How "accurate" is the DR Meter?

I've done some comparisons recently between different pressings and masterings of a couple of albums. In the process, I've discovered that the foobar DR Meter isn't always a real indicative of a recording's dynamic range. For example, some of the DCC titles score consistenly lower DR values (sometimes by 3 or even 4) than some other pressings/masterings, yet we know they aren't compressed. How is that possible? Can something else, like EQ for example, "fool" the DR Meter and influence its measurement?

It makes me wonder...

 

No idea, but the Celebration Day DVD stereo mix (which contrary to what was said at the time of release is a different mix) looks less dynamic and more compressed as a waveform, but is somehow less taxing on my ear than the cd mix, which looks less compressed. And don't forget how many factors in a recording contribute to its dynamics. A normal Rock song usually is less dynamic than a normal jazz tune, simply by design of the music

 

It's probably very accurate at measuring whatever it measures.
Perhaps the question should be "what, precisely, does it measure?".
There is some degree of subjectivity to perception of dynamics and loudness.
Replay Gain does something similar - how does it correlate to DR scores?

 

The dr meter just shows the lowest dr score in each song. For instance, a song could be super compressed at one part of the song and have say dr 3 for 10 seconds but the rest of the song could hardly have any compression at all and have lets say dr 15 and the meter will show it as dr 3, it doesn't show the average dr. It can be a good indicator of how compressed it is for shorter songs at least but I don't trust it at all with songs say 30 minutes+ in length.

 

This is also what I have noticed. When trying to decide which version of a song or album I'm going to go with, the DR score is one of many factors I consider. It is not the be-all, end-all. I do hope that heightened awareness will cause a change in attitudes among those in the music industry for the better.

 

This is also what I have noticed. When trying to decide which version of a song or album I'm going to go with, the DR score is one of many factors I consider. It is not the be-all, end-all. I do hope that heightened awareness will cause a change in attitudes among those in the music industry for the better.

 

This is also what I have noticed. When trying to decide which version of a song or album I'm going to go with, the DR score is one of many factors I consider. It is not the be-all, end-all. I do hope that heightened awareness will cause a change in attitudes among those in the music industry for the better.

 

I don't take the readings as Gospel but more as a guide, been doing quite a bit of needledropping lately and have noticed "especially with 80's recordings" the lp has a lower DR than the original cd counterpart, I have some DR6 cd's that actually sound pretty great, and a few with 15+DR that are awful

 

They aren't Gospel, especially when it comes to classical, lots of jazz, or many prog rock recordings. When it comes to stuff like mainstream pop, country, CCM, etc. the readings are accurate. EQ, the width of the stereo mix, etc. can fool the DR meter as well. With a lot of DR narrow stereo mix but not heavily compressed in the mastering and already have a DR 10, you can widen the stereo mix using a sound editor and magically, the reading will turn out to be DR12+. On modern recordings that typically have 5-7 DR readings, you can lower the gain, use the Subsonic Rumble filter, raise the gain, save the editing and the DR readings will be DR 10+.

 

I remember reading another thread, where the poster wonders if a DCC/AF mastering is inferior to the remaster because of a smaller DR. I believe an answer was that the remaster used treble boost, which will give the recording a higher DR. Though this practice does not preserve the sound fidelity of the original recording.

 

Perhaps you can provide some specific titles and the other masterings as examples of this difference, as I’ve only found a few titles that may have 1-2 DR difference and that small of an amount really doesn’t matter.

The DR value is an average of averages and really should only be used as a quick snapshot of what the album dynamics may be.

 

The question is what causes that difference? On some albums it is fairly consistent, for example the Doors catalog when compared to the originals, or the Eagles Hotel California...

 

Well I’ve never seen one that had that much difference in DR values (3-4). So unless you provide an example, how can anyone respond accurately?

A DR value difference of just 1-2 can be due to a lot of different things including EQ moves by a mastering engineer and really shouldn’t be referred to as compressed audio even though it may reduce the DR value.

 

This isn't true.

http://www.pleasurizemusic.com/en/our-aim

Look under "What does the DYNAMIC RANGE METER do?"

 

Well I tested it before, with an album that has 3 long songs that are about 20 minutes long each (disc 2 of Boris- Dronevil final) the first song is Dr 9, the second song is dr 3, the last song is dr 5. I used Audacity to combine the 3 songs together into an about hour long song. I then used the meter on the 3 songs combined and the dr rating was only 3, which is obviously not correct.

 

Even if it's not perfectly accurate, it sure shows the crappy 'industry norm' of masterings going on of late. Check the standard cd release vs. the 'unmasters'...WHY do people put out such horrible sounding product ? I don't get it.

 

This doesn't seem to be a very realistic or typical example though either.

 

Its not very accurate for longer songs but for comparing originals to remasters its quite useful, its good for shorter songs at least.

 

Yes, I would of course agree with that. It's value it for comparisons.

 

Got a better cart & stylus recently and was wondering how, if any, this would affect the DR values of my needle drops. I was hoping the improved frequency response would help at least a little bit. I was surprised when the DR values overall fell slightly compared with those that I did with a cheaper cart & stylus (I'm talking fractions of a number, such as a track going from a DR of 11.57 down to 11.29, but sometimes these cumulatively were enough that the DR meter lowered an album by a full point).

I went back and read this again:

Now I don't feel so bad. I think my new needle drops present an improved spectrum of sound, and with more definition than the older ones, but are probably a little flatter in terms of EQ than before. I went back and did a slight bass boost (1 - 2 dB) to a few of the new songs and their DR scores were right back to around the levels of the old cart. Made a test disc and did some A/B testing on the main home system and in the cars, and the new versions without the bass increase usually sounded best.

Moral of the story: Don't lose sleep over minute fluctuations in DR values. Gotta keep telling myself this is just a tool.