The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Streaming services LUFS confusion. Whats the answer??
Old 3 weeks ago
  #1
Here for the gear
 

Streaming services LUFS confusion. Whats the answer??

Hey guys, apologies if this has been answered before but i just can’t seem to wrap my head around this...

It’s to my understanding that spotify requests masters with approx -12 LUFS (give or take) and if it’s louder than this they will supposedly turn it down to match.

Well i’ve been doing some experimenting with this theory. I’ve noticed that if i push a mix to say around -9/-8 LUFS and play that original audio file, it still doesn’t match the loudness of playing a song on spotify over the same speakers at the same volume setting. So if Spotify really do turn down the volume of a song if it’s louder than -12 LUFS, how come they still all sound louder than my mix at -8 LUFS??? And this isn’t just on one song either.

Am i missing something here? What’s really going on and what should i be aiming for? To compete with spotify i’m having to push to, at the very least -8 LUFS or above.

Many thanks
Old 3 weeks ago
  #2
Lives for gear
Quote:
Originally Posted by ItsMisterVee View Post
...
Never tested that but on the settings spotify gives you 3 output options. Default is normal and there's quiet and loud.

Just guessing but maybe quiet option is the original lufs match between songs and normal (and loud) hypes loudness (?).
Old 3 weeks ago
  #3
Here for the gear
 

Quote:
Originally Posted by cooker View Post
Never tested that but on the settings spotify gives you 3 output options. Default is normal and there's quiet and loud.

Just guessing but maybe quiet option is the original lufs match between songs and normal (and loud) hypes loudness (?).
Hey man, good point.
In order to see those 3 options, ‘audio normalisation’ has to be checked. I religiously set this to off as i don’t want any alterations to the music that i’m hearing. I think by default this is switched to ‘on’ though. So maybe their algorithms are dependant on this feature being switched on. I just don’t know which one is the true default.
Weirdly though, when audio normalisation is off, songs that are supposedly capped at -12 LUFS are noticably louder than my original wav files at -8
Old 3 weeks ago
  #4
Here for the gear
 

I just compared the volume settings on spotify. When normalisation is switched off, the songs are actually louder than the ‘loud’ setting when normalisation is on.
I wonder if this setting bypasses spotifys -12 LUFS algorithms? Anyone know anything about this?
Makes me wonder still what the best level is to send my masters to them. I still think aiming for louder must be better... in my opinion -12 LUFS is very quiet. I can push my mixes beyond this and still not get much limitting occur until i get to around -9 LUFS.
It would seem others are sending in their masters hotter too as when normalisation isn’t checked, a level more consistent with -8 LUFS is apparent.
Old 3 weeks ago
  #5
Lives for gear
Quote:
Originally Posted by ItsMisterVee View Post
...
Yeah its switched on default, and I bet almost no one gets into these settings so if mastering for spotify then its better to keep the settings as provided.

But on the other hand they may update something any minute and all considerations may be useless too. I think it was just 6 or so months they changed the lufs target.

Its common but I like the answer given here to just make it sound good and not pushing further for loudness sake.

My tip, focus on experiment/researching and if possible experience mfit. At first I also checked subjects 1 by 1 (isp, loudness, normalization, lufs...) but things started to make more sense to me when I got into the mfit process from a (although I consider myself a hobbyist) ME's point
Old 3 weeks ago
  #6
Lives for gear
 
SmoothTone's Avatar
 

Tracks play at their full volume unless loudness normalisation is switched on. The targets are Normal -14, Quiet -22 and Loud -11LUFS (approximately).

Spotify webplayer doesn't use loudness normalisation and it must be turned on in the app. A lot of people are still delivering loud masters because of this, particularly in EDM genres where the loudness war still rages in DJ sets.

If you have normalisation turned on and your track still sounds quieter it may be something skewed in the spectral balance of your mix or in your monitoring environment.

If you listen to a bunch of stuff that you want to match in terms of loudness and set your monitoring level to play those at a comfortable level, set your monitoring to that level when mastering your tracks and you'll be able to do this completely by ear. Your masters will sound as loud as the others because they sound as loud as the others, no LUFS measurements needed. If you then start to notice that your tracks sound like they lack clarity, punch or power compared to others, then you have a good reference point to work from.
Old 3 weeks ago
  #7
Here for the gear
 

Quote:
Originally Posted by SmoothTone View Post
Tracks play at their full volume unless loudness normalisation is switched on. The targets are Normal -14, Quiet -22 and Loud -11LUFS (approximately).

Spotify webplayer doesn't use loudness normalisation and it must be turned on in the app. A lot of people are still delivering loud masters because of this, particularly in EDM genres where the loudness war still rages in DJ sets.

If you have normalisation turned on and your track still sounds quieter it may be something skewed in the spectral balance of your mix or in your monitoring environment.

If you listen to a bunch of stuff that you want to match in terms of loudness and set your monitoring level to play those at a comfortable level, set your monitoring to that level when mastering your tracks and you'll be able to do this completely by ear. Your masters will sound as loud as the others because they sound as loud as the others, no LUFS measurements needed. If you then start to notice that your tracks sound like they lack clarity, punch or power compared to others, then you have a good reference point to work from.
Many thanks for your reply. I’ve noticed that with normalisation turned ‘on’ my mixes are now far louder than whats playing in Spotify. Even when the normalisation level is set to ‘loud’ so thats reassuring! And is consistent with the LUFS readings you mentioned.
I’m still unsure as to what level i should submit my tracks to spotify though. If i send in a track a -12 LUFS for example (like they ask for) it will most likely sound consistent with everything else on all three of the normalisation settings... however if the end listener has their normalisation switched ‘off’ then my mix will still be -12 LUFS, but most other tracks will be at a more expected level of -9/-8 as i’ve just found out when testing.
Old 3 weeks ago
  #8
Lives for gear
 
Ben F's Avatar
 

Verified Member
Mastering to a number when there are no set standards will always be a moving target.
Old 3 weeks ago
  #9
Here for the gear
 

Quote:
Originally Posted by Ben F View Post
Mastering to a number when there are no set standards will always be a moving target.
What would you recomend to at least get in the ballpark??
Old 3 weeks ago
  #10
Lives for gear
 
SmoothTone's Avatar
 

Quote:
Originally Posted by ItsMisterVee View Post
Many thanks for your reply. I’ve noticed that with normalisation turned ‘on’ my mixes are now far louder than whats playing in Spotify. Even when the normalisation level is set to ‘loud’ so thats reassuring! And is consistent with the LUFS readings you mentioned.
Great. Problem solved.

Quote:
I’m still unsure as to what level i should submit my tracks to spotify though. If i send in a track a -12 LUFS for example (like they ask for) it will most likely sound consistent with everything else on all three of the normalisation settings... however if the end listener has their normalisation switched ‘off’ then my mix will still be -12 LUFS, but most other tracks will be at a more expected level of -9/-8 as i’ve just found out when testing.
Spotify doesn't require a LUFS target, they recommend -14LUFS. But one of the benefits of loudness normalisation is that if it sounds good at -10 or -9 or even -8 LUFS, it will still sound just as good when turned down to whatever the target is and play at the same perceived level as everything else.

Mastering and loudness – FAQ – Spotify for Artists

My recommendation is to aim for the loudness (crest factor) that sounds best for the music. If that happens to fall somewhere between noticeably quiet in non-normalised scenarios and noticeably squashed in normalised scenarios, then you're good to go.
Old 3 weeks ago
  #11
Lives for gear
 
Ben F's Avatar
 

Verified Member
Quote:
Originally Posted by ItsMisterVee View Post
What would you recomend to at least get in the ballpark??
Master the music so it best suits the mix, genre and target delivery.

It's not an arbitrary number or bunch of numbers. People are obsessed by this. All of my masters from 10 year ago still sound great on iTunes, youtube, Spotify, etc as do the newer masters. If I did these masters to exact LUFS numbers then it would have changed several times already. Like I said, music loudness is a moving target with no set standard.

People peddling exact numbers are confusing everyone.
Old 3 weeks ago
  #12
In general, more dynamic means the loudest peaks will tend to sound more clear, and the material will potentially sound more punchy, and it will sound like the peaks hit harder on the loudest peaks of a track that has been squashed and normalized.

In general, it always depends on the content.

However, I suggest that you take a look at something from Airwindows, in which Chris J uses soft saturation in his usual genius-way of pulling no punches.

Spotify might not do normalizing unless its turned on, on the user end. However, youtube does. Also, if albums are sent out to aggregate companies which distribute them to all of the leading streaming radio websites, then I would suggest considering that most of those platforms do normalization on louder material. They also have a set standard, which amongst the leaders seems to be -14 LUFS (I am unsure if Spotify updated and is now saying that -12 is best for their platform). Also, iunes radio is at-16, and I doubt they're going to budge. And, aggregate companies are likely to send it to itunes radio, as long as it has met apple's standards. I would recommend -14 or -13 LUFS for most material, and then to use several strategies that are aimed towards pushing the content to sound louder on streaming platforms.

Check out his description of how Chris J was able to handle Youtube, in 2015. Since then, he has released an updated Righteous4, which you can download for free, since Chris J is supported by Patreon.

Righteous 2 | Airwindows
Old 3 weeks ago
  #13
Lives for gear
 
SmoothTone's Avatar
 

Man, I wish people would just check the specs at the source instead of endlessly speculating and clouding the issue further.

Here it is again: Mastering and loudness – FAQ – Spotify for Artists
Old 3 weeks ago
  #14
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by SmoothTone View Post
Man, I wish people would just check the specs at the source instead of endlessly speculating and clouding the issue further.

Here it is again: Mastering and loudness – FAQ – Spotify for Artists
From that page:

" - 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.
- 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.
"

That alone is reason enough to NOT try to target certain LUFS level for streaming.

I'd rather have my limited track turned down than have Spotify's limiter turn it up!
Old 3 weeks ago
  #15
Lives for gear
 
RightOnRome's Avatar
I just finished and put out a record at -14 lufs ..to make sure it was right ..and it seems that the record comes on quieter then most new records on itunes..not doing that again! ..im going with my gut next time !! this is ridiculous!
Old 3 weeks ago
  #16
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by RightOnRome View Post
I just finished and put out a record at -14 lufs ..to make sure it was right ..and it seems that the record comes on quieter then most new records on itunes..not doing that again! ..im going with my gut next time !! this is ridiculous!
I've heard this same complaint many times.

Sadly, loudness normalization algorithms aren't standardized across platforms and they just plain don't work sometimes.

If/when the technology improves enough, then I'll be happy to recommend those loudness targets. But we aren't there yet...
Old 3 weeks ago
  #17
Quote:
Originally Posted by Trakworx View Post
If/when the technology improves enough, then I'll be happy to recommend those loudness targets. But we aren't there yet...
Not sure if it's the technology. My take on this is that the true underlying objectives are not primarily about an equal loudness high quality playback for the listener.

Occasionally listening to music via Spotify free, Deezer and Youtube, I can clearly hear that loudness control primarily guarantees ad customers a generous safety margin. Ads are systematically louder, i.e. twice as loud as the "filling material" heard between ad blocks.

IMHO, all this is not at the independent musician's advantage. But for ad agencies and their customers, loudness control (over the others) is a gift.
Old 3 weeks ago
  #18
Here for the gear
 

Quote:
Originally Posted by Trakworx View Post
From that page:

" - 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.
- 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.
"

That alone is reason enough to NOT try to target certain LUFS level for streaming.

I'd rather have my limited track turned down than have Spotify's limiter turn it up!
Great post! Based on what i’ve read in this discussion, ithink this is the way forward to bypass any destructive processing on their end.
Plus it ticks alot more boxes because if the end listener has normalisation activated, your mix will sound consistent with everything else (within reason) and if they have it deactivated, your track will still be loud enough in accordance with everyhting else.
Once again, when normalisation is deactivated, most tracks i played on spotify were WAY higher than -14 LUFS.
So to summarise, imo it would appear that the loudness wars are still relevant, unless spotify permanently activates normalisation without the possibility of turning it off.
Unless i’m mistaken :/ haha
Old 3 weeks ago
  #19
Lives for gear
Quote:
Originally Posted by ItsMisterVee View Post
imo it would appear that the loudness wars are still relevant
LUFS isn't perfect by the way just like how rms was but it does help, unlike cd the perceived/added output isn't as much so specially eq tweaks favoring loudness can sound really bad. Many 90's metal artists as far as I noticed remastered their old albums because of this.

If spotify released a free check tool just like how apple did, things would be so much clearer for everyone. Instead they copy what mfit does and give it as a written description instead of a program. The program responds to various scenario's in mastering, written advice does not.
Old 3 weeks ago
  #20
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by ItsMisterVee View Post
imo it would appear that the loudness wars are still relevant, unless spotify permanently activates normalisation without the possibility of turning it off.
Yes, Spotify and also every other streaming service would have to do that. And iTunes along with every other media player. That would leave CDs as the only remaining bastion of the loudness war and we could eventually stop worrying about loudness once clients realized what had happened.

But in addition to that, a standard loudness that will not change in the future, plus a standard way to measure loudness would be needed so we could actually know what we're doing.

I don't see all of that happening very soon...

And even if it did happen there would still be the problem on services offering "Loud" settings where their own limiters kick in. Oy.
Old 3 weeks ago
  #21
Lives for gear
 
Trakworx's Avatar
Quote:
Originally Posted by cooker View Post
LUFS isn't perfect by the way just like how rms was...
I've been thinking this for a while. Despite the science behind it, LUFS isn't all it's cracked up to be. I use RMS meters alongside LUFS meters and often I find RMS to be just as good an indicator of perceived loudness. It depends on the spectral content of the music. RMS responds more to low end and LUFS to upper midrange. Neither of them tells the whole story by itself, which is why I use both.
Old 3 weeks ago
  #22
Lives for gear
 
Trakworx's Avatar
" - Negative gain is applied to louder masters so the loudness level is at ca - 14 dB LUFS."


Interesting how Spotify talks about LUFS even though they don't use LUFS in their algorithm (rather ReplayGain).

And they say they're going to switch to LUFS in the future, so yet another moving target!


"We currently use ReplayGain, which was the most recognized standard for calculating loudness when Spotify first started.

In the future, we plan to use a new standard for calculating loudness, called ITU 1770 (from the International Telecommunication Union). This defines the integrated LUFS (Loudness Units Full Scale) measure, and it’s what we recommend you use to measure the loudness of your tracks.

ReplayGain doesn’t specify a measurement unit for loudness, so we’re unable to give an exact measure in LUFS used by ITTU 1770. However, we adjust tracks to 3 dB higher than ReplayGain algorithm specifies, which is roughly equivalent to -14 dB LUFS, according to the ITU 1770 standard.
"


Mastering and loudness – FAQ – Spotify for Artists
Old 3 weeks ago
  #23
Lives for gear
 
SmoothTone's Avatar
 

Quote:
Originally Posted by Trakworx View Post
Interesting how Spotify talks about LUFS even though they don't use LUFS in their algorithm (rather ReplayGain).

And they say they're going to switch to LUFS in the future, so yet another moving target!
ReplayGain 2.0 adopted the same k-weighting filter as ITU BS. 1770-3 (LUFS), so the target shouldn't move too much.

Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Matt Pumpkin / Mastering forum
4
jajohu / Mastering forum
12

Forum Jump
Forum Jump