Jump to content

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


Chris A

Recommended Posts

One of the subjects of the thread on "Digital vs. analog" included a discussion of the loudness war (http://en.wikipedia.org/wiki/Loudness_war) and an online database that is systematically measuring the dynamic range of recordings: http://www.dr.loudness-war.info/. Anyone can query this database for their favorite CDs, SACDs, and vinyl rips to disk to see if their particular version of a recording has been subjected to Loudness War compression techniques. In order to do a database query, you put in an artist name and/or album into the search fields.

 

The DR database allows its users to measure their own disks if they happen to not already be in the database using a freeware downloadable application called DR Offline Meter. You can then upload the results of their analysis of their own disks so that other may share with others and expand the database.

 

Observations

 

1) CDs produced in the 80s and in 1990 uniformly have much more dynamic range than those produced/remastered in the last 20 years, particularly in the last 10 years. I've not been disappointed with most of my 80s-vintage CDs but I've been very disappointed with CDs from the late 1990s/2000s, especially "remasters".

 

2) Some disks that are recorded at a very quiet level sound "dull" if played back at levels that are typical for most of my other disks. However, when the preamp gain is increased, the recording comes alive. Horn-loaded loudspeakers, such as Klipsch, are fully capable of handling very dynamic recordings without accompanying audible distortion that is typical for direct radiator [cone-type] loudspeakers.

 

3) Disks that have very high DR ratings but that also sound very good without a great deal of boost typically have a great deal of relatively quiet instrumentation and a lot of dynamic percussion. The "James Newton Howard and Friends" CD is a gold-standard example of this. Other recordings include "Bolero" and "The Planets"--especially "Mars, the Bringer of War", which in both cases build to a very high SPL from very low SPL beginnings. These recordings are notable in that I find myself jumping up to turn the volume down at least twice during the performance when there is anyone else in the room that doesn't prefer to have loud music playing, or they want to hold a conversation.

 

Chris

Link to comment
Share on other sites

Most of my audiophile test disks and recordings all sound dull until turned up, then they come to life. Loud recordings sound loud at mid and high volume and "background level" at low loudness levels - like elevator music. Perhaps a better approach is to campaign for more use of signal compressors when playing dynamic music at lower than concert levels, similar to what DrWho was saying in the "Digital vs. analog" thread about hand-held portable music devices being able to compress on the fly.

Chris

Edited by Cask05
Link to comment
Share on other sites

I was wondering if any other folks here have experience this same effect using their most dynamic recordings: do they sound "dull" to you until boosted a great deal?

Sure. Depeche Mode's Violator, Lionel Richie's Cant Slow Down, and Pink Floyd's Wish You Were Here come to mind. All of them sound great, but require a healthy turn of the knob (usually on the order of 15-20 dB higher) in order to really get them signing.

What you describe above is not uncommon in Home Theater where people are dealing with wide dynamic material all the time. Well engineered movies frequently contain a dynamic range of up to 20 dB.

The number one complaint:

"I have to turn it up to listen to the dialog, only to get slammed to the back wall when something loud happens."

That's why HT AVR's have compressors built into them. It's usually labeled "Dynamic Range", "Night Mode", or "DRCOMP" in their menus.

The user can turn that on to squash down and level the mix. Quiet won't be so quiet and loud won't be so loud.

On an AVR, it's just the push of a couple buttons. For a 2-CH rig, you're stuck with having to insert a compressor unless you're already using a DSP.

Link to comment
Share on other sites

What you describe above is not uncommon in Home Theater where people are dealing with wide dynamic material all the time. Well engineered movies frequently contain a dynamic range of up to 20 dB.

I have noted that quite a few two channel folks here don't use/invest in HTs. I've been aware HT dynamic range issues since my two-channel rig is also my HT rig: we split about 50-50 on video and music playing time.

I'm also acquiring a lot more multi-channel music disks now. Having 5 speakers in surround + 2 subs makes the job of reproducing high SPL transients easier, and playing multichannel music disks can make that performance stunning, i.e., the digital formats beyond red book CD in reproduction quality.

Chris

Edited by Cask05
Link to comment
Share on other sites

I systematically downloaded the DR Database and placed the data into a spreadsheet in order to do more analysis--and I see some interesting features: one of note is the vinyl vs. digital DR aggregate values spread over time.

Anyone see any issues with this plot? (Remember that CDs have at least 30 dB more SNR than vinyl.)

391055d1388616174-loudness-war-dynamic-range-compression-dr-database-observations-drdb_vinyl_vs_digital_timeline.gif

Link to comment
Share on other sites

About the time that CDs came out the studios began to use near field monitors for recording and mixing. These speakers have a sensitivity of around 86 dB or so and can't reproduce the dynamic range of the soffit mounted monitors they replaced. This seems to coincide with a general reduction of DR on all consumer formats as studio use of nearfield monitors became prevalent.

  • Like 1
Link to comment
Share on other sites

Here is a plot using only the reported "maximums" for each album/CD/digital track collection--it also shows that even the "uncompressed" tracks, relative to vinyl, are still compressed:

 

391030d1388614448-loudness-war-dynamic-range-compression-dr-database-observations-dr_max_vinyl_digital_timeline_cor1.jpg

Link to comment
Share on other sites

About the time that CDs came out the studios began to use near field monitors for recording and mixing. These speakers have a sensitivity of around 86 dB or so and can't reproduce the dynamic range of the soffit mounted monitors they replaced. This seems to coincide with a general reduction of DR on all consumer formats as studio use of nearfield monitors became prevalent.

Thanks for that insight, Don.

I noticed that even Doug Sax in his new mixing/mastering studio uses much lower efficiency cone-type loudspeakers than he used during the days of Sheffield Lab. It must be an industry-wide thing with no one able to withstand market forces driving to "lo-fi".

Chris

Link to comment
Share on other sites

I'll let you know how that exercise turned out.

Curiosity piqued. Standing by until further notice. [&]

Anyone see any issues with this plot? (Remember that CDs have at least 30 dB more SNR than vinyl.)

Yup. I noticed that too. Lame, to the point of, "why even bother?" [N]
Link to comment
Share on other sites

Yup. I noticed that too. Lame, to the point of, "why even bother?"

I believe that it is strongly related to WHAT are you listening to. Multiply the standard deviation (SD) bars by ~3 and add to the "max" data values of each bin to see the actual maximum Dynamic Range (DR) data values for the two recording formats.

Chris

Edited by Cask05
Link to comment
Share on other sites

Chris et al,

I have been following this thread and have been hesitant to jump in. I still don't know if I should but .............Live music perfomances are dynamic. The more dynamic a recording is the more alive it sounds IMHO. Any compression beyond what is absolutely necessary to squeeze it onto the recording medium is a shame, again IMHO. And the overcompression and then boosting the overall level is a disgrace. But hey I'm just the average listener that has a system that is capable of concert level reproduction at low distortion levels.

Eric

  • Like 1
Link to comment
Share on other sites

Live music performances are dynamic. The more dynamic a recording is the more alive it sounds IMHO. Any compression beyond what is absolutely necessary to squeeze it onto the recording medium is a shame, again IMHO. And the over-compression and then boosting the overall level is a disgrace. But hey I'm just the average listener that has a system that is capable of concert level reproduction at low distortion levels...Eric

 

Amen, brother: no arguments here.

 

What I had intended to do at this point was to report on different year releases of the same music from the '80s vs. 2000s on CD.

 

In at least one instance, the 1981 CD (Best Shots, Pat Benatar, Chrysalis), which is generally very high SPL music in its live format, sounded much worse than the 2005 CD with most of the same tracks (Greatest Hits, Pat Benatar, Capital). The 1981 disk suffers from peaks that are above 0 dB (see the database details for the word "over"...) implying that the mixing or mastering engineer really didn't do their job very well. However the 1981 CD has much higher DR ratings across the board (12, 9, 15) vs. the 2005 version (7, 6, 8). Needless to say that the 2005 version sounded better, i.e., no clipping. It also sounded very "loud", but listenable.

 

If you want a disk that doesn't compress, seek out the 1981 version of the Tower of Power's Direct (CD-17, Sheffield Lab) with DR Database ratings of 17,15, and 18. I had to boost playback levels in excess of 16 dB to bring this disk's (-24 to -27 dB average level) playback up to about the same listening level as the Greatest Hits disk mentioned above (recorded at -7 to -8 dB average level). However, I detected no loss of resolution or quantization effects on the Tower of Power disk even though it was recorded at much lower average level.

 

What I'm beginning to see is that CDs ought to be recorded at the -20 dB average level or below in order to capture the natural peaks: what I'm witnessing now is that most CDs today are being put out at the -5 to -7 dB average level. That's the problem. If recording standards addressed average loudness levels like the standards do for THX and other DVD/BD formatted disks, we wouldn't be talking about this subject here.

 

Chris

Edited by Chris A
Link to comment
Share on other sites

Amen, brother. No arguments here.

Chris,

That's why I hesitated. I knew I was preaching to the choir, just stating the obvious. Most of us Klipshophiles love high DR(because our systems can reproduce it). We know what live music sounds like and we want our home systems to sound as close as possible to that experience.

What I'm beginning to see is that CDs ought to be recorded at the -20 dB average level or below in order to capture the natural peaks: what I'm witnessing now is that most CDs today are being put out at the -5 to -7 dB average level. That's the problem. If recording standards addressed average loudness levels like the standards do for THX and other DVD/BD formatted disks, we wouldn't be talking about this subject here.

Yes and what can we do to make the recording industry take notice? Has the ship already sailed too far out to sea?

Eric

Link to comment
Share on other sites

Yes and what can we do to make the recording industry take notice? Has the ship already sailed too far out to sea?

In response to the loudness war CDs: I'm buying digital disks that have more dynamic range by using the DR Database as a guide.

I don't buy popular, i.e., big record company, CDs/SACDs with low DR as new disks. If I want one of these discs, I wait and buy them used off of Amazon Marketplace. That way the big music companies don't directly increase their profit on low DR disks. I really don't buy vinyl. I'd steer clear of low DR vinyl disks if I did buy/use them.

If more people started to do this, the record companies might eventually start to put out higher DR disks. Sales feedback is apparently the only thing that matters to big music.

Chris

Edited by Cask05
Link to comment
Share on other sites

Here is a plot of the total DR Database ratings counts by format (i.e., vinyl, digital). This one should really begin to raise eyebrows. Look at the bi-modal characteristics of the digital data, implying that there are two types of digital disks in terms of Dynamic Range:

 

391032d1388615105-loudness-war-dynamic-range-compression-dr-database-observations-count-dr-rating-dig-vinyl.jpg

Link to comment
Share on other sites

What I find more disturbing as I plug away at my collection using the DR offline meter software is:

An absolutely positive correlation between albums that I know sound tough to my ears (fatiguing) and clipped recording levels.

If it's over 0 dB, it's toast... junk... done... regardless of dynamic range. [N] There's no way it'll ever sound clean.

I can't begin to describe the harmonic nightmare that ensues when my amp is fed a clipped signal. Straight-up ***, whether it's Huey or Yo-Yo Ma.

Over? It's all over for this one alright... [8-)]

post-42237-13819832642364_thumb.png

Link to comment
Share on other sites

Quote

An absolutely positive correlation between albums that I know sound tough to my ears (fatiguing) and clipped recording levels.

 

Agreed: I hear exactly the same thing. Finally I have a way of measuring this form of clipping in any CD - and it's free of charge using the DR Offline Meter application.

 

When I look at CDs produced in the early 1980s, I have seen one or two discs that have clipped transients as you identified above AND average recording levels of -10 to -15 dB (instead of the -20 to -25 dB levels seen in the best disks that I own).

 

EDIT (30 December 2013): I've noticed that many discs from Japan in the 1980s had "pre-emphasis" applied to them. (See sites on CD pre-emphasis for a discussion of this technique used in the early days of CD players that had digital filtering problems.) When these CDs are played back with almost all computer-based software or some later DVD players and gaming centers (such as the PS3), they don't handle pre-emphasis EQing properly and these discs sound like vinyl records played back without the RIAA curve applied - overly hot on the high frequencies. I believe this is at least one source of dissatisfaction with CDs produced in the 1980s. There are simple fixes for computer-based players that rip these discs to lossless format available. Ive found that my Oppo player correctly decodes these CDs, but my PS3 (second gen) doesn't.

 

Additionally, I've found that most other CDs of that time period that were recorded using the exact same techniques used for producing vinyl disks, i.e., already reduced dynamic range due to signal compression being used during the recording process. In other words, the record giants systematically never used the advantages of the CD format.

 

Only a few boutique companies seemingly tried to use the increase in dynamic range, SNR, and freedom from other kinds of signal degradation associated with analog editing processes. One of those companies was DMP (e.g., Flim and the BB's CDs), and Sheffield Lab (James Newton Howard & Friends, Tower of Power Direct, etc.), which are to this day my reference CDs in terms of recording and production quality.

 

Chris

Link to comment
Share on other sites

Some more observations now that I've uploaded perhaps 50 or more DR rating test scan results:

 

1) My initial observation that there is a correlation between disks that I like to listen to and their track-level DR ratings have borne out, even more than I was expecting. I've found that certain disks were louder than others from the same artists, and I found that these are the disks that typically stay on the shelf. This is a significant finding--perhaps the most significant of all.

 

2) I've done a little digging on the subject of dynamic range (DR) and have found that the same expectations of DR on images and video also apply to audio. But while video and images usually enjoy at least 10 stops of dynamic range (i.e., -30 dBFS, or a DR rating of about 40) with 14 stops being the dynamic range of the human eye (i.e., -42 dBFS, or a DR rating of 51), audio tracks typically have maximum DR on the order of 7 "stops" (20 dBFS, or a DR Database reading of about 16-17). For reference, the latest digital movie cameras have a DR or greater than 15 stops (45 dBFS).

 

3) The latest loudness-war CDs can have an average of less than 2 "stops" of dynamic range (i.e., -6 dBFS or a DR Database rating of ~3), but I've found that I cannot comfortably listen to CDs with a DR rating of less than ~9 (-6.6 dBFS, or about 2.2 "stops").

 

Chris

Edited by Chris A
Link to comment
Share on other sites

I'm in total agreement with you on this, Chris. [:S]

I really like Paul McCartney, but I got his 2005 album 'Chaos and Creation in the Backyard'. I find the songs to be great McCartney songs, while the CD is almost unlistenable. Highest DR is 8. His newest is even worse, which is telling me I won't even bother to get it.

Interestingly, his 'Liverpool Oratorio', a 2 CD set, max is 16 and 18. I'm guessing the more classical instrumentation persuaded them to keep some dynamics.

Bruce

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...