The Unofficial Dynamic Range Database

At the 2012 NYC hi-fi show, I was chastised by a random showgoer for using a “dynamic-less” recording as a reference in the Legacy room. Now with the help of the Unofficial Dynamic Range Database, I can find the dynamic range measurements of many of my favorite recordings, including the record ridiculed by one spiteful audiophile. Was he right about its dynamics?

The “Unofficial” Dynamic Range Database uses a sliding scale from 1 to 20 (1 being the worst, 20 being the best) to rank the dynamic quality of each of the recordings they list. This number represents the difference between the peak decibel level on a recording and the recording’s average loudness. DRD applies the following descriptors to these ranges: 1-7=bad; 8-13=transition; and 14-20=good. Evaluation for each album on the website includes the album’s average dynamic range, the track with the weakest dynamic range, and the track with the greatest. The Unofficial DRD also provides individual dynamic range measurements for each track on the album.

For example, Phish’s Lawn Boy, the record which drew so much resentment out of a fellow audiophile, scored an average of 12 in terms of overall album dynamics, which is considered close to “good” or “in transition” by DRD’s sliding scale. So still a little compressed, but certainly not as bad as that guy made it out to be: my CD made him storm out of the room! The weakest track clocks in at 9 (“Split Open and Melt”) and the most dynamic song on the record, “Bouncing Round the Room”, scores a 14. Interestingly, “Bouncing” is the only song on this record that received a medium level of radio airplay, and it was the most dynamic! Just goes to show that not all radio singles have to be LOUD.

Participants on the Dynamic Range Database use the Dynamic Range Meter, a foobar2000 component, to take their measurements. The Dynamic Range Meter is designed to give the same information as the popular Pleasurize Dynamic Range Meter. Any user can submit their test results to the site.

On the site, users can search by artist or album. The mainpage allows you to sort alphabetically by Artist or Album or rank their dynamic rankings in ascending or descending order.

After sorting the albums by worst dynamic range, I was not surprised to see some of my favorite bands and records at the top of the list, including My Morning Jacket, The Mars Volta, and the Red Hot Chili Peppers. What I found in the ultra-dynamic recordings was also refreshing: Van Halen, Talking Heads, and Rush.

Kudos to Stereophile.com user lastgoodbye for the tip. Sites like these give consumers the power of information, and with that comes the power of smarter decision making, which will ultimately lead to better quality by demanding more in what we choose to buy.

COMMENTS
Kal Rubinson's picture

I wish it was searchable or in a coherent order (did I miss it?).  Couldn't find anything that I could relate to except some older classical which didn't do too bad.

Kal

michaelavorgna's picture

...before listening for the same reason I listen to paintings before determining what's worth viewing. 

nile49's picture

of certain formats, I believe the old foobar plugin optmistically miscalculated the dynamic range of mp3's and possibly other formats.

matt123's picture

Ironic that Elbow's Build A Rocket Boys!, the 2011 Dynamic Range Day winner for best-sounding dynamic album scores a "bad" 7 or 8 in this data. The 2012 winner Björk for Biophiliascores an 8 also.

http://dynamicrangeday.co.uk/award/

I checked the other finalists and only one vinyl copy scores a "good"
Laura Marling - 8
LCD Soundsystem’s This Is Happening, 10, 15 (vinyl)
The Coral’s Butterfly House, 8, 13 (vinyl)
Massive Attack’s Heligoland, 8, 12 (vinyl)
Neil Young’s Le Noise, 9
Mumford & Sons’ Sigh No More, 8, 12 (vinyl)
Four Tet’s There Is Love In You, 9
Norah Jones’s …Featuring, 9

I think this is a really valuable service, and I don't want to dis what seems to be a volunteer effort - but this has to bring it a bit into question, maybe its the plugin as nile49 says above.

Would be interesting to understand why the vinyl scores better - I assume they are vinyl rips, so is it the ripping process, the format itself or mastering?

Ladyfingers's picture

It's indicative of how little dynamic range the average non-classical recording has at the moment.

I use this database very reliably to see which version of a favourite recording to buy, as while a high DR is no guarantee of quality, a low one almost certainly guarantees poor sound for anything with percussion. 8 actually isn't too bad, but the 12s and 13s in comparison really do sparkle and punch. Good late 80s and early 90s CDs sound AMAZING when you turn them up.

Using the DB, look for instance at Rush's albums, and compare the modern remasters to the older CDs. They are something of a disgraceful benchmark in overcompressed, clipped remastering, and it's clearly shown there.

Standard mastering practice is so abysmal now that it beggars belief. Vinyl scores higher in general because it is now accepted that digital audio MUST sound a certain way. My friend has a audio software from Steinberg whose final step in the recording process is a 0dBFS brickwall limiting with no explanation why it must be done.

You'll find that older CDs actually had higher DR than same-era vinyl and that at the time this was a selling point mentioned on the packaging ("may reveal limitations of the master tape"). Now comparatively limited vinyl, hilariously, is the audiophile medium because the better format has been abused.

Ariel Bitran's picture

Kal, you can organize the results either alphabetically or numerically by clicking on the category. You can also search by artist or album in the search toolbar just above all the listings (right corner).

 

ML: are you being sarcastic or was that a typo? just wanna be sure. ("listen" to your paintings)

 

Nile and Matt: the question of the validity of the listings is a great one, and one that I honestly avoided answering b/c I can only assume that some of these listings are incorrect and I just wanted to bring light to what a great idea/site this is. Content is user submitted. As the wikipedia of dynamic ranges, user submitted content is surely up to question. Hopefully if this website grows, they can enlist editors/moderators or a system which ensures accuracy.

 

Re: certain formats -- the listings indicate whether the recordings were measured with lossy or lossless formats, so hopefully that can assist in making a judgment.

 

Re: dyanmic range day winners -- this is crazy! none of those albums really scored above a 14 except the LCD vinyl! Great research. Certainly raises some interesting questions....

michaelavorgna's picture

I think it’s important to realize that the relationship between these “DR Scale” scores and perceived sound quality is not 1 to 1. In other words, not all dynamic range compression sounds the same even if it gets the same “score” from the Foobar2000 software tool.

Or to put it in plainer language, you have to actually listen to music to know if you’ll enjoy it or not.

I also think it would be a shame and a mistake if people started to make music purchase decisions based on this information. I’m not promoting dynamic range compression, and we’d all be better off without it, just trying to avoid another unnecessary witch hunt.

VILLAGER #1: If... she.. weighs the same as a duck, she's made of wood.
BEDEMIR: And therefore--?
VILLAGER #1: A witch!

kovan yarrum's picture

Michael, you are spot on.

Sometimes how the files "look" have very correlation with how they sound. 

It can work both ways...one that looks like it has plenty of headroom may sound just ok, and one that looks like it may be clipped can sound good.

I sent some Audacity screen caps of the recent Queen remasters to a friend who said they "looked" good and should sound great, when in reality they are as harsh and clipped sounding as any remastering I have had the (dis)pleasure of hearing.

Kal Rubinson's picture

Sorting was useless as there is no naming consistency.  I did sort by date and gave up looking when I found no classical in 2012, 2011 and a page or two of 2010.   I guess the problems lie elsewhere.  cool

John Atkinson's picture

...tend not to suffer from unnecessary dynamic compression, so the database's problems are less important than in rock, where too many recordings from the past 10 years have had too heavy on a hand on the "Squash-it-to death" knob.

As an example, I mentioned in my presentation on this subject at last October's AES Convention the 1932 recording of Elgar's Violin Concerto, which had greater dynamic range than almsot all modern rock recordings - see www.stereophile.com/asweseeit/177/index.html.

John Atkinson
Editor, Stereophile

JItterjaber's picture

I appreciate the chief over at Audiostream for speaking up.  

It is also important to remember that dynamic range is not just a result of an engineer compressing/limiting...It is also a product of the type and style of music being played. More so, It can be over compressed in the mixing stage...before it is then squashed in mastering. 

Seriously, use YOUR EARS! 

I explored this topic a bit on my blog: http://bit.ly/IqmDVU

I for one appreciate the mastering engineers over at "Dynamic Range Day." They are trying to be advocates in an industry where non-engineers make final sound level decisions. Learn, pass it on.

Ariel Bitran's picture

the relationship between audio hardware and measurements.

there are many times when an amplifier sounds great to the reviewer, but JA finds that it measures terribly! (I'm loooking at you tube amp)

I think its more of a question of what do these measurements bring to the table and what can they help us learn and can awareness of the issue positively affect the current standard? 

If you sort the recordings by worst average dynamic range, almost all of the ones at the top of the list are "new" recordings released in the past 10 years. Just b/c something is compressed surely doesn't mean that we shouldn't buy it, but to ask for better isn't a sin either. like with Metallica's Death Magnetic, fans heard the record, thought it sucked (not the music, but the sound quality), and asked for a better quality release. these measurements are a great form of documentation and community building and education to perpetuate the attitude of wanting better. not a bible but a travel-guide.

michaelavorgna's picture

Well put and agreed. 

But...The notion that we can cleanly and clearly separate sound quality from music quality in order to focus on the former runs the greater risk of losing sight of the latter. Case in point being a lot of the music you hear at hi-fi shows.

Ariel Bitran's picture

I agree.

My question is the following:

Is there any room for absolute measurement in terms of deciding the quality of an audio recording? does this have place or is this a lost cause since the ratio of quality to dynamic range rating is not 1:1? or what is the healthiest way to use these numbers?

michaelavorgna's picture

It’s music. Just listen.

The longer answer…

Your Metallica example is an interesting one in that people objected to what they heard and asked for better - the fact that they objected to the amount and severity of dynamic compression used is an important and interesting fact, after the fact. It allows for a common understanding of the issue as well as a solution.

An attempt to reverse this order by looking at a list of recordings and their associated “DR Scores” shifts our focus from listening to looking which strikes me as a step in the wrong direction. To reiterate, the ultimate quality of a recording must also take the music into account. Otherwise we’re left with nothing but sound. Since there’s no way to measure musical content to determine value, there’s no way to graphically represent a value that takes both sound and music into account. Just as there’s no way to listen to the value of a painting.

So no, I cannot see “any room for absolute measurement in terms of deciding the quality of an audio recording” unless we’re uninterested in the music.

Ladyfingers's picture

There are plenty of things that indicate objectively poor sound quality, though. Most of the low DR recordings in the database do, in fact, sound pretty terrible.

michaelavorgna's picture

But the only way to know if we're interested in the music is to listen.

Some terrible sounding records are certainly worth listening to, just as some great sounding records are not. In my opinion of course because ultimately we're talking about music not sounds.

Ladyfingers's picture

But, given the choice between a poor and good master of the same recording, which would you choose?

michaelavorgna's picture

The one I prefer listening to.

Ladyfingers's picture

...how bad any 5-6DR mastering sounds compared to a 12DR version of the same record?

There has to be something very seriously wrong with the old mastering for it to sound worse than the modern, compressed-to-hell version.

michaelavorgna's picture

I was not thinking about using the DR Ratings to compare reissues/remasters with original recordings. Here I agree – a reduced DR Value in the reissue would indicate something’s not right.

Regarding reissues, I tend to prefer the original vinyl (if available).

TriodeDave's picture

would be to ignore them. If you already have a recording and enjoy it, will you chuck it when you discover that it's not on an Approved Dynamic Range List? Conversely, if you're told that a recording you're considering purchasing 'cause you like the artist or you've heard the tunes are great isn't on the ADRL, will you change your mind buy something "approved" instead?

If you do either or both, you'll end up with a shelf of ADRL members, but you'll spend very little time listening to them, perhaps taking up aggregating a giant ball of twine instead.

If you do neither, you'll end up with a collection of music you enjoy with varying amounts of dynamic range represented.

Unless your favorite thought is, "Well there's two (minutes, hours, days) I'll never get back.", why bother?

roscoeiii's picture

Based on a quick search I am also seeing that vinyl copies are tending to have a greater dynamic range than digital, including 24/96 digital. To choose an example of recently recorded albums, check out the dynamic ranges for the the Fleet Foxes: 

http://www.dr.loudness-war.info/index.php?search_artist=fleet+foxes&search_album=

Ladyfingers's picture

You can't apply the same brickwall limiting to vinyl. Vinyl mastering needs a certain RMS volume level to sound good because of all the surface noise, and if that is exceeded, then the needle can literally jump out of the groove. It's an inferior medium, but its limitations encourage better mastering. Very sad.

PredatorZ's picture

Some of my favorite recordings are live , IM assuming that they must have minimal production and tweaking by some "engineer", for example Judas Priest Live, and Muddy Waters Folk Singer, have a real tactile feel of the venue. I wonder how these 2 recordings would rank in the tests. It might be a usefull tool for exploring unfamiliar music and making choices of recordings to audition. Bottom line, as mentioned in prior  posts, is listening. Great thing is,  thats why we are all here, our love of music and reproducing the best possible experiences with the systems we have.

 

Cheers Bret

Edit: Downloaded and ran the foobar2000 component with results now posted on the website, Muddy Waters Folk Singer rated 15 overall, 17 on its hiest track, Judas Priest...Live rated 13 overall, and 14 on its hiest track. Im going to submit my favorite albums as far as my "percieved sound quality" and see how this works out :)

tmsorosk's picture

I always chuckle when someone kicks into one of these discussions and says, " use your ears ". doesn't that go without saying, or at least hope everyone partaking in an audio forum understands that by now. I find it interesting to understand a little bit about one perimeter of why we may or may not like the sound of a certain album, musical content aside.

                                                                                                    Tim

james's picture

A Picture of Nectar, which is listed as Phish's most dynamic album is also the worst sounding to my ears. It sounds so thin and brittle.

Ariel Bitran's picture

haha james! I actually agree. I could never understand why Nectar was so brittle. I think Lawn Boy sounds great.

Rift is a little squashed but super clean sounding (like Lawn Boy), and I'm dying to hear the Junta just released on HD tracks.

james's picture

glad someone agrees. try making it all the way through Cavern.

The new Junta sounds great. I got the vinyl copy and it does sound excellent. Quickly A/B'ing back and forth between the original and the new version, it's not jaw dropping, but it does sound very good. 

I could listen to something like "Anything But Me" from [i]Round Room[/i] all day. The sound of Trey's guitar on that track does it for me.

Patrik Lundquist's picture

Google "MasVis analysis" for more in-depth mastering analyses of CDs.

Also, the Dynamic Range DB isn't completely trustworthty due to a bug in the official DR meter:

The reason for the usefulness of the DR Meter despite of its flaws had always been that large DR value database /.../ - now because of that bug all the thousands of records scanned from lossy sources are basically rubbish.

http://www.hydrogenaudio.org/forums/index.php?showtopic=88373&st=0&p=769036&#entry769036

jasenj1's picture

I just recently discovered the DR database. I promptly downloaded the tool and began running a bunch of my albums through it - and submitting entries.

Some of the discussion here has pitted audio quality against musicality. I would think anyone reading Stereophile values BOTH, and the balance between them is dynamic. I might have an album/song that really speaks to me and moves me for its lyrics, rhythm, emotion, etc. I would enjoy such music whether it was played through an AM radio or a $100,000 Stereophile-worthy rig. But, because I care about such things, I would choose the expensive rig given a choice. Likewise there may be recordings that are exceptional - great dynamic range, clarity, subtleties, sound stage, etc. In these, I admire the craft of capturing the performance; a high-quality playback rig enhances (and is probably necessary for) that enjoyment. And certainly, certain types of music lend themselves to high DR while others do not.

So with all those caveats and qualifiers in mind, a higher DR recording is an indicator of a better recording than a low DR. And if the DR is low enough it is a strong indicator that the recording is severely compromised.

Snarf's picture

I must say it is a lot off line last few weeks.
Furthermore it is for me a very usefull tool. I listen (mostly)to 60 70 80 music based on my CD collection (ha ha i'm 59). I remix an make Dj set based on Deep house beats with 60 70 80 undertones. Now for the latest (Deep house beats) you want to have the opposite of a High DR rating all this Deep House suff is quite compressed. But even so it is an art to compress it such that you can listen for houres.

X