LUFS and compression

[Aaron Straley]

When I get close to finishing a song and get a rough mix, it's usually no where near -11LUFS, which is the level I want to achieve for streaming.
So usually my mix will end up somewhere between -16 to -20 LUFS. Then I'll try to jack it up with compression on the channels that need it. Also maybe some slight compression on the master track. This changes the entire mix and then I need to go back and eq everything again an re-adjust the levels on each track.
Does it make sense to get the proper compression, FX, and level on each track before I EQ each track and set the levels for the entire mix? How do you guys do it?
Is there a general rule on what types of compressors to use for each instrument? I currently only use Reacomp and the 1175 compressor in Reaper. Do I need other compressors, or should these cover it?
What LUFS level do you try to achieve and do you ever have issues getting to that point without having to use compression?

[Kwackman]

When I get close to finishing a song and get a rough mix, it's usually no where near -11LUFS, which is the level I want to achieve for streaming.
So usually my mix will end up somewhere between -16 to -20 LUFS. Then I'll try to jack it up with compression on the channels that need it.

If your mix sounds good and is -16LUFS, why not just turn up the master out (either by the fader or a gain plug-in) by 5dB?
Same mix but now -11LUFS?

[Aaron Straley]

When I get close to finishing a song and get a rough mix, it's usually no where near -11LUFS, which is the level I want to achieve for streaming.
So usually my mix will end up somewhere between -16 to -20 LUFS. Then I'll try to jack it up with compression on the channels that need it.

If your mix sounds good and is -16LUFS, why not just turn up the master out (either by the fader or a gain plug-in) by 5dB?
Same mix but now -11LUFS?

It will clip and I need to compress anyway. From what I can tell, you cant just get clean gain without changing the dynamics of the mix. I want my overall LUFS to be closer to -11 before I start mixing individual track levels

[James Perrett]

When mixing you should ignore the LUFS readings. All you need to worry about when it comes to the master level is that it isn't clipping. Just make your mix sound as good as you can. If the mix ends up at -20LUFS then that's fine - and probably around the level that I would expect.

Once you have a mix that you are happy with, you can either send it off to a mastering engineer or leave it alone for a week or two. If you opt for the latter then at least you can come back with a fresh set of ears.

If you are self mastering then you can probably do most of the things you need with ReaLimit, ReaEq and ReaComp. If you compare your mix to others in a similar genre you may want to use ReaEq to match the sound of your track to the reference. If you feel it needs a little glue to bring things together then ReaComp can give you that - it also does parallel compression which can be useful for some songs. And finally you can use ReaLimit to bring the level up. I would say that -11LUFS is fairly high and you aren't going to achieve that without compromising something.

Of course, there is quite a bit more to mastering and there are quite a few SOS articles on the subject so it may be worth digging around and doing some reading up on the details.

[Drew Stephenson]

Assuming we're talking about an integrated or programme loudness measure
-11LUFS does seem a bit high for streaming. Every service is going to turn that down, some by quite a bit. You can see how much by having a look at the Loudness Penalty site https://www.loudnesspenalty.com/

If I've kept an eye on my gain-staging throughout the project I'll generally find that my un-mastered mix will sit between -20 and -16LUFS depending on the kind of song it is. I'll generally aim for a mastered* mix of around -14LUFS but again this will depend on the song. 2-3LUFS is generally fairly easy to get with a little master bus compression, potentially some clipping, and some transparent limiting.

Ian Sheppard advises not being too fussed about the integrated LUFS measure and focusing more on the short-term maximum figure - if you're going to focus on any number that is - and he suggests a STM of around -10LUFS is generally going to be in the right ballpark.

* I do my own "mastering" so it's not really what you'd get from a proper mastering engineer like James.

[Aaron Straley]

Thanks, I need to read those articles and learn the mastering process better. compose a check list and order of operations.

My mixes are always low, then I need to get more level from somewhere. After applying gain and compression to the master, seems that all tracks need eq'd again and levels adjusted. Why not get the individual track levels compressed and to the correct level before trying to add gain or compression on the master track?

What is a generally accepted average LUFS level that works well with all streaming platforms....what do you guys shoot for?

[RichardT]

The streaming platforms do seem to be coming more into line with each other as time goes by.

Spotify, for example, will now bring anything hotter than -14 LUFS down to -14 LUFS. They will raise anything quieter than -14 LUFS until it reaches -14 LUFS or -1 dBFS whichever comes first. No limiting is applied. This is for individual tracks. For albums they retain the relative volume of the tracks.

This means the dynamics of the original are now preserved.

The only reason for going above -14 LUFS is if you want a particular sound, IMO. It won’t make your track louder.

[Drew Stephenson]

My mixes are always low, then I need to get more level from somewhere. After applying gain and compression to the master, seems that all tracks need eq'd again and levels adjusted.

Applying gain to the master won't change the tonal balance. If you're finding you want to re-eq after applying gain then you're probably hearing the effect of way that our ears are naturally more sensitive to different frequencies. (Look up Fletcher Munson curves if you're not familiar with them already: https://en.wikipedia.org/wiki/Equal-loudness_contour )
If that's the case then the simple check is to switch your monitoring levels occasionally when you're mixing.
Mixing at lower levels can be great for highlighting when certain elements are poking out of the mix but it is important to listen louder sometimes as well.
Don't go overboard though. As well as leading to poor mixes, doing too much at high volume can permanently damage your hearing.
Listen up and down occasionally but always come back to your reference levels.
Compression, on the other hand, might well change the tonal balance of the mix, especially if you've got a lot of low frequency content triggering the compressor.

Why not get the individual track levels compressed and to the correct level before trying to add gain or compression on the master track?

Well, ignoring for the moment the idea of 'top-down mixing', this is absolutely what you should be doing. Get your mix as good as you can before any kind of mastering stage (whether it's self-mastered or external).
Ideally mastering should be about checking with a second set of ears (and a full-range system) and a little bit of gain.

[ConcertinaChap]

I can only say for the sort of music I record, which is acoustic, aiming for a "natural" sound. My mixes generally end up around -16 LUFS and I'm very content to leave them like that. Your mileage will very likely vary.

Why not get the individual track levels compressed and to the correct level before trying to add gain or compression on the master track?

I agree with Drew. Spot on!

CC

[Hugh Robjohns]

As has been said, 11 LUFS is roughly 3dB too loud for all the major streaming platforms, but if that's really your target the little available headroom means your track will have to be very dense and with little dynamic range.

The only practical way to achieve that is to squash pretty much every source element during the mix build. if you leave all the dynamic reduction to a final mastering process of a dynamic mix you will suffer all the problems you've already highlighted.

If you want a squashed mix, you need to build a squashed mix, rather than build a dynamic mix and squish it afterwards.

And use the LUFS meter as you mix to make sure you're heading in the right direction, using the Short-term or Momentary scales.

[RichardT]

I guess an important question is - why do you want to target -11 LUFS?

[Zukan]

First and foremost, clean every channel using BP to remove redundant and problematic frequencies. Then gain stage each channel sensibly leaving ample headroom for gain dependent processes. Aim for a LUFS target if you want to but pay more attention to what is happening short term, momentary, dynamic range and TP. If you are over or under the desired LUFS target don't let it stress you. This can be fixed very easily.

The mix itself and how the dynamics breathe is far more important than a selected target.