You are here

Very confused about LUFS and RMS; can someone please clarify

For everything after the recording stage: hardware/software and how you use it.

Re: Very confused about LUFS and RMS; can someone please clarify

Postby Dave B » Fri Feb 09, 2018 4:54 pm

I wasn't criticising Hugh, just curious. Traditionally, I come from a prog background so overall dynamic changes can be a given and I'm curious to see how intelligent the calculations are. Sounds like they are at least trying to take things into account which is great.

Interesting thread this. And well timed. I've just been fiddling with some live recording mixes and was curious about levels and loudness, so I'm just re-doing the mixes with the final level at -12 lufs-i. As I have to wait for the tracks to play to get the values, it's nice to read and learn...

;)
User avatar
Dave B
Jedi Poster
Posts: 5506
Joined: Thu Apr 03, 2003 12:00 am
Location: Maidenhead
Veni, Vidi, Aesculi (I came, I saw, I conkered)

Re: Very confused about LUFS and RMS; can someone please clarify

Postby Hugh Robjohns » Fri Feb 09, 2018 6:10 pm

Dave B wrote:I wasn't criticising Hugh, just curious.

Yes, I appreciate that Dave, but you're not the first to wonder about it and it is a potential weakness in the current system with some specific types of material.

As I have to wait for the tracks to play to get the values, it's nice to read and learn...

There are some loudness metering apps that will analyse a file and report back in moments, rather than playing through in real-time if that helps... although It's often no bad thing to be forced to listen to the whole track ;-)

H
User avatar
Hugh Robjohns
Moderator
Posts: 27644
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Very confused about LUFS and RMS; can someone please clarify

Postby CS70 » Fri Feb 09, 2018 6:20 pm

jmcmmusic wrote:So I've just gone back to mastering one of my tracks.

I've got an integrated LUFS of -14.9. The WAV is bouncing out at -7.9 dB, so it's a quiet file. RMS is around -14.

But the waveform is flat; there's not much dynamism there. But it is a very energetic song all the way through, so I'm not overly surprised. I think Pro L2 flattened it out even more, so I think I need to consider other options, perhaps not using a limiter, or reducing the signal going into it so that the song's dynamics are preserved by the limiter.

So am I on the right track now?

Interestingly, I also did a quick free master on Landr for comparison purposes, and that's a much bigger waveform and a lot louder, peaking a 0dB.

If you are peaking at -8dBFS, the last thing you need is a limiter :)

Keep in mind that dynamic range (and interest) is much more in the arrangement, song and musical expression than in the technical mixing/mastering. A bass line is punchy because there’s little else around it. A vocal stands out because there’s space around it. A chours is exciting because is a sudden surge in instrumentation, complexity and (just a little) level - but you can’t surge if you’re already 100%..

Landr masters by peak limiting, so no surprises there.
User avatar
CS70
Jedi Poster
Posts: 5575
Joined: Mon Nov 26, 2012 1:00 am
Location: Oslo, Norway
Silver Spoon - Check out our latest video and the FB page

Re: Very confused about LUFS and RMS; can someone please clarify

Postby jmcmmusic » Sat Feb 10, 2018 3:32 am

"If you are peaking at -8dBFS, the last thing you need is a limiter "

Yeah, I had to turn everything down to get it within the -14 LUFS limit, as measured by Levels by Mastering The Mix, which luckily I remembered I had.

So the peak is at 0dB now, but the troughs are much lower.

And yes I understand about musical interest, building the energy etc, and that's all part of the current arrangement.

But it's the technical levels I'm concerned about getting right now, and understanding whats required level-wise for distribution.
jmcmmusic
New here
Posts: 7
Joined: Fri Feb 09, 2018 12:14 pm

Re: Very confused about LUFS and RMS; can someone please clarify

Postby jmcmmusic » Sat Feb 10, 2018 6:27 am

And I think I found the tool I need to help me out here : EXPOSE from Mastering The Mix. It will at least give me some degree of confidence when I am submitting the album tracks for streaming.

According to LEVELS, the track I'm currently mastering is ready for release in terms of levels and dynamic content. I'd attach a screen shot but it doesn't seem to be possible.

Many thanks to you all for demystifying a very confusing topic! :) :thumbup:
jmcmmusic
New here
Posts: 7
Joined: Fri Feb 09, 2018 12:14 pm

Re: Very confused about LUFS and RMS; can someone please clarify

Postby Hugh Robjohns » Sat Feb 10, 2018 11:20 am

jmcmmusic wrote:So the peak is at 0dB now...

This is a mistake, and will result either in the track getting turned down further, or being peak limited by the streaming service.

You need to use a meter that measures the TRUE PEAK level (dBTP), and then adjust the track (or set an oversampling limiter) so that the true peaks are definitely no higher than -1dBTP, and ideally no more than -3dBTP -- the latter level is best if the streamer is using a lossy data reduced audio format like MP3.

H
User avatar
Hugh Robjohns
Moderator
Posts: 27644
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Very confused about LUFS and RMS; can someone please clarify

Postby jmcmmusic » Mon Mar 19, 2018 10:44 am

Sorry for resurrecting this topic, I've been ill for the last month and am only getting back to this now.

Hugh, I take your point. I'm now using Fabfilter Pro L 2 as a true peak limiter with the limit set to -3Dbtp as you said, and the song is coming out at -14 LUFS. So that should all be good. The output is still nicely dynamic and there's not much limiting being triggered.

So I ran a simple experiment using Spotify, Google Play Music and my mobile phone speaker,

If I play e.g. Calvin Harris - "Faith", or Sia - "Move Your Body" on Spotify, they are both very loud, or apparently so. But Spotify say they should be no louder than my song.

Then I play my song through the Google Play Music app, and it's reasonably loud, but not as loud or apparently loud as the Spotify tracks.

So are the Spotify tracks really at -14 LUFS, or is there something else I need to do to increase the apparent loudness of my track?

I'm really keen to get this right, I don't want to submit a 10 track album in June only to find out later than it doesn't match commercially acceptable loudness standards, or what the listener expects from a commercial track.

Again apologies for raising this again, but it really is a vital issue and I have to be 100% certain that I've got it right before I distribute the album online.
jmcmmusic
New here
Posts: 7
Joined: Fri Feb 09, 2018 12:14 pm

Re: Very confused about LUFS and RMS; can someone please clarify

Postby permanent_daylight » Mon Jun 01, 2020 2:07 pm

Dumb questions but:

-when using this system you'd ideally end up using less 'dynamics' potentially? If we are only concerned about loud sections? or does this just mean 'set the quiet sections wherever you want'. I could see it as meaning 'avoid loud sections standing out'

-if every song is to a similar LUFS standard, then that's not dynamic. So since naturally some songs might fall a lot lower than others, for example, in an album i'm doing with a very 'metal' volume track, is next to a piano solo. ...then maybe the best place is that they lie either side.

of course, I am talking about actual dynamics, the musical piano to forte. Not what's refered to here which is about transients being squashed or something akin to a 'peak-crest' i.e. in the micro, not macro scale.
permanent_daylight
Regular
Posts: 143
Joined: Sun Jun 25, 2006 12:00 am

Re: Very confused about LUFS and RMS; can someone please clarify

Postby John Willett » Mon Jun 01, 2020 2:44 pm

With LUFS you don't have to compress so much - and heavily compressed music can actually sound worse (thank goodness).

I once heard a demo by the AES's standsrds man (Mark Yonge) who had a quiet classical piece followed by loud rock music - both at -23dBLUFS - it worked beautifully. :thumbup:
User avatar
John Willett
Jedi Poster
Posts: 7176
Joined: Tue Mar 07, 2000 1:00 am
Location: Oxfordshire UK
John
Sound-Link ProAudio
Circle Sound Services
Sound-Link are UK Distributors for: Microtech Gefell, ME-Geithain, AETA, HUM, Håkan, Meyer Turtle

Re: Very confused about LUFS and RMS; can someone please clarify

Postby Hugh Robjohns » Mon Jun 01, 2020 3:07 pm

permanent_daylight wrote:-when using this system you'd ideally end up using less 'dynamics' potentially?

No. One of the desirable side-effects of loudness normalisation is that it allows and actually encourages the use of greater dynamics.

...If we are only concerned about loud sections?

The Integrated Loudness value is calculated across the entire duration of a programme or music track etc. The level of the loudest elements contributes to the Integrated figure, but so too do the (slightly) quieter bits.

... or does this just mean 'set the quiet sections wherever you want'. I could see it as meaning 'avoid loud sections standing out'

You mix the track so that it sounds good with appropriate dynamics. You can then measure the Integrated loudness value and adjust the overall level if necessary to comply with the required Loudness Target level.

With practice, and assuming you have optimised your loudspeaker monitoring reference SPL, it's surprisingly easy to mix and hit the right target straight off.

-if every song is to a similar LUFS standard, then that's not dynamic.

Dynamic changes happen within a track. The Integrated LUFS figure is measured across the whole track.

So since naturally some songs might fall a lot lower than others, for example, in an album....

Yes, if you measure and calibrate each track individually to a set target loudness that would mess up any intentional dynamics between tracks across an album collection. So in that case it is the whole album that should be considered the 'programme' for the Integrated Loudness figure. This is an issue that affects classical music albums quite significantly, too.

The problem, which I don't think anyone has really solved yet, is what happens when someone takes an album track and plays it individually amongst other material in a playlist? Should it retain the loudness level required when sat amongst it's partnering album tracks, or should it have a standard loudness to compare with everyone else's material in a random playlist?

So some compromise is involved, depending on whether you think the tracks will be played mostly within an entire album context, or mostly as independent tracks selected for streaming in a random playlist.
User avatar
Hugh Robjohns
Moderator
Posts: 27644
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Very confused about LUFS and RMS; can someone please clarify

Postby CS70 » Tue Jun 02, 2020 2:08 pm

jmcmmusic wrote:So are the Spotify tracks really at -14 LUFS, or is there something else I need to do to increase the apparent loudness of my track?

Didn't see then back in March, but for what's worth...

I think often this LUFS thing is taken a bit the wrong side.

The general idea is that, if the loudness of track is higher then a certain amount, it will be "turned down". Turned down means as in turning down the volume knob. The process is called loudness normalization - in the sense that it is the kind of uniform, linear amplitude reduction that you get when you move the volume knob counter-clockwise.

This is in opposition to compression/limiting: there's no attack, hold and release that depend on a threshold or anything (and would change the sound of your mix): the sound of your mix is absolutely unchanged. If say you exceed the target LUFS of 5 dB full scale, the normalizer will turn everything down 5 dB... but the listener can easily defeat this by turning up his volume knob of 5dB and the mix will sound as loud as it sounded on the mastering monitors. In rough terms, the control of the playback level is back to the user, as opposite to the mastering engineer.

Also, this is in opposition to what happened before: without LN, the level of the master is exactly the level that reaches the user (give or take radio processing etc which is the same for all songs so doesn't matter) . So before LN it was important to make a master as loud as possible, while with LN, all masters will be brought down to the same average loudness (so long they are loud enough, as usually they will not be brought up). Before LN, "as loud as possible" of course meant compression/limiting, which do change the sound of the mix, and so as a producer you had to decide how much to sacrifice the sound to get to a "competitive" loudness.

The result of all this is that now you really don't need to worry about loudness (well, you still have to worry to be "loud enough" but that's usually not a big issue): just make the best mix you can.

That means that the (mastered) mix must sound good at low volume level, sound good at medium level and sound good at high level.

If the LUFS are too high, fine - it will be turned down a little... to be at the same average level of all the other material. Not lower, not worse - just the same average.

But since you made sure it sounded good at any reasonable level, it will still sound good. So if it gets turned down, it is no big deal.

Obviously to make sure your mix sounds good at low levels, you generally end up mixing it at low, conversational level... which is a well known trick of the best mixing engineers out there.

The conclusion is that you don't really need to obsess about the LUFS at all.

Just make the best, most dynamic, most good-sounding-at-all-levels mix you can, and have it mastered in a way that preserves the goodness (aka no hard limiting, and the mastering engineer will raise the average level in case it's too low), and you're all set. Spotify here I come.

If the final LUFS are a little on the high side, it just means that you like to listen to music (and mix) at a little higher level than what the broadcasting community thinks is a reasonable average, but that's it.

If you turn down your volume knob you will get the same level that streaming listeners will hear - and since you have checked your mix at all levels - nothing untowards will happen: the streaming service will normalize it down, but most likely your final listener would have done the same with his volume knob... but your mix will still "pop" since you've made sure it's damn good also at lower volume.

Basically LN removes the need of excessive limiting at the mastering stage, so there's one less thing to worry about, not one more.

It's a bit like your studio customers couldn't see the badges on the microphones - a Neumann is still a Neumann and a Behringer is still a Behringer but may another mic is not a Neumann and sounds great anyways and the user just appreciates it for what it is.
User avatar
CS70
Jedi Poster
Posts: 5575
Joined: Mon Nov 26, 2012 1:00 am
Location: Oslo, Norway
Silver Spoon - Check out our latest video and the FB page

Re: Very confused about LUFS and RMS; can someone please clarify

Postby The Elf » Tue Jun 02, 2020 3:16 pm

:clap: Well said.

It seems that people feel that they are still wearing the shackles of peak normalisation, but it's been given a different name and target.

Loudness normalisation = CHOICE!

You can choose to create a very dynamic mix, or you can choose to slam a limiter, just as you always did. There are mixes that I will still slam, because the music sounds good that way. And there are mixes I can now give full dynamic range to, because it is suitable in that context, whereas before LN I would have to join the herd.
User avatar
The Elf
Jedi Poster
Posts: 14563
Joined: Tue Aug 14, 2001 12:00 am
Location: Sheffield, UK
An Eagle for an Emperor, A Kestrel for a Knave.

Re: Very confused about LUFS and RMS; can someone please clarify

Postby RichardT » Tue Jun 02, 2020 8:56 pm

Good post by CS70.

The only comment I’d make is that Spotify for example will change the sound of your track if it is at less than -14 LUFS and needs to be louder. In their own words:

Negative gain is applied to louder masters so the loudness level is at ca - 14 dB LUFS. This process only decreases the volume in comparison to the master; no additional distortion occurs. (...This is exactly what Cs70 was saying)

Positive gain is applied to softer masters so that the loudness level is at ca - 14 dB LUFS. A limiter is also applied, set to engage at -1 dB (sample values), with a 5 ms attack time and a 100 ms decay time. This will prevent any distortion or clipping from soft but dynamic tracks.
RichardT
Regular
Posts: 50
Joined: Fri Aug 13, 2004 12:00 am
Location: London UK
My last album sold over a million copies - if you don't believe me, come and look in my garage.

Re: Very confused about LUFS and RMS; can someone please clarify

Postby Hugh Robjohns » Tue Jun 02, 2020 9:05 pm

RichardT wrote:The only comment I’d make is that Spotify for example will change the sound of your track if it is at less than -14 LUFS and needs to be louder.

To be fair, for the sound to be changed noticeably the track's Integrated Loudness level would need to be well below -14LUFS AND it would need to have a lot of transient peaks hitting or very close to -1dBTP.

The processing being applied is simply a lift in level combined with some protective limiting with typically fast dynamic parameters, and it will be pretty much inaudible in the vast majority of real world situations.
User avatar
Hugh Robjohns
Moderator
Posts: 27644
Joined: Fri Jul 25, 2003 12:00 am
Location: Worcestershire, UK
Technical Editor, Sound On Sound

Re: Very confused about LUFS and RMS; can someone please clarify

Postby RichardT » Tue Jun 02, 2020 9:30 pm

Hugh Robjohns wrote:
RichardT wrote:The only comment I’d make is that Spotify for example will change the sound of your track if it is at less than -14 LUFS and needs to be louder.

To be fair, for the sound to be changed noticeably the track's Integrated Loudness level would need to be well below -14LUFS AND it would need to have a lot of transient peaks hitting or very close to -1dBTP.

The processing being applied is simply a lift in level combined with some protective limiting with typically fast dynamic parameters, and it will be pretty much inaudible in the vast majority of real world situations.

Yes, this is true.
RichardT
Regular
Posts: 50
Joined: Fri Aug 13, 2004 12:00 am
Location: London UK
My last album sold over a million copies - if you don't believe me, come and look in my garage.

PreviousNext