Do you like to tell us which?

 

Last edited: Jun 15, 2018

Indeed, hence why I mentioned PLR and DR; and a balance/trade-off between the loudness and dynamics

 

I think we're in violent agreement here too... "without" dynamic compromise i.e. the dynamics are not compromised. The point being if you target a loud master you ARE compromising dynamics and yes it will just be turned down but if you are not targeting one of these non-standard standards when uploading a song to YT etc then you have already compromised the dynamics - not the platform - you - the engineer. Knowing the platform LUFS target (and monitoring against this) will help, so understanding this will help the engineer get the balance they want!

 

@LeviD Ah, a simple misunderstanding, I use this one

https://dmgaudio.com/products_tracklimit.php

 

Either we have pretty different information about this matter, or something got lost in my translation. I don't understand this sentence

I furthermore don't understand your connection between loundness, be it in LUFS or RMS and dynamics. If you're refering to something like this article and video, there are a few thing I totally disagree with
http://productionadvice.co.uk/plr/

When it comes to what suits best to a track, I really don't care about the standard of any platform. They are not made to make the music sound good, which is the only standard I (try to) follow, when I mix or master a track.

And please, don't get me wrong, I don't mean this offensive or aggressive in any way, I'm just confused and curious.

 

Last edited: Jun 15, 2018

No offence taken. I'm quite passionate about dynamics (which is why I mentioned PLR and DR; two measurements of dynamics and exactly what you linked above) in music (and of course quality of a final mix/master). As I said - I think we are in agreement. My point was to initially just address the original topic as to "why" these numbers exist; and my answer is -14 is a good balance between dynamics (an interpretation of music sounding "good") and loudness. The trouble being people tend not to mix with a dynamic meter and go by loudness which is a much simpler (non-standard) "standard" to implement and create a consistent volume/listening experience to the user. The point being as a mix is increased in loudness it's generally via compression, so as a consequence dynamics are compromised. -14 (which is what is generally quoted) is an average of the streaming platforms, and the reason for this is to help musicians, producers or authors to these platforms increase quality (which could be addressed as dynamics). So yes, loudness is NOT the point, but understanding what loudness means when it comes to dynamics is complex, so the powers that be (YT, Apple, Spotify etc) say -14 and IF you aim for this as a target, you're dynamics will likely be better than if you target -8 and therefore, at the same loudness will sound "better".

I whole-heartedly agree that the most important criteria is it sounds good... 100%... However, there are also dependencies on the platform or media that we cannot control ( most people know e.g. that a mix @32 bit float will not distort but can easily go over 0dB/peak... If this goes onto CD or into the digital realm like this at 16 Bit it will be horribly distorted ) so knowing the constraints or dependencies of the media is an important factor too. Again, yes they are all turned down to the prescribed loudness normalisation (not all of them turn up e.g. YT) but if your mix is finished as -8 or -6 LUFS Integrated and turned down to -14, yes it will be the same volume as other tracks on that platform/site but you have comprised the dynamics to do so; generally meaning that snares and transients will sound less crisp, more muted, duller than if that mix was prepared with less compression (overall) in order to achieve a loudness that becomes redundant on a streaming platform... That's probably slightly more than 2 cents worth but the end goal is 100% definitely, absolutely "make it sound good"

 

What I have problem with is, that there is no connection between dynamics and loudness. You can easily produce a song at -14LUFS with more dynamics than at -16LUFS.

 

Of course that may be possible as those values are low enough to provide that flexibility... But very much dependent on genre/material. It may be tenuous but the link between the two is clear. If you have extreme dynamics, there is no way you will reach extreme loudness and vice versa. If you master the same track louder, you will lose dynamics - that is a connection! It is about getting the balance between the two.

 

I can make a song with extreme limiting, 0 dynamics, and then apply a gain reduction of (x) db so it finally have -14LUFS.

But will it sound better than a song that has -14LUFS with good dynamics, transients preserved and also its macrodynamics present?? Of course not, and that's the point of the LUFS reference. There is no benefit of sacrificing dynamics to sound louder because in the end Streaming platforms will reduce the volume to their LUFS limit.

The point is: the limit of (x) platform is (x) LUFS, now you decide how much of the dynamics available you want to take advantage for your music.

 

There is a connection between dynamics and loudness and it is called peak to loudness ratio (PLR). If you use a valueo -10 to -8 you're pretty safe as far as online streaming is concerned. Meterplugs dynameter is a intuitive and simple tool to work with for that matter.

 

I always have DynaMeter on when mixing... So useful! Awesome plugin!

 

so could i get conclusion that those standard was wrong? the music contents not just only a audio track sound wave but its art, every genre has a 'ideal perceived loudness' which should not be same, and for the listeners, even in same genre, every track also perceived different regardless the loudness standard.
someone who posted a 'curve' pic at prev page but i should say its only working at static track, the music is dynamic so its useless, or probably everyone was wrong?

 

I used to use TT DR Meter (on- and offline), but it's hardly anymore neccessary for me, I can see the dynamics in the level meter too.

Those standards are made only for compilations, so that the different tracks have a similar perceived loudness. But tbh I don't want to hear a piano played ballad with the same LUFS value like EDM or HM.

That pic on the prev side shows the Fletcher-Munson curve, how we percieve loudness of the different frequencies. For instance, 1kHz at 40dBSPL are percieved as loud as 60Hz with 70dBSPL.

 

In my opinion, the conclusion would be there are no loudness standards in the music industry (there are in Broadcast), however due to the prominence of streaming platforms there are loudness limits/targets due to "loudness normalisation". An average could be made as about -14LUFS (which is why this is quoted so often, and perhaps why this was quoted by the OP).

So all loudness on these streaming sites will be the same (some will not turn up songs though) in an attempt to provide a consistent listening experience regardless of genre. The reason for this is to allow different genres to play on the same field. The differentiating factor in genres (and in my opinion, more important a point) is not loudness but Dynamic Range [DR] (and by this I mean some kind of Peak to Loudness Ratio [PLR], there are different ways to measure this). General guidance advised by many is a value of DR8 (or approximately -8 PLR, which is perhaps why this other number is quoted by the OP) however THIS is what is genre dependant and allows the "creativity" or "art" as you put it, in the musician or producers hands (and also measured across the entire duration of the track).

For example we know that Hardcore metal or Techno is intense and full-on and will probably have lower dynamic range (maybe DR6 or even less) than say a ballad which may be soft through verses with increased instrumentation or simply playing louder in a chorus (providing musical dynamics) giving it a larger DR (maybe DR16 or even more). This is very much a generalisation though as if the Hardcore metal song has periods with no drums or distorted guitar then sudden bursts of extreme "loudness", over the full duration of the song the DR will be higher (as in a bigger differentiation between the average loudness and peaks. And classic ballads may use drastic key changes to produce "dynamics" (which we could potentially hear as "emotion") but the overall average loudness remains the same. Of course, as per the post above you generally wouldn't want to hear a piano ballad at the same level as EDM but the levels provided by streaming services allow for a target level but it's too simplistic to say it's just using LUFS (YT does incorporate a level of DR management) e.g. the ballad will likely have a much larger DR than the metal track. I'd rather have a level of constistency than have the volume jump all over the place and this is why -14 works as a good average, in my opinion.

So for me, ignoring the ambiguity of the OPs question, the key takeaway is, of course, use your ears and adjust you mix to suit what you want, what you like. But the -14LUFS and -8 PLR are good guidelines for those that maybe aren't as experienced as a Bob Katz!

 

Last edited: Jun 16, 2018

so if we accept the standard, the standard will standing above and destroyed the musicality of a track, of course the online streaming service is more important but they claimed 'high quality' is just a lie which produced by mixing/mastering engineers and online service providers.

 

I'm not sure I understand the point you are making or the question you are asking. What do you mean by quality? The format of the file/streaming bit rate? If so this is a different point. Some streaming services do offer high quality (I equate this as high bit depth - again, this is dynamic related) but many compress to a lossie format due to the "streaming" nature. However if you opt to download this is different.

However if by quality you mean production quality of the music production I would disagree. I think this target of loudness normalisation helps retain musicality and quality. I think if producers, engineers are targeting this "loudness normalisation" level (again, this is NOT a standard but an approximate average) then this only helps improve musicality. Yes, you can kill/crush a mix and reduce it to -14LUFS but if you actually target this final loudness level during the mix/master at -14LUFS then most modern styles of music have plenty of room for musicality/dynamics (e.g. DR12-DR18 - again very dependant on the style/genre and actual composition/arrangement) and the loudness is taken care of by the platform (i.e. it allows you not to compromise "quality" (dynamics) by feeling forced to push the loudness to e.g. -6 LUFS to be comparable to commercial peers (or even louder on some recent commercial productions). Again, this makes loudness not the main factor but dynamics/musicality which is better for music production... In my opinion.

As mentioned before, this is not a standard, and the numbers are largely arbitrary; the main goal is make your music sound good, balanced, dynamic, exciting, emotional (and not kill musicality). These numbers just help provide a guide which supports this and the loudness normalisation being approximately -14LUFS (some lower some higher) promote this type of thinking.

Oh, and some mix/master engineers and streaming providers may well lie (not sure what about) but the point is the target they are setting, if applied correctly, will actually help return some musicality and "quality" to productions...

 

@meow
Here is a list with album DR levels starting from '81, but there are also some earlier albums in there, lots of famous ones.
http://dr.loudness-war.info/album/list/year/40

The dude who runs the site has categorized 122.588 albums until today. You can change the sorting and see how it got louder and louder during the years.

Fun fact: In 2017 japanese artist Ninja McTits produced an album that has a min, max and an average DR of "0"

It is the only one with that specs in the whole list.
I wonder how that sounds and how long people are able to listen to it.

 

This is only slightly off-topic; I was wondering what everyones prefered meters are, especially the ones that ones can't live without. I'm not new to mastering, but I'm also not where I want to be at (and yes, I know that it's always something that you can improve upon,) so, I've been toying around with different types/brands and I was hoping to narrow down the long list of plugins (I am running Win7Ult. 64,) to try out that will ensure that it is only me and my decisions fucking up a song, not faulty or less-than-stellar meters.
A quick comment/list of shameless promotion for a metering plugin(s) of any type would be appreciated.

 

Last edited: Jun 17, 2018

If OP is fine with it.
DMGAudio TrackMeter, RNDigital IXL Level Meter, WaveLab Level Meter (and lately less but, still great, TT DR Meter).

EDIT: Voxengo SPAN free and if you're looking for VU meters, I can recommend TBProAudio MV Meter2, it's also free (you'll have to adjust the calib level on the upper left for some measurings).

 

Last edited: Jun 17, 2018