Jump to content
IGNORED

Why DR Values are so Misunderstood...


Recommended Posts

Rather than keep discussing this in the Beck thread, I thought I would ask my question in a separate thread. How is it that a tool that doesn't give an accurate indication of the mastering quality (compressed to an inch of a song's life) or the actual dynamics of the whole song (meaning the softest to loudest range) get to continue to be the tool used to find a really important component of the mastering quality? And I don't really care about what they do for TV commercials where the loudest 20% should be used to measure how loud the commercials are compared to the main programing. What I am saying is that shouldn't be the test for audiophiles. Why hasn't this standard been replaced for audio?

 

We are all in agreement that it can't be used for vinyl as it produces inflated DR values using the exact same master for the CD version as the vinyl version.

 

However, I have shown that it is misleading simply because it only uses the loudest (most compressed) 20% of the song which is not an indication of the dynamics of the song in some cases. In other words, the tool is bad and produces bad information.

 

The song in question is Wilco's Misunderstood with a DR value of 5. This is a popular song by a popular artist, not some obscure song without generally meaningful application.

 

Based on the DR5 rating, one would think this song is overly compressed and portions are artistically compressed to DR5. The problem is that whole other sections, perhaps more than 50% of the track, are DR11 and would be considered "not bad to good" in general, particularly if there were other positive qualities in the mastering.

 

7888d1379650828-new-wilco-hdtracks-wilco-misunderstood.png

 

To demonstrate that the DR value of 5 was basically meaningless, I clipped out the middle 120 seconds of the song and the compressed end of the song (some 43 seconds). I ran the TT DR Meter on the two tracks plus the original full file.

 

The middle 120 seconds is DR11.

The end 43 seconds is DR5.

The "two clips album" is rated at DR8.

 

----------------------------------------------------------------------------------------------

Analyzed folder: /Users/JK/Music/DR TEST FILES/DR Test Files

----------------------------------------------------------------------------------------------

DR Peak RMS Filename

----------------------------------------------------------------------------------------------

 

DR5 -0.11 dB -5.58 dB Misunderstood End Section.aif

DR11 -0.11 dB -13.63 dB Misunderstood Middle Section.aif

----------------------------------------------------------------------------------------------

 

Number of files: 2

Official DR value: DR8

 

================================================================================

 

Since the DR Meter uses the loudest 20%, the whole song is shown as a DR5 when it is really a DR11 for about half the song (or perhaps the average of the two, a DR8 is better?). I would argue that none of these numbers has ANY meaning in describing the dynamic range of the song.

 

In no way does the DR5 give an indication of the compression of this song. It does say that the artist wanted a huge wave of sound to crush you at various places for artistic reasons that fit the song perfectly.

 

In no way does the DR11 indicate the compression of the song although I would argue that the MOST dynamic 50% is a much better indicator of the dynamics of the song than the least dynamic 20%. A far, far better impression, quite frankly.

 

Thoughts and comments? I'm trying to learn why I should trust such a blunt tool that seems to mislead by it's very design.

 

Best,

John

Positive emotions enhance our musical experiences.

 

Synology DS213+ NAS -> Auralic Vega w/Linear Power Supply -> Auralic Vega DAC (Symposium Jr rollerball isolation) -> XLR -> Auralic Taurus Pre -> XLR -> Pass Labs XA-30.5 power amplifier (on 4" maple and 4 Stillpoints) -> Hawthorne Audio Reference K2 Speakers in MTM configuration (Symposium Jr HD rollerball isolation) and Hawthorne Audio Bass Augmentation Baffles (Symposium Jr rollerball isolation) -> Bi-amped w/ two Rythmic OB plate amps) -> Extensive Room Treatments (x2 SRL Acoustics Prime 37 diffusion plus key absorption and extensive bass trapping) and Pi Audio Uberbuss' for the front end and amplification

Link to comment

What if you compare the "Dynamic Range (DR)" and "Dynamic Range (R128)" values in JRiver?

 

I suspect that a heavily compressed album like Beck's Morning Phase would have low values for both, but the example you posted above will have low DR values (because it's heavily compressed in places) but a higher R128 value.

 

 

"Dynamic Range (DR)" values should match the TT-DR meter, and are a measurement of Peak-to-Loudness ratio. (PLR)

 

"Dynamic Range (R128)" values are R128 Loudness Range (LRA) measurements which are a statistically weighted dynamic range measurement.

 

 

Ultimately the tools are a guide and not a definitive statement on the quality of a track.

They are very useful when you are trying to compare different versions of the same album, and almost always correspond with what I hear, but they are not an absolute measure you can use to compare different albums with.

Link to comment
Ultimately the tools are a guide and not a definitive statement on the quality of a track.

They are very useful when you are trying to compare different versions of the same album, and almost always correspond with what I hear, but they are not an absolute measure you can use to compare different albums with.

 

Couldn't have said it better (bold by me).

 

To evaluate audio on a more sophisticated level, one should not use one tool alone, but try to use different tools (i.e. RMS and DR analysis plus waveform analysis, ...), and also listen to the music to get the whole picture - as good as it would get - together.

 

I have found the DR meter very helpful for my evaluations, but I am also aware of its restrictions ...

Esoterc SA-60 / Foobar2000 -> Mytek Stereo 192 DSD / Audio-GD NFB 28.38 -> MEG RL922K / AKG K500 / AKG K1000  / Audioquest Nighthawk / OPPO PM-2 / Sennheiser HD800 / Sennheiser Surrounder / Sony MA900 / STAX SR-303+SRM-323II

Link to comment

I suppose that I was "fooled" into believing it was a good tool to measure with when it actually has a lot of limitations. Limitations such as not representing a mixed DR track such as posted above or not showing if a high DR track actually sounds good or not showing if a low DR mix actually sounds bad.

 

Perhaps I thought it was more like a framing hammer but discovered it acts more like a 10 lb sledge hammer. Not useful for finishing nails at all…in fact quite misleading about the quality of the nail…but perfect when you need a simple tool for demo work.

 

As I do with my tool box, the more I learn about craftsmanship, the less I use simple tools and the more they stay at the bottom of the tool box.

 

Here is an analyzer that may be more what I'm looking for: http://www.computeraudiophile.com/f14-music-analysis-objective-and-subjective/easy-analysis-tool-trackalyzer-19674/#post303728 (thanks to jlohl for posting it).

 

Best,

John

Positive emotions enhance our musical experiences.

 

Synology DS213+ NAS -> Auralic Vega w/Linear Power Supply -> Auralic Vega DAC (Symposium Jr rollerball isolation) -> XLR -> Auralic Taurus Pre -> XLR -> Pass Labs XA-30.5 power amplifier (on 4" maple and 4 Stillpoints) -> Hawthorne Audio Reference K2 Speakers in MTM configuration (Symposium Jr HD rollerball isolation) and Hawthorne Audio Bass Augmentation Baffles (Symposium Jr rollerball isolation) -> Bi-amped w/ two Rythmic OB plate amps) -> Extensive Room Treatments (x2 SRL Acoustics Prime 37 diffusion plus key absorption and extensive bass trapping) and Pi Audio Uberbuss' for the front end and amplification

Link to comment

The TT-DR meter measures Peak-to-Loudness Ratio.

This is not a measure of dynamic range. The naming of this tool is very unfortunate.

 

When looking at the peaks and comparing that to the average loudness in your example, it would seem to be an accurate assessment.

 

What you want to be looking at is the dynamic range covered by the track (how loud and how quiet it gets) which is what the R128 Loudness Range (LRA) is a measurement of.

 

 

This is why I suggested that you look at and compare the results between the "DR" and "R128" measurements in JRiver.

JRiver's analyzer measures both, in addition to some other useful measurements such as Sample Peak Level, True Peak Level, R128 Volume Level (average Loudness) etc.

 

The "Trackalyzer" tool you have linked to also appears to use the LRA measurement.

 

LRA is useful to characterize dynamic range, but it is often not very useful for assessing whether one version of a track is better or "more dynamic" sounding than another.

In most cases, the TT-DR results are a far better indicator of this.

Link to comment
The TT-DR meter measures Peak-to-Loudness Ratio.

This is not a measure of dynamic range. The naming of this tool is very unfortunate.

 

When looking at the peaks and comparing that to the average loudness in your example, it would seem to be an accurate assessment.

 

What you want to be looking at is the dynamic range covered by the track (how loud and how quiet it gets) which is what the R128 Loudness Range (LRA) is a measurement of.

 

 

This is why I suggested that you look at and compare the results between the "DR" and "R128" measurements in JRiver.

JRiver's analyzer measures both, in addition to some other useful measurements such as Sample Peak Level, True Peak Level, R128 Volume Level (average Loudness) etc.

 

The "Trackalyzer" tool you have linked to also appears to use the LRA measurement.

 

LRA is useful to characterize dynamic range, but it is often not very useful for assessing whether one version of a track is better or "more dynamic" sounding than another.

In most cases, the TT-DR results are a far better indicator of this.

 

I am sorry I didn't reply to your earlier post…I got distracted. I will absolutely try the JRiver R128 tool…if that is more in line with what I'm looking for.

 

Thank you for "getting my attention" with another post. I truly appreciate it and the time you have taken.

 

Best,

John

Positive emotions enhance our musical experiences.

 

Synology DS213+ NAS -> Auralic Vega w/Linear Power Supply -> Auralic Vega DAC (Symposium Jr rollerball isolation) -> XLR -> Auralic Taurus Pre -> XLR -> Pass Labs XA-30.5 power amplifier (on 4" maple and 4 Stillpoints) -> Hawthorne Audio Reference K2 Speakers in MTM configuration (Symposium Jr HD rollerball isolation) and Hawthorne Audio Bass Augmentation Baffles (Symposium Jr rollerball isolation) -> Bi-amped w/ two Rythmic OB plate amps) -> Extensive Room Treatments (x2 SRL Acoustics Prime 37 diffusion plus key absorption and extensive bass trapping) and Pi Audio Uberbuss' for the front end and amplification

Link to comment

Great example Inferno. IMO if you're going to have one number for music than the calculated value should use all of the track data in the calculation, not game it for the 20 loudest passages to peak ratio.

 

Even the title has DR in it which it is not. Does it use R128 even? Is the code public? No, the code is proprietary and a black box other than the description the company posted in the help section of the offline app. It's certainly better than nothing, and it is an AWESOME concept, but I just feel that people rely in this single number the holy gospel in this which I feel in many cases is misleading. I may be wrong on this and I look forward to comparing to other tools and running it on more tracks so I can evaluate further.

 

It's a great discussion to have and I really look forward to looking at various analysis tools and how they compare to TT-DR. Thanks for all the info guys!

A Digital Audio Converter connected to my Home Computer taking me into the Future

Link to comment

Inferno thanks for starting this topic. I have the TT DR Meter Tool and have a question. The note on the face of the app says that it is good for 16/44 offline. Does that mean that a 24/96 file that I load in is not accurately processed and if so, what is it doing incorrectly? Other free apps to consider that perform the same task? Thanks again Inferno, I am learning a lot.

Link to comment

There is a plugin version of the TT-DR meter for Foobar 2000, which also enables writing the DR ratings (track and album) to custom tags. This version allows processing of all samplerates.

Esoterc SA-60 / Foobar2000 -> Mytek Stereo 192 DSD / Audio-GD NFB 28.38 -> MEG RL922K / AKG K500 / AKG K1000  / Audioquest Nighthawk / OPPO PM-2 / Sennheiser HD800 / Sennheiser Surrounder / Sony MA900 / STAX SR-303+SRM-323II

Link to comment
Sorry that I was not more specific, I am on Mac OS Mavericks.

 

oouhhps, you're screwed ... :-)

 

Maybe others might chime in, as i don't know of any Apple software for this task.

Esoterc SA-60 / Foobar2000 -> Mytek Stereo 192 DSD / Audio-GD NFB 28.38 -> MEG RL922K / AKG K500 / AKG K1000  / Audioquest Nighthawk / OPPO PM-2 / Sennheiser HD800 / Sennheiser Surrounder / Sony MA900 / STAX SR-303+SRM-323II

Link to comment
oouhhps, you're screwed ... :-)

 

Maybe others might chime in, as i don't know of any Apple software for this task.

 

There are certainly some packages out there like Adobe that provide a suite of tools, however the cost does not seem practical for the casual hobbyist/audiophile. I certainly don't have any problem with paying a software fee as I have with Audirvana. Thanks for the response.

Link to comment

This was posted here on CA before for Mac.

 

http://www.dynamicrange.de/sites/default/files/DR-Offline-Meter-for-MAC.zip

 

Can't say how it works as I don't use a Mac myself.

And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. 

Link to comment

Cross posting an excellent article for reference only…thanks Audiventory for posting this!

 

Positive emotions enhance our musical experiences.

 

Synology DS213+ NAS -> Auralic Vega w/Linear Power Supply -> Auralic Vega DAC (Symposium Jr rollerball isolation) -> XLR -> Auralic Taurus Pre -> XLR -> Pass Labs XA-30.5 power amplifier (on 4" maple and 4 Stillpoints) -> Hawthorne Audio Reference K2 Speakers in MTM configuration (Symposium Jr HD rollerball isolation) and Hawthorne Audio Bass Augmentation Baffles (Symposium Jr rollerball isolation) -> Bi-amped w/ two Rythmic OB plate amps) -> Extensive Room Treatments (x2 SRL Acoustics Prime 37 diffusion plus key absorption and extensive bass trapping) and Pi Audio Uberbuss' for the front end and amplification

Link to comment
  • 2 months later...
oouhhps, you're screwed ... :-)

 

Maybe others might chime in, as i don't know of any Apple software for this task.

I've used it offline for other sample rates and it seems to work fine. I get comparable/same readings to others I see posted online. I've just converted to use of the JRiver tools and as mentioned above the DR value seems the same as the TT tool.

 

As as for the "meaning" and use of the tool for me the primary thing has been to compare different masterings. When I first started using the tool I would compare the two masterings that I might have typically comparing a remaster CD with the original release or an audiophile remastered version with either the original or another release. Typically, but with exceptions, I have preferred the versions with a higher DR rating. I'm at the point that I have made most all of the remaster versus original changes I wanted and now it's only in the case of an audiophile release that I add do I pay attention to the values since JRiver implements them automatically for me.

"If you fly a flag of hate you are no kin to me"

Ry Cooder

Link to comment
  • 3 weeks later...
After Silence was recorded with the musicians playing together in the same room, without headphones.The reason being that in our opinion that creates a number of musical and technical benefits....... The musicians interact much more as they would do in a concert situation...and as they are not ''separated'' by headphones the musicians are forced to create a musical balance...the need for compression to control levels is no longer necessary...we can use a minimalist microphone setup and there by reduce phase problems...since everybody is in the same room, the boxed sound which is so common in many modern recordings is absent...the sound of the room helps ''glue'' the sound of the recording. That sounds like an easy solution but bear in mind that in order for this to work:e studio has to have a good sound.....the musicians have to be very good and well prepared as it is very difficult to repair mistakes because of the ''cross talk'' between the instruments....we have to be very precise when choosing and placingthe microphones...and the puzzle of placing the musicians at the right distance to the main stereo microphone pair and at the right distance to each other is very time consuming.

from the liner notes of the After Silence album.

Is there a DR reading on that album?

I would find that interesting since this album according to the above quote has a natural musician controlled ''compression''.

After Silence (FLAC)

or After Silence (WAV)

 

Andre%20300%20met%20schaduw%20320_300.png

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×
×
  • Create New...