# Mastering level question

Page 1 of 2

### Mastering level question

Posted: Sun Aug 04, 2019 5:47 pm
Im pretty new to mastering, so Ive been reading up and watched some video tutorials about levels for streaming services and Im a bit confused about the information I found.

When leveling for Youtube, (-14LUFS) in one video was said to level at -14LUFS average, while another said to level at -8LUFS short term. I understand from this that average should be -14, with the max peaks at -8. However, now Im trying to level, my average and short term are very close together, about 1.5LUFS difference max, so I tried to increase short term, without affecting average, to meet the best vallues. But no matter what I do, when I inscrease peak, my average will follow in the same way, I cant find how to increase the difference between those two vallues.
Am I overlooking something in the mastering process that can increase the difference between peak and average to such extend, or did I misinterprete the given information about these 2 levels?

### Re: Mastering level question

Posted: Sun Aug 04, 2019 5:57 pm
Ok, couple of things here before the experts (I'm not one) chip in...
Short term LUFS isn't the same thing as peaks - it's measured over a few seconds. So your loudest section short term figure can be around -10 whilst your peaks are up at -3dB.
Average LUFS is measured over the course of the piece (excluding silences) but if you're finding that there's not much difference between your Average and Short Term figures then arguably you're mixing too hot in the first place.
It's all horses for courses and will be programme dependent, but if you want to meet those suggested levels (and they're just suggestions remember) then you'll need to look at the mix before you get to mastering.

### Re: Mastering level question

Posted: Sun Aug 04, 2019 6:14 pm
Thanks, that already gives better insight. :)

Im trying to meet the given vallues to have a starting point for the exercise. When I learn more about it I'll go my own way, anyways. ;)

I mixed at k-20 and I have a dynamic range of 6-9dB, measured with BX Meter.
After processing during mastering, my dynamic range is 5-7 (also BX Meter) and around 10LU dynamic range and that is measured with Youlean Loudness meter.

### Re: Mastering level question

Posted: Sun Aug 04, 2019 8:58 pm
Demious wrote:Im pretty new to mastering, so Ive been reading up and watched some video tutorials about levels for streaming services and Im a bit confused about the information I found.

When leveling for Youtube, (-14LUFS) in one video was said to level at -14LUFS average, while another said to level at -8LUFS short term. I understand from this that average should be -14, with the max peaks at -8. However, now Im trying to level, my average and short term are very close together, about 1.5LUFS difference max, so I tried to increase short term, without affecting average, to meet the best vallues. But no matter what I do, when I inscrease peak, my average will follow in the same way, I cant find how to increase the difference between those two vallues.
Am I overlooking something in the mastering process that can increase the difference between peak and average to such extend, or did I misinterprete the given information about these 2 levels?

There are a number of LUFs. Read the standard. Then read how your streaming service changed/interpreted the standard for their use.

There is long term whole show LUFS and there is shorter term LUFS as well as very short peak LUFS. There might even be another one in the standard somewhere depending which one you use.

I think the intent was to limit DR to something reasonable to avoid the loudness war, while also preventing commercials and short term content from overwhelming the main show and longer term content , while not letting very short things like a gunshot in a movie affect any of those others. It also ignores long periods of silence or near silence for the same reason.

### Re: Mastering level question

Posted: Sun Aug 04, 2019 9:01 pm
Demious wrote:Thanks, that already gives better insight. :)

Im trying to meet the given vallues to have a starting point for the exercise. When I learn more about it I'll go my own way, anyways. ;)

I mixed at k-20 and I have a dynamic range of 6-9dB, measured with BX Meter.
After processing during mastering, my dynamic range is 5-7 (also BX Meter) and around 10LU dynamic range and that is measured with Youlean Loudness meter.

That DR sounds just a bit low for my taste, but not really bad.

What is the source material?
Are you compressing things? What are the settings on that ?

### Re: Mastering level question

Posted: Sun Aug 04, 2019 9:13 pm
The important thing is to start with something that sounds right to your ears. Now check the level. If the integrated level is around -14LUFS you are pretty much there. If it is lower then you need to look at the reasons for it being lower. If there are just a few short spikes then bringing up the overall level and limiting those spikes is the obvious answer. If limiting doesn't work then you need to ask yourself 'do I really want this louder?'

If the integrated level is too high then you need to try bringing the level down and see if the mix still works at a lower level as that's the level people will hear it when it is mixed in with other songs in a playlist. You then need to decide whether to master at the lower level or master at a higher level and have the level automatically reduced when it is played as part of a playlist.

The LUFS level is a guide - I'm still delivering certain masters higher than -14LUFS because that material is often dirty grungy garage rock that works at a higher level on CD (and that's how it has been mixed).

### Re: Mastering level question

Posted: Sun Aug 04, 2019 9:38 pm
hobbyist wrote:
Demious wrote:I have a dynamic range of 6-9dB...
That DR sounds just a bit low for my taste...

And yet only a few posts ago we were told:

hobbyist wrote:Good dynamics are those under 10dB for me.

Hmmm... ;-)

### Re: Mastering level question

Posted: Sun Aug 04, 2019 9:53 pm

Hmmm again! It might be more helpful if you (re)read it yourself, as your post is vague to the point of valueless at best, and fundamentally incorrect and misleading at worst!

There is long term whole show LUFS and there is shorter term LUFS as well as very short peak LUFS.

Broadly right, but more helpful to use the correct meter terms. The standard defines three types of loudness measurement. Integrated, which is measured across the entire programme duration, Short-term or Sliding, which is measured over a sliding 3-second window, and Momentary which is measured over 400ms (slightly slower than the integration time of a VU meter). There is no 'short peak' loudness measurement -- and using the term 'peak' is rather confusing in that context!

There might even be another one in the standard somewhere depending which one you use.

Nope. Only those three loudness measurements. The other associated measurments described in the standards are the True Peak value, and the LRA or loudness range. (It's worth noting, though, that the LRA figure is not directly comparable to the DR figure shown on meters like the BX).

I think the intent was to limit DR to something reasonable to avoid the loudness war...

Nope. Quite the reverse, actually. Loudness normalised systems encourage wider dynamic ranges, not restricted ones.

It is through matching programmes by their perceived loudness, rather than their peak level, that avoids the 'loudness war'. Dynamic range was inevitably minimised in peak normalised systems to try and maximise loudness.

However, loudness normalisation systems -- which all use a relatively low target loudness --enjoy reasonably generous headroom margins and so inherently encourage the use of a much larger dynamic range.

H

### Re: Mastering level question

Posted: Sun Aug 04, 2019 10:13 pm
Demious wrote:I mixed at k-20 and I have a dynamic range of 6-9dB, measured with BX Meter.

If you're aiming your material for streaming services with a -14 LUFS target, you might find working with the K-14 meter variant more helpful.

Your existing dynamic range is fine from a technical perspective, so I'm wondering why you feel the need to reduce it further in your 'mastering' processing?

Regarding your question about opening the 'gap' between the Integrated and Short-term loudness measurements, this is all about the micro-dynamics of your material. It's determined by the dynamic nature of the music itself and how much compression you have in the mix across individual sources, buses, and the stereo mix two-bus (as well as any more added in 'mastering'.

### Re: Mastering level question

Posted: Mon Aug 05, 2019 12:40 am
Hugh Robjohns wrote:
hobbyist wrote:
Demious wrote:I have a dynamic range of 6-9dB...
That DR sounds just a bit low for my taste...

And yet only a few posts ago we were told:

hobbyist wrote:Good dynamics are those under 10dB for me.

Hmmm... ;-)

Not that much under?! :)

### Re: Mastering level question

Posted: Mon Aug 05, 2019 8:23 am
hobbyist wrote:
Hugh Robjohns wrote:
hobbyist wrote:
Demious wrote:I have a dynamic range of 6-9dB...
That DR sounds just a bit low for my taste...

And yet only a few posts ago we were told:

hobbyist wrote:Good dynamics are those under 10dB for me.

Hmmm... ;-)

Not that much under?! :)

Not as much as 1dB under?!

### Re: Mastering level question

Posted: Mon Aug 05, 2019 9:26 am
Demious wrote:Im pretty new to mastering, so Ive been reading up and watched some video tutorials about levels for streaming services and Im a bit confused about the information I found.

When leveling for Youtube, (-14LUFS) in one video was said to level at -14LUFS average, while another said to level at -8LUFS short term. I understand from this that average should be -14, with the max peaks at -8. However, now Im trying to level, my average and short term are very close together, about 1.5LUFS difference max, so I tried to increase short term, without affecting average, to meet the best vallues. But no matter what I do, when I inscrease peak, my average will follow in the same way, I cant find how to increase the difference between those two vallues.
Am I overlooking something in the mastering process that can increase the difference between peak and average to such extend, or did I misinterprete the given information about these 2 levels?

What James says is important, the track must sound good to you. That said, most often the reason is your arrangement.

Technically, short term and long term loudness are both (more or less) integration of loudness over a time window (i..e the area below your waveform and the horizontal time axis, for a certain amount of time). The only difference is that the window is a sliding one of a few seconds for the short-term, while it's as long as the full track for the final loudness.

If their values always very near to each other, it means that you arrangement is about equally loud everywhere. In other words, you don't have breaks, stops, places where you use less instruments etc.. it's a wall of sound from start to tend.

Now there's nothing inherently wrong with that, but in general people keep being interested when there's some arrangement variation. So mute stuff away, take away instruments, introduce less dense areas or conversely makes certain areas pop more etc - and your short term loudness will change accordingly. Another way to keep interested is to add ear candy - unexpected sounds and effects.. that's used in lots of modern dance productions, for example.

### Re: Mastering level question

Posted: Mon Aug 05, 2019 3:31 pm
Ramirez wrote:
hobbyist wrote:
Hugh Robjohns wrote:
hobbyist wrote:
Demious wrote:I have a dynamic range of 6-9dB...
That DR sounds just a bit low for my taste...

And yet only a few posts ago we were told:

hobbyist wrote:Good dynamics are those under 10dB for me.

Hmmm... ;-)

Not that much under?! :)

Not as much as 1dB under?!

It depends on the type of music.

One size fits nobody. Perfection is not possible.

I mostly listen to classical so closer to ten is my preference for that.
I can see some pop music being lower DR.

### Re: Mastering level question

Posted: Mon Aug 05, 2019 10:17 pm
hobbyist wrote:It depends on the type of music.

One size fits nobody. Perfection is not possible.

I mostly listen to classical so closer to ten is my preference for that.
I can see some pop music being lower DR.

Yes, but you said “under 10dB”. You can’t get much closer to 10 than 9!

### Re: Mastering level question

Posted: Mon Aug 05, 2019 10:29 pm
Classical with a DR of >10dB would be awful, the dynamic range of an orchestral concert is 60dB plus, why would you want to compress that to >10dB?