HOW TO lower the Volume of a project? LUFS are too high

[Caine123]

hi guys.

im wondering what method is "transparent" or better to use or if there is ANY difference at all?

i got a project which is at around -5 LUFS, i dunno how i did it but the Clippin on the Limiter is only ar around -1 GR. still it is too loud cause my other tracks are around -8 LUFS.

how would you go lower? the Limiter is only pushed +1 db!

1. i got in my chain PRO-Q3 as the first instance. shall i just lower the volume here?


2. shall i lower ALL mixer volume faders (the picture is a different example project) without the BUS Drum?


or how would you do it? i want to have the sound like it is and not changing it when pushing the volumes different so the comp etc. work different on master?

thx

[BertKoor]

Export to wav. Grab any ole audio editor. Lower the overall volume by 3 dB. Save as; Done.

[legendCNCD]

Does it sound good still?
I'd go and drop the mix -6dB on all the channels. I hope you did not automate the mixer levels

[enroe]

How would you go lower? ...

Yes, that's of course a difficult question - and it sometimes
happens that a track is simply too loud and you can't easily
turn it down again.

But: There is a top secret, well-kept trick. Mmmh, should I really
reveal it? So, we help each other here, and that's why I'll tell you
this fantastic trick:

Turn down the master control a little!

But don't give it away!

[lobanov]

If there is no clipping (the master channel and individual channels don't go red) don't bother. LUFS level is a goal for the mastering stage. Mastering deals with one file (a whole exported song or, better, songs in an album). It's not plugins on the master channel as it is said erroneously sometimes.

If your project clips lower all channels, don't touch the master channel. Also you may use any gain plugin. In FLS it's really easy to save a preset (for example, -6 dB) and then load it on all the active channels. But if there is no clipping don't bother.

[Caine123]

thanks a lot to all the replies!

i have to add some more infos .

first i found this video about gain staging


seems i did mix a different way so far. in the past i used a lot the channel rack volume knobs or adjusted in the VST themselves the volume. then i added effects in the mixer channel and after that adjusted some volumes in the mixer. seems i did that the correct way.

but i thought the channel rack might be just another volume place possibility and used more the mixer channel volume faders FIRST.
the video showed however that it is not the volume BEFORE the effects. i adjusted WITH effects on the volume fader in the mixer channel the fader to around -9db and so on.

second. so which way would be the better one?
i guess having without effects like he states in the video adjusting the kick volume at around -9db and adding the rest around it or like i did later, make the beat ready with all effects and then put all volume faders in the mixer channel to 0 and then raise each element step by step? (Master bus is dry then).

[sqigls]

i try to keep all faders at unity and adjust the volume of all tracks on the actual VSTi.

I put a VU meter on my master, select my reference level at whatever dBFS i'm aiming for.
depending if i'm giving a shit, i generally have one VU pre FX and one post.

I then get my bass and kick bending the needle to its desired area of heaviness... continue from there.
i keep my faders at or around unity, by habit. If only for the resolution/sensitivity factor.
Also cos so many soft synths and drums actually clip in their initial state.
If you just drag all your faders down instead of adjusting the volume of the original source, you hit the effects so hot - every channel is overcooked. Sometimes that's the desired effect, but i'd mostly prefer to keep things clean and dynamic.

[zerocrossing]

I don’t know if this is the right way, but I just always have a limiter before Youlean Meter, and I just turn down the output until I get the reading I need.

[aciddose]

You need to understand which signal types have high RMS/loudness, and which are low. This is easiest to understand if you start at the most basic level, with very simple waveforms. A sine waveform has an RMS value per cycle of sqrt(1/2) (0.7071~) but a square wave at the same peak level has an RMS of 1. A triangle or ramp has an RMS of sqrt(1/3).

So we can see the RMS/peak/loudness values are very different for a variety of waveforms. Suffice it to say, if you use a low-peak and/or low-RMS signal and put that into a limiter with input gain, you'll most likely dramatically increase the loudness by maximizing the peak, since loudness is going to increase along with the peak level (an addition of +3 dB of peak level will likewise increase loudness, in general, by approximately +3 dB.)

So most likely this particular track has a combination of signals that have low peak levels and high loudness. As a result, if you measure the loudness using a system like LUFS, you'll also get a very high value. In such a case, driving into a limiter or applying any gain whatsoever serves absolutely no purpose. You'll want to simply linearly reduce the level of those channels/signals during mixing which contribute most to the overall loudness. It's very easy to make the mistake of maximizing peak/RMS/loudness levels using peak/RMS limiters when this is in fact wholly detrimental to a good mix.

I personally use monitor-effects which include dynamics processing and EQ. This is so given the monitors/headset I'm using, a "perfectly mixed" track sounds ideal to me personally with the processing enabled. That same track might sound wrong, not enough bass, not loud enough to my ears using the equipment I'm listening with, but the goal is to mix to produce another track that very closely matches this "perfection", which is in reality a healthy middle-ground for loudness and EQ.

That strategy of processing all content allows me to produce mixes and outputs that are correctly balanced. So my compressor/limiter is only applied to what I hear, and not directly to my mix. The mix is set up so the overall levels and loudness don't activate the compression/limiting too much, and flicking the monitor effects on/off to compare still sounds good to me and matches other "perfect" mixes very closely with processing disabled.

[Caine123]

hi guys thx a lot again, so you mean you should always (of course besides the effect/distortion if desired) control the volume FIRST in the channels or vst itself?

(Left)


instead of the mixer faders? (Right)?

strange thing is i watched a lot of mixing tutorials and a lot of guys told that you can go by lowering all mixer faders to 0 and raise each element until you are happy then you start mastering.

what is the best way to level your volumes?
to tell you the truth i always thought channel volume/VST Volume is the same as the mixer volume. not that the effects chain is effected differently. wow. awesome
i mostly always started to put all elements until i was like ok now i need a structure, i effected everything already and raised/lowered the volumes mostly already in the channels/vsts but sometimes if the mixer was first open then used a mixer fader.

so it seems the way is this? all mixer channels without effects or if you use effects do you dial back the volume to the initial volume? e.g. you use a kick sample and balance it to -12db to start. then you enable effects and are at -5 db then, do you dial back e.g. the comp to get to -12 back again?
do it on all elements and then start using mixer faders to balance the track a bit further out?

also here in the fl studio video he uses also the mixer faders all setting to bottom and dial in each element (with noise as reference) and not with channel volumes? or maybe he did but didnt mention it clearly, cause i see on some parts the volume is a bit different on the channel volumes

[BertKoor]

This is roughly how it's done on an analog console.

First thing to do is set the Gain/Trim knob of all channels, so you have a healthy level before it hits further processing in the mixer (EQ, Aux/FX send, DirectOut for a multi-track recorder, etc.) A too soft level will have a bad signal/noise ratio, a too loud level will distort.
For this purpose there is a Gain/Trim knob at the top of the channel. You put the channel on Solo, so the main VU meters show its level and only that level. Adjust Gain/Trim so when the instrument is playing loud it does not clip (high in amber zone of the VU bar.) An ideal volume is where the Gain/Trim knob sits at the 0dB - Unity Gain position (usually 12 o'clock)
If you need to turn up the gain far above the 2 o'clock position, you yell at the dude at the instrument to turn his thing's volume up. If he cannot, you'll have to gain it beyond 2 o'clock for him, no problem.
When the incoming level is way too loud, use the "Pad" button next to the Gain/Trim which can shave off some 20dB of the signal. If it then still is too loud, you yell at the dude at the instrument to turn his thing's volume way down.

(leaving out the process of EQing of the channels, which might cause you to adjust the gain/trim again)

So all the channels come in at roughly the same level. Now it is time to mix. Pull up some important channel faders to around -10dB, adjust all to taste. The level of the main bus is sum of the channels. A handful of channels at -10dB will sum to roughly 0dB, which again is a healthy level in the amber zone and not clipping. It is not uncommon to have the fader of the lead vocalist way above 0dB. Maybe the total level on the master bus is then a bit hot, but with the master bus fader you can compensate.
Also a handy tool here is the grouping of channels. Route all channels of the drum kit to a group, and the group has its own fader to mix into the master bus. Now there is one fader for mixing in the drums, so you don't need to grab all 10 faders of the kit and destroy the balance between them.

Mastering though is an entirely separate process. The whole mix is recorded as a stereo track. The mastering engineer gets only this stereo track and sprinkles on his fairy dust to make it sound even better.


Now in a DAW. There are again a lot of places to adjust volume. Since the audio is represented by digital floating point numbers, it does not really matter which one you use. Noise and distortion are not an issue. You can run all channels way too hot or way too soft, and compensate with the master volume fader. It only is a concern for some effects that are dependent on the level they get: compression, limiting, saturation, distortion.

So you can put all faders at 0dB and adjust at the instrument. The consequence is your master fader will be set rather low if you don't want it to clip. Plus, if you are mixing this way, you're not using the mixer for what it is intended for: setting the level of individual tracks.


Still confusing? Probably yes...

[BertKoor]

Should I have added... in the analog world the reference level is zero dB VU. Clipping is to be expected somewhere above +18 dB, could be +26dB on gear with lots of headroom.

If you mix like that in a DAW, you aim at ca -20 dB fs (20 decibels below the "zero Full Scale" clipping point) and your peaks can be anywhere from -2 to -14 dB fs depending on the material. That is before any compression / limiting, that's the job of the mastering engineer.

Those are the practices from before the loudness war, but this is where we come from and how things were intended.

[aciddose]

hi guys thx a lot again, so you mean you should always (of course besides the effect/distortion if desired) control the volume FIRST in the channels or vst itself?

Typically, while mixing I find it useful to have -3 dB on a fader correspond not to an exact level of -3 dB, but somewhere approximately a constant "loudness". So for instance a guitar track might have a higher input level than that same track processed with distortion, because the distortion will typically make the peak/loudness ratio closer to 1 to 1 (like a square wave.)

So with all faders set to 0 dB (no effect on level), I adjust the "input gain", a combination of the instrument and input preamp levels you'd normally find on a console mixer. The preamp gain knobs are generally present because different instruments or microphones will have radically different peak output levels (in volts) and we want generally a neutral setting to produce either equal peaks (0 dB peak meters) or equal loudness before we adjust faders and start mixing.

It really depends upon the exact situation which method works best. You usually do not want to see faders set to -30 dB or other extremes. You want somewhere 0 dB to -18 dB, something like that, similar to the common range of a VU meter over 0 dB. Every mixer will have faders with a "sweet spot" (actually a range) that are different, so specifying exact values makes zero sense outside an exact context.

Likewise, different DAW mixers will have different faders and curves. So you have to come up through experience with what suits you best. Personally, I note that two things are negatives for me:

1. All faders sitting at the same point for tracks with radically different levels.
2. All faders sitting at vastly different points like +40 dB or -40 dB pushed to the very edge of their ranges.

So I typically adjust input gain for each track/channel to get the peak levels and loudness sitting somewhere in that "sweet spot", then when I mix and adjust the level of different tracks I end up with a blend of fader settings like -18 dB to 0 dB that roughly correspond to the loudness of each track, so there are some good visual cues.

One good tip might be that if you have a bunch of tracks at -12 dB or so and then one (like a flute, or a bell, or similar) that needs -28 dB to sit right in the mix, reduce the input gain or instrument level on just that track. You might reduce the gain by -12 dB, and that would place the track fader at -16 dB instead of -28. Most likely everything will then seem to sit at relatively visually matched positions for the levels.

[Caine123]

hi, i did this one on a laptop cause im still not back yet :/ so here i first switched off all effects, only the limiter is on.
i reset all mixer faders and put all elements in the channel volumes to 0db. then i raised the volume of each element until i found it pretty nice balanced. then i turned on the effects and adjusted these volumewise back so it matches nearly the volume i first dialed in (would you do it this way? e.g. you have without effects the volume at -9 db then you put in EQ, comp, saturation etc. and dial the volume back so it is at around -9db again in mixing stage?). then i made a quick mastering adjustment with eq, comp, clipper and more limiting until im happy and then i used the mixer faders to balance it slightly at the master stage.

https://www.dropbox.com/scl/fi/r656vflu ... 4xsjj&dl=0

[aciddose]

How did that turn out with your end result LUFS measurements? Are they sitting now at more typical average/windowed -12 to -18 dB LUFS levels with peaks near -3 dB?

A master limiter can throw you off a ton too, so ensure during this "gain staging" stage of the mix that you're not seeing it activating gain reduction at all if possible. I find it personally easier to not apply master effects like a limiter at all and rely upon a clip indicator instead to show me where I've messed up. Obviously if you have a mixture of inputs scaled correctly, the mixed result won't ever exceed 0 dB.

As I described, my monitor effects include a limiter, headphone filter (where using such) and EQ and I often switch these on/off to compare to the raw sound. Since modern processing is all in floating point anyway, it's entirely possible to see +200 dB peaks and the monitor effects limiter (the master's master!) will safely catch those peaks before they are output into any filters or analog domain.

For example, listening to a "hardcore" genre electronic track with high loudness played through my master effects, the peak level the limiter is seeing is -4.1 dB. This is after "loudness" boosting low/high shelf EQ, AGC, dynamics compression, bass harmonics, headphone filter and high-pass filters. So the actual output signal levels at peak loudness that my "DAC" or analog systems will ever see are safely far below -3 dB and way far from 0 dB.

With harder pumping hard kick house genre tracks, I can see -2.5 dB peak levels at the limiter.

This track which is a somewhat more amateur mix:


Gets as high peaking as -1.7 dB ... yet still the peak limiter's gain reduction never activates at all.

This "Otherworldly Foe remix" track seems to have the author's peak limiter pushed at exactly -3 dB threshold, then it rarely produces peaks in excess of the limiter due to the further processing stages applied by youtube or the audio codec. This is obviously a "failed" mix, as the signal levels pushing hard into the limiter at -3.01 dB (technically codec signal averages around -2.95 dB) result in the potential for clipping on "intersample peaks" due to later stages in the digital or analog signal chain.