LUFS - confused.

[DJ Warmonger]

Then run the CD release of Metallica's "Day that never comes" through it

But Metallica doesn't release club music as far as I know. Surely, if you hard compress a piece of music which was originally not meant to be compressed, it will sound like crap.

Labels like Warner and Universal Music also say "what's working and what's not" regarding releases. (...) clubs are completely ignored in this case

And vice versa

If you're working purely vinyl based

Wait, what is vinyl again? We are living in 2017 now

Sorry, but you seem to live in a different reality than I'm used to.

But if this is your thing - so be it. You just have to realize and accept, that there are people out there, fighting against this insane loudness that only damages productions.

I'm not fighthing against anytthing, I only state the facts. You can have your own opinion, but you can't have your own facts. You describe some perfect world which just doesn't exist.

What is "FOH" you're talking about, anyway?

[Googly Smythe]

Vinyl is on the up. There is even a thread somewhere here about cutting machines.

I think this thread demonstrates that the loudness wars are far from over.

[annode]

As far as I'm concerned this loudness scheme is only viable when the file is read completely before a loudness measurement is available.

Measuring/Calculating the ILk value takes no longer than 1-2s, which can then be written into a database for streaming services to easily read out (as META Data for example). So the handling of this these days (realtime) is pretty much painless

I updated my misconception just below that as an EDIT. I guess you didn't read my whole reply before writing.

[Compyfox]

At this point, I consider you trolling this thread.

But Metallica doesn't release club music as far as I know. Surely, if you hard compress a piece of music which was originally not meant to be compressed, it will sound like crap.

Same goes for electronic music, or what you call "club music".

Labels like Warner and Universal Music also say "what's working and what's not" regarding releases. (...) clubs are completely ignored in this case

And vice versa

As somebody that worked in the event technology field, this is nonsense.

"Clubs" these days use huge PA's that need to be limited in signal strength (sometimes they aren't even allowed to exceed a certain sound pressure level!). So already heavily compressed material gets compressed even further to sh*t.

If you enjoy that, that's your prerogative.

If you're working purely vinyl based

Wait, what is vinyl again? We are living in 2017 now

Sorry, but you seem to live in a different reality than I'm used to.

Trolling at it's best. Not biting, sorry.

I'm not fighthing against anytthing, I only state the facts. You can have your own opinion, but you can't have your own facts. You describe some perfect world which just doesn't exist.

This post alone tells me that you know jack about this topic at all, and want to shape "your" opinion on to others. See comment (and I paraphrase): "it is done like that".

You have no grasp on technological limitations (MP3/AAC has a loudness "limit", tech at "clubs" has a compressor/limiter array before hitting the PA). You don't seem to understand that so called "DJ Mixes" also need to be loudness normalized (volume pulled down) before you can properly mix. And you ignore that "loudness normalization" or proper loudness handling is an important topic these days, and not limited to "streaming only" (especially with the return of compact cassette tape, vinyl, the release of Blu-Ray HD Audio, etc).

Your "I master to -5,5LUFS, because commercial Trance sits there" (which it doesn't, btw) is the reason why the Loudness war keeps on continuing.

I also only state "the facts", and not "an opinion" - I've been too long in this metering game to know pretty much every angle of the discussion. Maybe educate yourself in this area first, please.

What is "FOH" you're talking about, anyway?

Front of House. Sometimes there is a tech sitting right in the middle, controlling what the "house hears" - sometimes he sits behind a glass window at the other side of the hall (theater). In other words... the person that controls the sound of the location (not the DJ!).

I updated my misconception just below that as an EDIT. I guess you didn't read my whole reply before writing.

I read it, but from my understanding, you thought of this be done in "realtime". "Offline" is on a whole different ballpark. And that is a task of about 1-2 seconds for a 5min production (at least if I look at my Wavelab Batch Analysis process).

[annode]

I updated my misconception just below that as an EDIT. I guess you didn't read my whole reply before writing.

I read it, but from my understanding, you thought of this be done in "realtime". "Offline" is on a whole different ballpark. And that is a task of about 1-2 seconds for a 5min production (at least if I look at my Wavelab Batch Analysis process).

Yes, this makes sense to me. For the most accurate measurement of material, reading the full file is best. It seems they found that using a gating process block in 'real-time' metering gets close enough to what people actually perceive as loudness.(as described here: https://www.itu.int/dms_pubrec/itu-r/re ... !PDF-E.pdf
This process is beyond me, but it seems to only need to look ahead 400ms to set a gate threshold where sound below that is ignored. And then it sets some point 10db above that for some reason I don't quite understand...and I don't really need to understand.

My own definition of loudness is: "How many,how much,for how long."

[Compyfox]

The gating is mainly there to "limit the measurement range" so to speak, to get a more objective readout of the "subjective/perceived loudness".

Too low signal is being ignored/discarded (< -70LKFS), while "sudden drops of loudness" is also ignored/discarded (> -10dB).


Source: https://tech.ebu.ch/news/itu-publishes- ... eb-15apr11

The relative gate makes sure that quiet periods, below a given threshold, are left out from the Loudness measurement, because human perception of how loud a programme is chiefly determined by the loudest passages - the so-called 'foreground sound'.

My own definition of loudness is: "How many,how much,for how long."

Excellent question.

Though then we still have to differentiate:

• "perceived loudness" as in "program stream/material loudness" (measured ITB, tells us if the production isn't "too hot" for the release medium, whether or not it has lost it's "dynamic", or show if it has sudden loudness jumps where you have to manually adjust your volume settings)
• or Sound Pressure Level (measured post speaker/headphones, tells us how "loud" the signal source is we currently listen to).

Two completely different things (as in two different realms) - but equally important in terms of ear fatigue and health.

[annode]

The "How many,how much,for how long" is less of a question and more of an equation as I meant it.
It applies to loudness whether it is SPL or perceived judgement, if you think about it. All that is missing is "gain".

The gating block still needs a relative comparator to do it's job. Same thing with making our own music post-production loudness decisions. I'm finding that the only way to really get it right without lots of trial and error is by using comparisons to 'like' material/genre.

The studies in that PDF used a comparator signal so the listeners could switch between the signal and the testing material, adjusting the material level to match. So I could master a song solely based on some average, genre specific LUFS level for some types of music I guess, but I'd think collecting some music that is very close to what I myself make, and matching with that level, is the way to go.

Unless my ears are fairly fresh, I won't really find the best loudness level on the fly. I can if I compare it to another target level track. Some music types of are more forgiving then others of course.
For example, this track of mine turned out too loud. Some might say it sounds fine, but I have a quieter mastering(as well as the original mix) and compared to that, this version is clearly missing the vibrancy and sounds fatiguing.(dl the .wav file to really hear it...if you care too. )
https://soundcloud.com/annode-1/annode-condensation

[Compyfox]

The thing with reference tracks however... they have a specific loudness to begin with. So in order to master to this reference, the reference needs to be pulled down to the loudness you want to distribute.

Then, an objective A/B is possible. However, your track will be more "dynamic" (due to still existing transients) compared to the pulled-down reference (e.g. -8LUFS/-0.5dBTP down to -16LUFS/-6,5dBTP)

[annode]

...I'd think collecting some music that is very close to what I myself make, and matching with that level, is the way to go.

"that level" = the loudness level from a reference track after analysis or metering then matching my track to that.
A/Bing was just meant on the fly comparison.

You lost me with the point you were making.
BTW, this just appeared in my Youtube...seems relevant to this topic.
https://www.youtube.com/watch?v=mDZcz-V29_M

[Compyfox]

...I'd think collecting some music that is very close to what I myself make, and matching with that level, is the way to go.

"that level" = the loudness level from a reference track after analysis or metering then matching my track to that.
A/Bing was just meant on the fly comparison.

You lost me with the point you were making.

My point was:

- if you use a track that you like, and use this as reference for both the sound and the loudness, then you'll never beat the loudness war
- if you use the track as "sound reference" however, but pull the loudness down to the value you want to release in, then A/Bing is easier

BTW, this just appeared in my Youtube...seems relevant to this topic.
https://www.youtube.com/watch?v=mDZcz-V29_M

Watched this documentary, was sadly cut and created in such a way, that it completely ignores the topic "Loudness War" in an understandable way. They brag on about MP3 releases and that things "changed". But they don't offer a possible solution.


It's like telling you "this cake makes you addicted", you know that it makes you addicted, but there is no possible "way out" given.

[annode]

Yeah, I was pretty disappointed with that vid as well. I posted it before I finished watching it...sorry.
Not too old...but still stupefied about media compression...odd.
Compression is what we do to our post-productions to allow more gain. This is how things get loud.(as we all know)
If I use a Steve Roach track to set my track I'm personally not involved with any loudness war. My track will sound more evenly balanced and at a more proper gain level for the genre. If I just used a mastering program preset I might just be wasting my time. But for argument sake, for my next work, I plan to get the right loudness balance by using shelving EQ and upwards compression...and if needed, a brickwall/true peak limiter for fast transients.

The track I posted above shows how a quiet track can sound dead when using a mastering program to bring it up to a level common for this music. I didn't do it right and you can hear it.(in the dl .wav not the SC player.)
In a post-production perspective here, it's too loud. But it's not because of the gain, it's in result of it.

It's all quite complicated because music is so varied. This LUFS concept is pretty simple in itself.

[Googly Smythe]

I find the LUFS "principle" easy to understand and use.
I think a lot of the confusion comes from (as in the op's case) reading about theoretical guidelines and then expecting to find them adhered to in the real world - and then discovering variances.
I believe that European broadcasters (ie tv and radio) have to abide by the rules laid down by the EBU (the -23dBs thing) but everyone else can do what the hell they want, for good or ill, and regardless of what iTunes, YouTube, Spotify, etc will do to their productions.
I also think the "loudness war" is being won (not by the loudness freaks!) but is a long, long way from being over.

I also, also, think that the biggest casualties of the war are older tracks that have been "remastered". So many have been ruined by slap-dash limiting - well, 'nuff said.

[Compyfox]

I find the LUFS "principle" easy to understand and use.
I think a lot of the confusion comes from (as in the op's case) reading about theoretical guidelines and then expecting to find them adhered to in the real world - and then discovering variances.
I believe that European broadcasters (ie tv and radio) have to abide by the rules laid down by the EBU (the -23dBs thing) but everyone else can do what the hell they want, for good or ill, and regardless of what iTunes, YouTube, Spotify, etc will do to their productions.

This is why I talked about this years ago already.

These days, the following two "regions" are considered important:

Broadcast:
-23LUFS (Europe), -24LUFS (US and France)

"Regular" distribution:
-16LUFS to -14LUFS
(the Audio Engineers lean towards -16LUFS, the consumers more towards -12LUFS, so -14LUFS is currently a good middle ground).

I also think the "loudness war" is being won (not by the loudness freaks!) but is a long, long way from being over.

The war will never be over, only change the version number (currently in v3.0 IIRC from AES#142)

I also, also, think that the biggest casualties of the war are older tracks that have been "remastered". So many have been ruined by slap-dash limiting - well, 'nuff said.

Can't agree with you any more on that.

[DJ Warmonger]

"Clubs" these days use huge PA's that need to be limited in signal strength (sometimes they aren't even allowed to exceed a certain sound pressure level!). So already heavily compressed material gets compressed even further to sh*t.

Sound system is limited by max amplitude (thus cone travel distance) and not rms power as far as I know. Thus, to achieve greater loudness, tracks need to have greater rms value with same max amplitude. Thus, lower dynamic range / LUFS or whatever the measure is.

I don't think anybody insisted on compressing signal by saturating amplifier, this is strongly discouraged.

I also, also, think that the biggest casualties of the war are older tracks that have been "remastered". So many have been ruined by slap-dash limiting - well, 'nuff said.

Well, yes. Don't compress stuff that originally wasn't supposed to.
Remasters could get general noise cleanup and adjusted bass lost due to technical limtations of the age, though.

[BertKoor]

I think you misunderstood. Club PA systems can deliver 120 dB SPL no problem, but regulations state it should be below 100 dB SPL or so. That's why the PA has a limiter as well!
Peaks can be 110 dB still, unless the source material is already limited.