Dynamic range measurement is different than http://dr.loutness-war.info

I think the bottom line might be that LRA is not a measure of dynamic range, nor should it be used as a proxy for dynamic range.

As an aside, it turns out I have five versions of Brown Eyed Girl, four of which appear on CD ‘samplers’. They are all different.

The version with the highest LRA number turns out to be the mono radio edit; it is by far the poorest ‘sound’. The two channels are not identical, leading me to suspect a transcription from vinyl.

The version with the lowest LRA number is on the ‘Still On Top’ compilation, which I guess is ‘remastered’… and also has many inter-sample peaks above 0dBFS.

What a mess…

1 Like

Thanks, I didn’t know about the Orban loudness meter program. Running it on a smattering of my files reveals a pattern: most music I have that is not expressly mastered in “audiophile” ways has plenty of reconstructed samples over 0dBFS, even if its loudness measures are pretty good.

I suspect that Roon’s volume leveling hides these from my DAC in most cases.

I think that’s my position too. I’ve not bothered adding headroom management. Although I have noticed a couple of albums where volume levelling is increasing level, so I have some nagging doubt…

I sure like using R128 for volume leveling. It is brilliant for that. I wouldn’t want Roon without it!

But I sure wish Roon users had the option to use the DR Database way of measuring DR for displaying DR when comparing album versions.

I haven’t given up hope that we will see this option. They revamped the waveforms for us local file junkies, and comparison of versions utilize both waveforms and DR values frequently.

Please? :upside_down_face:

This is another one of those issues with Roon where they would be better off giving users the choice to select the scale this wish to use and are more familiar with. More options is better than being told to just use an option we did not ask for or want.

So, please add your voice to the feature request .

As always, it’s not that easy. They’re not measuring the same thing, so it’s not a matter of changing ‘scale’. Including a crest factor DR number would require reanalysing all your files; and is (probably) not available for streamed files.

To borrow the earlier analogy, it’s not like Centigrade vs Fahrenheit, more like (I’m struggling for a good analogy - sorry) mass vs density. Both properties of the same physical ‘thing’, and both useful to know, but different…

Thanks. I went ahead and made the feature request. I would encourage other to do so too.

1 Like

This is the EBU document describing LRA. This quote may be relevant…

“Loudness Range should not be confused with other measures like dynamic range or crest factor, etc.”

Perhaps Roon should not describe LRA as a DR measurement, or at least make it clearer that it’s not a reliable proxy.

Thought?

How much clearer would you like it to be? It’s got its own entry in the KB.

I know. I re-read it half an hour ago to be sure what it says.

To me, the KB conflates LRA and DR then, quietly, at the end, says it’s not really the same. Given the proportion of Roon users who never seem to find the KB, how many read to the end of the page? Is that good enough?

(I’ll get off my soapbox now… sorry… )

It seems absolutely crystal to me, but ymmv.

One should not have to read the knowledge base to discover that what Roon labels as the dynamic range number is indeed NOT a dynamic range number. It is labeled incorrectly and I believe it is done on purpose. Why? So Roon does not have to provide the “hobby standard” dynamic range number as that would require a license, more work, and may cause problem with Qobuz and Tidal.

1 Like

Apart from it consistently reading ‘dynamic range’ when the EBU use pretty much the same words and call it ‘loudness range’, and tell you it’s not the same thing at all (not ‘[not]directly comparable’), no problem at all!

Edit - bother - must burn my soapbox…

I just can’t get bent out of shape about it. I’ll leave you chaps to it.

Another thread muted.

@anon55914447

@AndyR is not “bent out of shape”. He, and I, think that Roon should not call it “Dynamic Range” when it is really “Loudness Range”. Even the EBU, who created the R128 standard, says is should not be conflated with “Dynamic Range”.

Burying the clarification in the knowledge base with a justification for using the EBU “Loudness Range” as “Dynamic Range” doesn’t really cut it.

1 Like

Roon is not the only program which label’s R128 as Dynamic Range so does JRiver, granted the type is specified in parentheses . So, Roon’s use of the term is not so far fetched or unique.

Well, I don’t consider anything in the knowledge Base to be buried. However, what the DR value means is clearly specified when looking at a Track’s File Info. See:

This topic has been discussed before in detail including responses by some devs, it is a valuable read, see:

This is an Argumentum ad Populum. Just because JRiver does it too, it must mean it is correct to do. That is a VERY weak and fallacious argument. Look, the EBU (the creators of R128) document that describes the “Loudness Range” measurement says this:

“Loudness Range should not be confused with other measures like dynamic range or crest factor, etc.”

In other words, it is not meant to be used as a dynamic range number because it is not a dynamic range number. Yet that is exactly what Roon does. Uses it as a dynamic range number. What JRiver does is irrelevant. Roon is misleading because it is representing the number as dynamic range when it is a number used for loudness control in broadcasting.

In that thread you said was a valuable read, you effectively said that the crest factor-based number DR uses provides you with meaningful information while the R128 number is misleading. What I read from Roon “devs” was justifications for using that number over the crest factor DR Database number. What you wrote was spot on and what the devs wrote is an effort to justify a poor choice.

3 Likes

One can argue the pro’s and cons of either method until the cows come home, but that doesn’t help if all I want to do, is compare an albums DR value with http://dr.loudness-war.info, which is not an unreasonable or even unusual request.

2 Likes

If it were labelled as Loudness Range (R128) in ‘file info’ that would be better, but I suspect people would still persist in misinterpreting it. I guess I hope correct naming might stop ‘Roon’s DR numbers are rubbish’ threads… but on reflection, probably not…

If what we really want is the numbers from the loudness war site, visiting the site gives us that and (usually) a bunch more information.

I’ve not looked, but maybe someone has already written a script or program to analyse local files and embed a DR tag. Now there’s a thought…