tracks with better ones subsides. 

 

However, for stereo music tracks produced since 1991, the success rate is lower since this is the music that's been significantly compressed.  Using the TT Dynamic Range Meter (or it's equivalent Dynamic Range plugin within foobar2000) will tell you how badly damaged the music files are when you first evaluate your music tracks, and using Audacity to un-clip the tracks and re-balance the equalization back to something more approximating what the musicians themselves would play. 

 

I've also found that for acoustic music (like classical, folk, Celtic, some New Age titles, and classic jazz) removing low frequency line noise--such as 30, 60, and 120 Hz for US-produced music, and 25, 50, and 100 Hz for European-produced music. Once you clear those offending non-well-tempered scale frequencies, the resulting tracks exhibit significantly lower modulation distortion, and sound much more transparent and cleaner.

 

As far as Steely Dan - some of the best unmastered tracks that I own come from this band and related music genre(s): Doobie Brothers, Alan Parsons, Supertramp, Genesis, Norah Jones, Joni Mitchell, etc.  All my classical music strongly responds to unmastering, except those recording from Linn Records, which typically need no unmastering at all. 

 

Chris

 

[NBPK402]

Would using a DBX expander help? I remember back in the 70s they were very popular.

[Chris A]

I've found that upward expanders are very difficult to use well, and the more de-compression you're dealing with, the more difficult to get a good result.

 

I've used Red Phatt Pro, which can do okay for very mild de-compression.  Bigger jobs will require more expensive (i.e., non-freeware) applications.  This includes multi-band compressors/expanders, and the difficulty in finding the right expansion bands and expansion gains will be much more difficult, IME.

 

In general, I've abandoned using them since I've never actually liked the results ("breathing" included...).

 

Chris

[Chris A]

Here is a link to the portion of this thread discussing that subject: https://community.klipsch.com/index.php?/topic/143320-loudness-war-and-the-dynamic-range-dr-database-some-observations/page-4#entry1631214

[mustang guy]

Chris, you are a brilliant guy, and I find that your threads on the multi entrance horn and this remastering thread to be two of the best ever on the forums. Some day, you should put up a streaming website for all your remastered stuff. Licensed and royalties paid, of course. I would subscribe...

[Quiet_Hollow]

I would subscribe...

Me too...

[Chris A]

It's also a good skill to develop as a music lover and an audio enthusiast--since it is more than capable in saving lots of money spent on frivolous attempts to improve the music through buying new and expensive gear, and multiple versions/copies of music in the quest to find better sounding recordings. Once you unmaster a recording with good starting dynamic range--I find that the need to replace those music tracks with better ones subsides.

 

One thing that I might've mention much earlier is that the best CDs to start the unmastering process with--are the oldest CDs that you can find on Amazon marketplace and they're usually the cheapest ones, usually selling for less than a dollar(US) plus $3.99 shipping.  If you check the DR Database site you can find some of these CD versions having 3-8 dB higher DR ratings than the ones remastered in the early 2000s to the present.  Those old versions by far yield the most outstanding results after unmastering, IME.  This all assumes that you're choosing music originally released before the 2000s.  But I've also noticed that there are many others are listening to these recordings from earlier time periods even though they're younger. 

 

There are also a few instances of released music within the last 25 years that also has higher DR Database ratings.  Unfortunately it's a pretty small number of titles.  I've also run into some music titles that have what I'd call excessive levels of limiting (clipping) that can be reconstructed back to their nominal un-clipped states using clip fix, after which their DR Ratings jump by as much as 8 dB with judicious use of the add-on "brickwall limiter" to bring the excessively large reconstructed peaks back down to something more reasonable, but without hard clipping injecting huge amounts of transient HF noise.  These tracks, once unmastered, sound extremely dynamic and natural without the stressing levels of HF clipping occurring continuously in the tracks before unmastering. These examples include Hiromi, Esperanza Spalding, Arnold McCuller, Avril Lavigne, Stacy Kent, Loreena McKennitt, Enigma, The Crystal Method, Tal Wilkenfeld, Jean-Luc Ponty, Mickey Hart, Stevie Ray Vaughan, Marcus Miller, Diana Krall, and several movie soundtrack CDs, etc.

 

Chris

[Schu]

Isnt there any "auto program" available?

[Chris A]

I wish there were.  The problem is that, when you get to the more popular music genres (i.e., not classical or jazz), the mastering varies all over the map from track to track on the same album/title.  This includes clipping and residual line noise, also. I've found that I have to check each track after de-clipping, re-equalizing, and sometimes line noise removal--sometimes even the left track vs. the right track in terms of noise removal. 

 

Classical, jazz, fusion, and a few other genres are typically much more consistent in their mastering tactics than the genres mentioned above.

 

The worst thing that can happen to a track is if soft limiting is used to excess--more or less continuously on every cycle instead of just the peaks in the mixdown tracks, or if EQ and soft limiting are performed differently every few seconds on a track and joined together (...yes, I've seen that...).  Then the music is basically lost.

 

Chris