Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Dynamic Range, LUFS - meaning and measurement (Read 2895 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Dynamic Range, LUFS - meaning and measurement

Hi,
I've been using DR Meter for a long time and it was always doing it's job - meaning the more DR the audio file had, the better (more pleasant in the long run) it sounded.
Some time ago I found that everybody uses LUFS now because it is better (more accurate and takes human hearing "capabilities" into account). So I started to use it also.
Let's not get into technical details of DR and LUFS now. I do (more or less) understand the principles of those both meters and what they do.
Only recently I found that LUFS seems to be some kind of "lie" and does not tell anything (or tells much less than DR) about actual music file. That is what I've noticed:
- I used HDCD file in Foobar (44,1/16 file with HDCD decoded by HDCD plugin).
- When I measured that file's DR - it was let's say 10dB
- When I mesured that file's LUFS - it was let's say 12dB
- I changed HDCD plugin options to Halve the volume when HDCD is detected and measured the file once more
- It's DR was still 10 - which is not suprising, because plugins options do not change the actual music contained in the file
- But LUFS changed, it wasn't 12 now. It was 18!!!! That occured, because the player (plugin) lowered the output volume by 6dB.
But my file was exactly the same, nothing changed except lovering the volume.
So I assume LUFS does not measure the actual music or file. It measures "what we hear or feel" but not what we hear or feel "from the file" but from the output. So I can turn the volume knob of my amplifier down and voila....I have a better sounding file - that is not true. The file is just the same but I just listen to it quieter.
That leads me to assumpion that I can theortically produce a file with DR of 2dB (which would be horrible) and then lower it's level in DAW program and get brilliant measures of LUFS - basically any dB I would want - if I got it to 15dB I could lower it more to get 20dB and so on....but what would be the point? The actual music still would be horrible 2dB but very quiet so I would have to turn the volume knob up.
What is the purpose of LUFS then? I've read that people were hoping that Loudness Wars would seize with LUFS being incorporated by streaming services but it is not a problem to deliver very badly produced and compressed file with very good LUFS.
If the purpose of LUFS is just leveling the volume of all files in peoples playlists it does not change anything in the matter of Loudness Wars.
Am I missing something?

Re: Dynamic Range, LUFS - meaning and measurement

Reply #1 –
LUFS means Loudness Units relative to Full Scale. It's not about quality, it's about how loud the audio is.

Quote from: wojak on
If the purpose of LUFS is just leveling the volume of all files in peoples playlists
Yes.
Quote from: wojak on
it does not change anything in the matter of Loudness Wars.
Who said it will?

Re: Dynamic Range, LUFS - meaning and measurement

Reply #2 –
Quote from: lvqcl on
LUFS means Loudness Units relative to Full Scale. It's not about quality, it's about how loud the audio is.

Quote from: wojak on
If the purpose of LUFS is just leveling the volume of all files in peoples playlists
Yes.
Quote from: wojak on
it does not change anything in the matter of Loudness Wars.
Who said it will?

So if I do not use streaming services at all LUFS should have absolutelly no meaning to me? I listen to files ripped from CDs and I listem to full albums (not random compilations). And if I want to listen louder or quieter I just turn the volume on amplifier.

I've read many articles about predicted end of L.Wars due to LUFS:
https://hyperbits.com/ultimate-guide-to-the-loudness-war/  (just one of them).

Re: Dynamic Range, LUFS - meaning and measurement

Reply #3 –
LUFS metering is just a tool, an indication.
What's affecting the "loudness war" more directly is the fact that streaming services are equalizing the loudness for the listeners. This means that heavily compressed music doesn't have the same advantage it would have, say, on CDs played by clueless listeners, who think louder=better.
But in practice, like 95+% of the time, a lower LUFS value does mean that the music has more dynamic range, because the peaks are still probably close to 0dBFS.

Edit: And there's also Loudness Range (LRA) which is part of the same EBU R128 spec and is more similar to the DR measurement, I suppose.

Re: Dynamic Range, LUFS - meaning and measurement

Reply #4 –
Quote
.but what would be the point?

What is the purpose of LUFS then?
It's exactly that...  A digital loudness measurement.  ;)  It's not really intended for listeners.  It's for content producers, broadcasters, and streamers, etc.  But listeners can use it for loudness matching (ReplayGain, etc.).

Quote
That leads me to assumpion that I can theortically produce a file with DR of 2dB (which would be horrible) and then lower it's level in DAW program and get brilliant measures of LUFS - basically any dB I would want - if I got it to 15dB I could lower it more to get 20dB and so on....
Right, but...   If you flip that around, starting with a highly-dynamic recording and boost the volume you can only go so-far without clipping.   If you want to "win" the loudness war you have to use dynamic compression and limiting (and maybe push the levels into clipping).   That's why some people use LUFS as a proxy for dynamic range. 

Quote
I've been using DR Meter for a long time
There is no perfect-way of measuring dynamic range (or "dynamic contrast") but "DR" (crest factor) has a flaw...   We don't hear very-short peaks as "loud" and all it takes is one short peak to give a better reading.      Or if you compress* to MP3 or cut & play a vinyl record the wave shape changes making some peaks higher and some lower (without affecting the dynamics of the sound).   That makes a "better" crest factor measurement.

LRA compares the loudness of the loud & quiet parts so I feel it's a better measure, but the important thing is how it sounds and what the artist & producer intended.  More dynamic range isn't always better.  And there are short-term & long-term dynamics, and "it's complicated".   I have some songs where the drums & cymbals really stand-out so to me they sound very dynamic (and good!), but otherwise the volume is consistent throughout the whole song (no long-term dynamics).


* Of course that's file compression, not dynamic compression.  ;)