LUF loudness for spotify and youtube

[alexhedleymusic]

Hi,
Im having a bit of confusion as to what to make my master luf level. Iv'e looked online and it says spotify and youtube luf is around -14.... but this only seems about right for youtube... to me spotify sounds a lot quieter which is annoying. my master is peaking at about -14 and this sounds roughly the same as youtube music videos... but to be honest mine might even sound a little bit louder which is weird as that would sugguest that youtube is actually quieter than -14.... but spotify on the other hand my track is way louder that spotify... but if i send my track to upload to spotify is the volume reduction going to ruin the sound or would it only work the other way round if my track was too quiet. the thing is is at the moment from looking at the luf meter on my track this doesnt seem to be in line with the volumes of youtube and spotify that is being suggested from articles online. im starting to wonder wether i should just go with my ear instead and make two different masters one for youtube and one for spotify and just match them using my ear. confusing anyway... the youtube ones not a problem as i can check that before it goes out but there doesnt seem to be a feature with spotify where you can check the sound before publishing.
thanks anyway all help is appreciated

[Kwackman]

Iv'e looked online and it says spotify and youtube luf is around -14........ my master is peaking at about -14

I think these are different things.
The "-14dB" loudness value you found for Spotify & Youtube refer to the integrated loudness value for your tune.
Peaking around -14dB is different, and I'd guess it would give a lower than "-14dB" integrated reading.

There's info in this article that might help?

https://www.soundonsound.com/techniques ... udness-war

And this site..
https://productionadvice.co.uk

[Mike Stranks]

Kwackman's right.

A LUFS value of -14 is not the equivalent of a peak of -14 dBFS.

Do you have LUFS meter in your DAW or as an ancillary?

I know not of Spotify, but my stuff generally ends up at about -14 LUFS prior to uploads to YouTube. And of course, if you're markedly higher than -14 LUFS, YouTube will adjust your mix to be quieter anyway.

[alexhedleymusic]

Kwackman's right.

A LUFS value of -14 is not the equivalent of a peak of -14 dBFS.

Do you have LUFS meter in your DAW or as an ancillary?

I know not of Spotify, but my stuff generally ends up at about -14 LUFS prior to uploads to YouTube. And of course, if you're markedly higher than -14 LUFS, YouTube will adjust your mix to be quieter anyway.

hey mike. i think maybe i was using the loudness metre wrong. im not sure if it makes a difference but i had the loudness metre before compression and limiters on the output channel. im not sure if that makes a difference but now putting the loudness metre at the end of te chain im getting readings which seem to make more sense... on the metre my track is reading as LU range - 7.7 and integrated is -9.9 and i cant see anything to do with true peak on the metre..... as far as lu range and integrated is this anywhere in the correct ball park for either spotify or youtube. sorry these probably seem like stupid questions but any really straight forward answers would be great to know what to aim for. also how to find true peak and what that would need to be for spotify and youtube would be good to know. am i essentially looking for LU range, integrated and true peak as things to aim for for different streaming services? also i did look through the article sent by Kwackman(thanks for that) but it did go over my head slightly and i got a bit confused with it
thanks!

[alexhedleymusic]

Kwackman's right.

A LUFS value of -14 is not the equivalent of a peak of -14 dBFS.

Do you have LUFS meter in your DAW or as an ancillary?

I know not of Spotify, but my stuff generally ends up at about -14 LUFS prior to uploads to YouTube. And of course, if you're markedly higher than -14 LUFS, YouTube will adjust your mix to be quieter anyway.

hey mike. i think maybe i was using the loudness metre wrong. im not sure if it makes a difference but i had the loudness metre before compression and limiters on the output channel. im not sure if that makes a difference but now putting the loudness metre at the end of te chain im getting readings which seem to make more sense... on the metre my track is reading as LU range - 7.7 and integrated is -9.9 and i cant see anything to do with true peak on the metre..... as far as lu range and integrated is this anywhere in the correct ball park for either spotify or youtube. sorry these probably seem like stupid questions but any really straight forward answers would be great to know what to aim for. also how to find true peak and what that would need to be for spotify and youtube would be good to know. am i essentially looking for LU range, integrated and true peak as things to aim for for different streaming services? also i did look through the article sent by Kwackman(thanks for that) but it did go over my head slightly and i got a bit confused with it
thanks!

LU range is actually 7.7 not minus 7.7 as i said before

[The Elf]

hey mike. i think maybe i was using the loudness metre wrong. im not sure if it makes a difference but i had the loudness metre before compression and limiters on the output channel.

It makes ALL the difference! You need to meter at the very end of the chain.

You owe it to yourself to get a grasp of loudness metering - it is the world we now live in. There are a number of YouTube videos that will likely help.

-9.9LUFS is high. You will need to examine the overall level and (arguably more importantly) dynamics of your mix to move towards the target reading. A couple of LUs either way is probably OK.

[Kwackman]

I'm assuming you're using a DAW and a software meter and not a hardware one?

What DAW and loudness meter are you using?

Yes, you're correct, the loudness meter should be at the very end of your chain.

I reckon the integrated loudness figure is the most important one. This will decide how much your streaming service wants to turn you're music up or down. Aim for around -14 and they'll hopefully leave it alone!
This integrated figure is the average loudness of your tune from start to finish, so your meter must "listen" to your tune from start to finish without stopping.

The true peak is important to avoid digital distortion (not a pretty thing), and I think -1dB is the rule.

The loudness range is an indication of the dynamic range of your tune, and I don't if there are guidelines here, as it will greatly depend on the style of music. A solo acoustic guitar will probably have a greater range than a dance track!

Your loudness meter should also show you, as you mix, the momentary loudness level. This is the rough equivalent to an "ordinary" meter showing peaks. This (and there might also be a reading of "short term" loudness -averged over a few seconds) should give you a clue to how close to the -14dB goal you are getting.

I would suggest, unless your mixes are far too loud or quiet, that you get too hung up about loudness as you are mixing. Keep the peak levels well down and mix your song until it sounds good. Then check on the loudness meter. If it's with-in a few dB of the -14 goal, then a simple gain control on your output channel will solve that.

This has probably confused you even more, and be aware I'm not an expert in this! Others hopefully will join in and correct (m)any errors!

[alexhedleymusic]

I'm assuming you're using a DAW and a software meter and not a hardware one?

What DAW and loudness meter are you using?

Yes, you're correct, the loudness meter should be at the very end of your chain.

I reckon the integrated loudness figure is the most important one. This will decide how much your streaming service wants to turn you're music up or down. Aim for around -14 and they'll hopefully leave it alone!
This integrated figure is the average loudness of your tune from start to finish, so your meter must "listen" to your tune from start to finish without stopping.

The true peak is important to avoid digital distortion (not a pretty thing), and I think -1dB is the rule.

The loudness range is an indication of the dynamic range of your tune, and I don't if there are guidelines here, as it will greatly depend on the style of music. A solo acoustic guitar will probably have a greater range than a dance track!

Your loudness meter should also show you, as you mix, the momentary loudness level. This is the rough equivalent to an "ordinary" meter showing peaks. This (and there might also be a reading of "short term" loudness -averged over a few seconds) should give you a clue to how close to the -14dB goal you are getting.

I would suggest, unless your mixes are far too loud or quiet, that you get too hung up about loudness as you are mixing. Keep the peak levels well down and mix your song until it sounds good. Then check on the loudness meter. If it's with-in a few dB of the -14 goal, then a simple gain control on your output channel will solve that.

This has probably confused you even more, and be aware I'm not an expert in this! Others hopefully will join in and correct (m)any errors!

Hey

no this has definitely cleared it up a lot more. im using logic pro x and theres a loundess metre which does all the things youve described... but cant find true peak. but iv'e resolved this by using the multimetre which has all of these functions. so put simply i should aim for integrated at around -14 and true peak under -1? pretty sure youve answered that already but thats the general jist of things?

[The Elf]

so put simply i should aim for integrated at around -14 and true peak under -1? pretty sure youve answered that already but thats the general jist of things?

-14LUFS (Loudness Units Full Scale) and -1dBTP (dB True Peak). Sorry to be pedantic, but these things are important to get right from the start - it will save confusion and problems later!

[alexhedleymusic]

so put simply i should aim for integrated at around -14 and true peak under -1? pretty sure youve answered that already but thats the general jist of things?

-14LUFS (Loudness Units Full Scale) and -1dBTP (dB True Peak). Sorry to be pedantic, but these things are important to get right from the start!

This is my first time mixing a track. This is the start!

[Kwackman]

so put simply i should aim for integrated at around -14 and true peak under -1? pretty sure youve answered that already but thats the general jist of things?

-14LUFS (Loudness Units Full Scale) and -1dBTP (dB True Peak). Sorry to be pedantic, but these things are important to get right from the start - it will save confusion and problems later!

Told you someone would correct my errors!
Elf is totally spot on here.
The correct terminology and correct units is really important.

[The Elf]

This is my first time mixing a track. This is the start!

All the more reason to grasp all these things right now - it will stand you in good stead!

[alexhedleymusic]

This is my first time mixing a track. This is the start!

All the more reason to grasp all these things right now - it will stand you in good stead!

yes this is true.... but im still a bit confused as it says on the loudness metre
m,s,i. (medium, short term, integrated)?

so you say to read the lufs to -14 but theres three things to look at so how do you read the lufs when theres three different readings?

Thanks

[Kwackman]

but im still a bit confused as it says on the loudness metre
m,s,i. (medium, short term, integrated)?

also i did look through the article sent by Kwackman(thanks for that) but it did go over my head slightly and i got a bit confused with it
thanks!

Yeah, it requires a lot of cups of coffee to get through, but an excellent article.
I read it about once a year!

If you scroll near the bottom there's a "box" with the title "Typical Loudness Meter Displays", which might help you?

[Wonks]

You're aiming for an integrated loudness of -14LUFS, which should tell you that you need to use the integrated value. This value needs to be for the whole track, so you can't just look at the integrated value for a short section, you need to run the track from start to finish and look at the result.

[Drew Stephenson]

They're measuring the same thing but over different time periods.
If I recall correctly:
M = momentary, fractions of a second.
S = short-term, a few seconds.
I = integrated, the whole piece.

It's the integrated one that the streaming services are interested in.

[Hugh Robjohns]

...but im still a bit confused as it says on the loudness metre
m,s,i. (medium, short term, integrated)?

Try reading the article that Kwackman linked above.

There are three different measurements options involved in loudness normalisation, but the only one that matters for streaming services is the Integrated measurement and that has to be calculated across the entire track, from start to end. It will generate bogus numbers if you only analyse a part of your track.

The Medium option works much like a VU meter and is intended to help find an initial level in the right ball park when you start a mix. The Short term option measures over a (roughly) 4 second rolling time window and is used primarily for assessing very short tracks (under 30 seconds) mostly in TV advertising.

so you say to read the lufs to -14 but theres three things to look at so how do you read the lufs when theres three different readings?

Ignore M and S, just worry about I, and only worry about it at the end of your track, not during!

And while the absolute limit for True Peak is -1dBTP, it's often better to keep it closer to -3dBTP as that avoids problems when the material passes through lossy codecs (MP3, AAC etc) which many streaming services are using.

[alexhedleymusic]

You're aiming for an integrated loudness of -14LUFS, which should tell you that you need to use the integrated value. This value needs to be for the whole track, so you can't just look at the integrated value for a short section, you need to run the track from start to finish and look at the result.

ok cool sorry to keep banging on about this but should i just make one master to send to all streaming and aim for these levels and it should be ok still?

[alexhedleymusic]

They're measuring the same thing but over different time periods.
If I recall correctly:
M = momentary, fractions of a second.
S = short-term, a few seconds.
I = integrated, the whole piece.

It's the integrated one that the streaming services are interested in.

Thanks for this!

[alexhedleymusic]

...but im still a bit confused as it says on the loudness metre
m,s,i. (medium, short term, integrated)?

Try reading the article that Kwackman linked above.

There are three different measurements options involved in loudness normalisation, but the only one that matters for streaming services is the Integrated measurement and that has to be calculated across the entire track, from start to end. It will generate bogus numbers if you only analyse a part of your track.

The Medium option works much like a VU meter and is intended to help find an initial level in the right ball park when you start a mix. The Short term option measures over a (roughly) 4 second rolling time window and is used primarily for assessing very short tracks (under 30 seconds) mostly in TV advertising.

so you say to read the lufs to -14 but theres three things to look at so how do you read the lufs when theres three different readings?

Ignore M and S, just worry about I, and only worry about it at the end of your track, not during!

And while the absolute limit for True Peak is -1dBTP, it's often better to keep it closer to -3dBTP as that avoids problems when the material passes through lossy codecs (MP3, AAC etc) which many streaming services are using.

Thanks this has helped a lot appreciate it! also i just have one other question and thought i might as well ask you as you seem to know what youre saying rather than start a whole new thread. the correlation metre on my track is fine for the majority of the track but when it goes into little instrumental sections it goes slightly over in to the red but only kind of jumps into it quickly and then goes back into the green. would you say this is a major issue. iv'e checked the song in mono and it does sound worse but thats probably to be expected to some extent... still sounds pretty ok to me though. also i can correct this issue on the correlation metre by changing the panning of the guitars but it makes the sound of the track a lot worse to me and just much less wide which is not whats wanted. shouldnt matter if it dips into the red every now and then though should it? as long as its not like sitting there for periods of time