Request for DR metric please

I am just enjoying the volume leveling that is the result of this measurement. Do we need more?

Ludwig, they havenā€™t, thatā€™s the whole point. They implemented R128, which is very different from DR :slight_smile:

I thought it was standing for ā€œdynamic rangeā€ which shows just how much I care.

It does stand for Dynamic Range, but the two metrics, ā€œDRā€ and ā€œR128ā€ use different algorithms, and the output of one is not compatible with the other. In my experience, DR is far more widely used, and therefore in one sense more useful. R128 is more relevant to volume levelling.

It is not my intention to belittle anything (and I donā€™t think I have given any reason to think I do). I am just trying to understand why this metric should be available in Roon.

What I really want to do is grow this feature.

Right now, between all of your libraries, you have a larger combined DR database than dr.loudness-wars.info ā€“ maybe we should make that info shared so you can see if there are better versions of the same album/trackā€¦ hrmā€¦

2 Likes

I use it to help me figure what version of a CD I have. All my 12,000 discs are ripped and have been put away in storage. I can take the DR number I generate using foobar2000 and look it up in the DR database and it gives me an idea of what release I have.

Once I know there may be a better sounding version according the people at the hoffman forums I can then start trying to track it down for albums I really care about.

For me is, and donā€™t get me wrong I think for the value levelling point of you this is a great advantage, that the values are per practically meaningless unless you are comparing two versions of the same album as per this post

If it is an easy thing to use what we more conventionally understand as DR VALUES well and good if itā€™s not, no harm done and still enjoy Roon as I always have done.

.sjb

That would require all of us to making sure we have the correct CD pressing and date data otherwise all we know is there is another version of an album with better or worst DR values.

Is it the original 86 disc, the 91 remaster or the 2015 remaster, the SHM version, the MFSL version etc etc

Indeed, excellent ambition but I agree it would be impossible to ensure people were assigning their DR rating to the correct pressing/release. Great to hear @danny and co are interested in this in some form though!

A design of a product that makes accurate mapping irrelevant would also be helpfulā€¦ for example, if N users had M discs each with a varying rangeā€¦ only the best of M would need to be identified accurately, and that might be doable manually upon request.

Building crowd-sourced products requires thinking about data in non-accurate-non-concrete terms.

Does that also mean that all our metrics in 1.3 may be used later as a crowd-sourced measurement?
As I expect currently nothing is send back yet to a big Roon database on your side, right?

we could theoretically allow you send edits back to us

correct.

I allow fully!

Please add a switch in one of the next releases so we can start supplying you that data :sunglasses:

Btw, I saw you mentioned something like this before in another thread:

I donā€™t want to get into too much detail about how weā€™re looking to improve Radio after 1.3, since the plans are fluid and could still change. I can tell you we think an important source will be what we capture about how people are already using Radio. Weā€™re now at a point where we have a critical mass of users generating this data, and 1.3 will contain client-side functionality to help us gather that data before itā€™s analyzed in the cloud.

1 Like

This is getting interestingā€¦ :grin:

I agree with you that a DR metric is most useful before buying an album. That said, I found it useful to analyze my existing albums. Sometimes Iā€™ll listen to a track and it sounds offā€¦something wrong. The DR metric helps me understand itā€™s a compression issue with the track and not something wrong with my setup.

Great idea. DR is so useful for comparing different versions of the same album. I have 4 versions of Off The Wall but 2 have a shockingly compromised DR that I donā€™t bother with them. I even use the DR database to decide which version of an album I want to buy.

Itā€™s not everything though of course. Most, if not all, versions of Blue by Joni Mitchell have a low DR (even the vinyl) but that will never stop it being a classic until the end of time lol. So I wouldnā€™t use it to determine whether to get a particular album - moreso which version of that album.

1 Like

The DR database gives Thriller the following values. Jap CD 1st pressing = 15 and 2008 CD remaster = 8

Roon is giving both a score of 5.

Itā€™s quite possible the DR score made me give biased preference to the 1st pressing and made me hate the 2008 remaster. Iā€™m very open to that idea.

Or maybe Roonā€™s R128 calc is off. Or both are wrong and we should actually listen and not be swayed by values lol.

Or we should have both values shown in Roon. Iā€™m confused but this is fun and interesting lol

Choose a Track and look at the Waveformā€¦and repeat the same for the 2008 remasterā€¦take screenshots and post them here if you like

IMHO, itā€™s best to look at BOTH the Waveform and the R128 value to get an informed opinionā€¦the Waveform will help give a guide to perceived Loudness, while the R128 value will guide towards the Dynamic Range within that relative Loudness

And lastly, the Crest DR should be called a ā€œGuideā€ rather than what you describe as a ā€œScoreā€, which sounds more like an ā€œinarguable resultā€, which IMHO it is not in all cases

Hi mate, no debates there. I did say in a post above a Dynamic Range value is not everything and gave an example.

Regarding the waveform, this is the first time Iā€™ve seen this in Roon (I had never turned on Audio Analysis until 1.3).

Is there a way to get the waveform in a larger view or is just at the bottom of the screen?

Cheers