LUFS small confusion here

[ManFromGlass]

I’ve read both SOS LUFS articles again and I get most of it but I am curious about something.
I have a track that sits around -16 LUFS, even occasional short term peaks never rise above -15. The meter on my master fader reads 0 with some peaks into the red. The LUFS plugin sits on this master fader.

So hypothetically - if I wanted to have my tune average around, say, -5 LUFS how could I even go about achieving that if my master fader is already showing the occasional red peak? I’m good with where the track sits but think I am missing understanding some relationship between the peaks on the master fader and the LUFS meter. Pushing track faders up so the master is always in the red feels like a crazy thing to do and I would expect distortion. I don’t think it matters but I am using Logic.

[sonics]

The short, simple answer is compression!

You have peaks in your audio that are a lot louder than the average levels, so you need to control them in order to increase the LUFS value.

Have you got to grips with the difference between peak, RMS, and LUFS properly?

[muzines]

The short, simple answer is compression!

Or the slightly longer, more complicated one is: develop your mix skills to be able to create a commercially-acceptably "loud" mix.

Getting a comparably "loud" track really starts at the mix process - with no thought to this, then you might need to use so much compression / limiting that the track gets to the loudness level required, but sounds absolutely terrible...

If you have a mix which you like, then push as much as you need to to get it where you want, but you might not be able to get it all the way there without destroying the track - and that case, it's mostly about finding an acceptable compromise.

[Hugh Robjohns]

I have a track that sits around -16 LUFS, even occasional short term peaks never rise above -15. The meter on my master fader reads 0 with some peaks into the red.

The LUFS meter registers the integrated loudness, which is the perceived loudness averaged across the entire track duration.

The master channel meters register the highest sample amplitude in concurrent blocks of samples.

The two are completely unrelated, but the disparity in their readings suggest that your mix has a high crest factor — a big difference in the peak-to-average ratio.

To raise the LUFS value you need to increase the average level. Pushing the fader up will do that... but will also raise the peak level, resulting in clipped transient peaks. Upwards compression will also do it without raising the peak level (much).

If you want to raise your LUFS value to (a fairly ludicrous) -5LUFS you'd need to introduce 11dB of overall gain, and also hard-limit the transients by at least the same amount, if not more... which is a lot of limiting.the mix quality will change drastically with that amount of limiting.

As said above, if you want a loud mix you really need to build it from the ground up, compressing and limiting sources individually and in subgroups, as well as in the stereo mix bus.

[Matt Houghton]

Yes, both approaches can make the figures read a certain way, but I'd definitely go with the 'slightly longer' answer if the aim is a good-sounding, loud, punchy mix: design your track from the writing and sound selection stage up to sound that way...

But I'm curious: why on earth would you want to master this hypothetical song to anything like -5 LUFS? I know a few commercial tracks that have been that loud but it *is* crazy loud, and not all material can be mixed that way. What's your actual aim here? (Hint: it shouldn't be LUFS figures!)

[Drew Stephenson]

bear in mind also that your master fader might be sitting post-FX so your plugin might not be reading your final output.

Going back to LUFS vs dBFS, the way I tend to think about it, and Hugh might shoot me down on this do I'm donning my hard hat , is it's a bit like waves hitting a harbour wall.
Peak dBFS marks the point of the highest wave, LUFS represents the something more like where most of the waves are hitting - the bit of wall that stays really wet.
And both of these are affected by the tide, which is your master fader.
A stormy day (wide dynamic range) at low tide can have louder moments than a quiet day at high tide.

You can also pointlessly extend this metaphor by saying that true peak is then the splashes of water that get thrown up out of the waves but drop back into the water never leaving a mark on the harbour wall.

[Matt Houghton]

You can also pointlessly extend this metaphor

Don't sea it. Hear it!

[Drew Stephenson]

You can also pointlessly extend this metaphor

Don't sea it. Hear it!

It's a reach to be shore.

[sonics]

Don't forget to try and trace the source of the peaks as well. Is it one track, or combination of tracks creating that peak reading? Is it a frequency you need? Perhaps you can't even hear it (like subsonics on small monitors)?

[Hugh Robjohns]

Going back to LUFS vs dBFS, the way I tend to think about it, and Hugh might shoot me down on this do I'm donning my hard hat , is it's a bit like waves hitting a harbour wall.

As analogies go, it's not bad at all....

[amanise]

.....
You can also pointlessly extend this metaphor by saying that true peak is then the splashes of water that get thrown up out of the waves but drop back into the water never leaving a mark on the harbour wall.

... then there's that flippin' monster seagull that always swoops down a nicks your chips! Nasty.

[amanise]

...If you want to raise your LUFS value to (a fairly ludicrous) -5LUFS...

Ouch - my ears are ringing already.

[RichardT]

Yes, you do need to raise the faders, but if you apply limiting at the same time the signal will not go into the red! It will be kept below 0dBFS or 0dBTP by the limiter, or whatever upper limit you give the limiter.

I understand this is theoretical and you’re not going to raise your track from -16LUFS to -5, thank goodness.

You could try it if you like and see how bad it sounds. The waveform would look like a solid block if you did.

[tea for two]

So hypothetically - if I wanted to have my tune average around, say, -5 LUFS .

ATB 9pm is one of my reference pieces for my EDM
https://m.youtube.com/watch?v=ri6Efk1SP ... bSBhdGI%3D

To get to the same loudness as ATB 9pm
My current EDM piece I'm working on in Logic :
Integrated LUFS is -22.2
True Peak is -6.8
This is because I have gain staged each track in the piece. So I can push the Master at the end to be loud in volume.

[RichardT]

.....
You can also pointlessly extend this metaphor by saying that true peak is then the splashes of water that get thrown up out of the waves but drop back into the water never leaving a mark on the harbour wall.

... then there's that flippin' monster seagull that always swoops down a nicks your chips! Nasty.

I see the seagull as a rogue synth that suddenly generates white noise at +100dBFS. Similar levels of shock.

[amanise]

The whole things a bit of a minefield. There's someone out there though! A chap called Ian Shepherd www.ProductionAdvice.co.uk has put out a really good guide on home mastering which goes into loudness pretty well. Someone alluded to it in an earlier reply - that the process of building a stand out loud track starts early at each track that gets laid down. If you control each track well you can have an easier time of pushing the master bus at the end of the mixing phase. Then on to your mastering efforts - and you shouldn't have to do too much. (howls from all mastering engineers here).
If a rock track averages -14 integrated LUFS, personally I'd find that pretty loud all the way through. For me, anything higher than -12 sounds like its really flat and harsh. Most of the old school stuff I like (and have measured) seems to fluctuate around -16 to -18. Things have been being remastered louder and louder for a long time though. Depends on the genre, your track, and your taste I'd say.

[amanise]

.....
You can also pointlessly extend this metaphor by saying that true peak is then the splashes of water that get thrown up out of the waves but drop back into the water never leaving a mark on the harbour wall.

... then there's that flippin' monster seagull that always swoops down a nicks your chips! Nasty.

I see the seagull as a rogue synth that suddenly generates white noise at +100dBFS. Similar levels of shock.

Usually the guitarists job!

[tea for two]

Pushing track faders up so the master is always in the red feels like a crazy thing to do and I would expect distortion.

So hypothetically - if I wanted to have my tune average around, say, -5 LUFS .

ATB 9pm is one of my reference pieces for my EDM
https://m.youtube.com/watch?v=ri6Efk1SP ... bSBhdGI%3D

To get to the same loudness as ATB 9pm
My current EDM piece I'm working on in Logic :
Integrated LUFS is -22.2
True Peak is -6.8
This is because I have gain staged each track in the piece. So I can push the Master at the end to be loud in volume.

I should have elaborated.

My Master Fader is in the Red permanently to get my EDM piece Volume as High Loud in Volume as ATB 9pm.
Yet it is Clean Clear, No distortion.

Because
1. I have Gain staged each track so that for the whole EDM piece
Integrated LUFS is -22 I reduced it to -25.2
True Peak is -6.8 I reduced it to -10.4

2. I managed the Gain on the Master so that pushing the volume slider to the Red permanently is still Clean Clear No distortion.

So when I crank up the volume on my Audio interface it is Clean Clear No distortion on my headphones on my speakers. When I play the piece through my hifi it is Clean Clear No distortion. Were a DJ to play this EDM piece the Volume could be cranked way high loud it would still sound Clean Clear No distortion.

I didn't use Compression, Limiter. BassRider is the only plugin I used (on the Bass track).

[amanise]

Pushing track faders up so the master is always in the red feels like a crazy thing to do and I would expect distortion.
..

ATB 9pm is one of my reference pieces for my EDM
https://m.youtube.com/watch?v=ri6Efk1SP ... bSBhdGI%3D
...
This is because I have gain staged each track in the piece. So I can push the Master at the end to be loud in volume.

...
1. I have Gain staged each track ...

Gain staging - that's the phrase that was eluding me earlier. Pushing things to their optimum as you add each track makes life a lot easier in the distortion stakes. There is an article on it in a past issue I'm sure - but there's nothing like doing it for real to embed it into your overall process. For me it starts at the beginning with the drum track. Push and push until it sounds rubbish, then back off until it sounds good again. Check your integrated LUFS. If its in the range you want to be in and sounds good, that's your datum. Put that on a bus and push it up to bounce around -1dBfs on that bus. Everything else you add then gets pushed on its own bus up towards that as you go. Eventually you end up with a pretty full sound which doesn't leave massive spikes for the mastering process to deal with. At least that's my way - that suits the music I do (which isn't EDM). We all have different methods I guess.

[Matt Houghton]

The whole things a bit of a minefield. There's someone out there though! A chap called Ian Shepherd...

Yes, it was Ian who I interviewed for one of the SOS articles mentioned at the outset.

A few years ago, he also created a series of video tutorials for us called Mastering Essentials, which are worth checking out.

And I gather he'll be a guest on one of our Q&A panels on the Saturday at Gearfest 2023 at Tileyard Studios, London.

[Drew Stephenson]

Gain staging - that's the phrase that was eluding me earlier. Pushing things to their optimum as you add each track makes life a lot easier in the distortion stakes. There is an article on it in a past issue I'm sure - but there's nothing like doing it for real to embed it into your overall process. For me it starts at the beginning with the drum track. Push and push until it sounds rubbish, then back off until it sounds good again. Check your integrated LUFS. If its in the range you want to be in and sounds good, that's your datum. Put that on a bus and push it up to bounce around -1dBfs on that bus. Everything else you add then gets pushed on its own bus up towards that as you go. Eventually you end up with a pretty full sound which doesn't leave massive spikes for the mastering process to deal with. At least that's my way - that suits the music I do (which isn't EDM). We all have different methods I guess.

Interesting. I work very differently, right from the 'definition' of gain staging.

For me, gain staging isn't about getting things loud, it's about getting them at the right level at each stage of the process: source to mic, mic pre input, DAW input, in and out through each FX in the chain, track to bus (and FX chain), bus to master (and FX chain).
The right level is one where I can make reasonably large changes without having to worry about anything going into the red. So that generally means inputs bumbling along around -18dBFS, the same for tracks (pre and post FX), busses may be a few dB higher, but I generally reckon to hit the master bus with about 6dB of headroom and around -20 to -16LUFSI without any processing on the master bus.
I'll then look use the master bus for a few light passes of compression, EQ and limiting to get things up to around -14 LUFSI and -1dBFS to simulate a mastered track.
I'll then drop the final limiting for my mastering stage so I end up with 3-6dB headroom for mastering.

[amanise]

Gain staging - that's the phrase that was eluding me earlier...

Interesting. I work very differently, right from the 'definition' of gain staging.
....

Well, its the only way to get everything up to 11 on its way in... I'm talking REAL gain. The unleaded sort.

[Wonks]

But that's not what's meant by 'gain staging'.

That's where you set the levels to maintain a similar average level of headroom (about 18-20dB) that you'd have had in an analogue mixer + tape machine setup.

It's not about getting the signal as hot as you possibly can at each stage of the process.

[RichardT]

But that's not what's meant by 'gain staging'.

That's where you set the levels to maintain a similar average level of headroom (about 18-20dB) that you'd have had in an analogue mixer + tape machine setup.

It's not about getting the signal as hot as you possibly can at each stage of the process.

Exactly - it's about minimising noise on the one hand and allowing enough headroom to cope with full signal dynamics on the other.

High recorded signal levels are not necessary to get a 'loud' track. That's all down to additional gain and compression / limiting applied during the mix or in mastering.

[Hugh Robjohns]

My Master Fader is in the Red permanently to get my EDM piece Volume as High Loud in Volume as ATB 9pm.
Yet it is Clean Clear, No distortion.

For this to have any useful meaning we need to know at what level 'the red' is on your system. Clearly, it's not 0dBFS...

For example, when I'm recording on location using my Zoom F8nPro, if the meters turn red I've hit 0dBFS and there will be distortion. If I use my Nagra VI the meters turn read at -9dBFS and I can tickle the red without problems.

(I set up all metering systems that allow it to change from green to yellow at -18dBFS and to red at -9dBFS. I find this works very well, warning me when things are getting too hot or too low so I can adjust as necessary. Having a red light come on after the signal has clipped is spectacularly pointless!)

So when I crank up the volume on my Audio interface it is Clean Clear No distortion on my headphones on my speakers.

Yay! The wonders of headroom...