Question about VSTs, Headroom & playback levels

[sjs94704]

Hi, All:

I understand the concept that if the approximate meter volume of a track is at or close to -18 that "it leaves lots of head room".

Using a single track as an example, I understand that once recorded, I have between -18 & -6 to stack VSTs in order to get the sound I am looking for.

The part that I am struggling with is, just as the pre-VST volume of the tracks should be at or very close to -18, what should the POST VST peak meter level be of each track including the MASTER track?
.
.
.
.

[lilith93]

It really doesn't matter where the levels of your individual tracks are, but it's good to keep them at ~0dB. The only thing that counts is your Master track which should not exceed 0 dBFS (Full Scale). If you upload your track to e.g. soundcloud it's good to leave a little bit of headroom (e.g. ~-1.5 dBFS).

https://www.youtube.com/watch?v=6vXiIlKCEgQ
https://www.youtube.com/watch?v=5efick6yJA4

[DVDdoug]

REAPER uses floating point so there is virtually no upper limit. Your ADC (recording) and DAC (playback and monitoring), "regular" WAV files, and CDs, all hard limited to 0dB.

So you can clip your DAC if you go over 0dB and you are monitoring at "full digital volume". Or if you are monitoring at lower levels you may not get clipping while mixing/monitoring but your rendered WAV may be clipped.

I assume most VSTi's are designed so they don't go over 0dB but effects can push-up the levels and mixing will push-up the levels.

[serr]

Let's put the calculators down for a minute...

-18 rms isn't the target
NOT CLIPPING is the target!
If you're unsure what peaks might come off the microphone you are recording, setting the level for -18 rms is usually safe.

But you're not trying to hit some specific number. You simply don't want to clip and distort.

After the fader, there is no target level whatsoever. You mix to achieve the balance you want to hear. The level you are concerned with is your master output. Again, you simply don't want to clip.

If you have something with peaks that hit zero but it's still not loud enough, that's when you get into the automatic volume leveling devices: compressors and limiters.


The really really simple instruction is: Don't clip.
THAT is your target.

[sjs94704]

Thanks, Everyone!

You know, I love working with Reaper, but, since I am self-taught, not just in the actual DAW itself but in mixing too a lot of the 'small' details can be kind of intimidating when putting them all together, but, I really enjoy the experience of learning and am very grateful to everyone's patients with me as I go about my journey in learning all of this.

In the past few weeks I have been distracted by some major issues with my PC that has caused a lot of distraction and distress, but, I am pretty much back up and running and looking forward to getting back into recording....

I have had some equipment failures and am having to rebuild my system a bit. But, this is a chance to upgrade on a couple pieces of of equipment such as a much better pair of headset! I am waiting for them to be delivered by WED and am excited to try them out!

[Steviebone]

Sound energy and the digital math behind it is cumulative. The more tracks you add together the larger the number and the more energy present at the output.

For this reason, how much headroom you need in a mix depends upon the number of tracks and the type of material you are recording/mixing. Cranking up all of the individual tracks and then lowering the master to compensate is not optimal math.

The headroom required to add 16 tracks together will never be sufficient for a 48 track mix. The key is how many tracks are playing at one time (max) and what their individual signal levels are. So there is no hard and fast rule other than the larger the mix the more headroom you need to leave. Fortunately, with a digital DAW like reaper you can select all as a temporary group and then lower all of the faders a bit together in one fell swoop. Likewise if you get a mix you like at the end and there's still headroom left its a simple enough matter to do the reverse. This is not a reason to avoid proper planning for headroom in your mix though.

PS: If possible, avoid doing a final mix on headphones. Ears need to hear reflections in order for a mix to sound right later in speakers, (unless you are mixing specifically for earbuds).

[ashcat_lt]

Originally Posted by Steviebone

Cranking up all of the individual tracks and then lowering the master to compensate is not optimal math.

But it ain't gonna hurt anything as long as you're not unintentionally overdriving any plugins along the way.

Fortunately, with a digital DAW like reaper you can select all as a temporary group and then lower all of the faders a bit together in one fell swoop.

I usually can't actually do that because I've got folders and busses and "aux" tracks with both pre- and post-fader receives and it could take me minutes to figure out which faders to ctrl-click to avoid actually just fucking everything up, and in some situations it really just can't be done at all. OTOH, I can just grab the master fader, or my "sub master" bus fader, or adjust the gain on one of the plugins on the master track. All of those are not only quicker and easier but actually more precise and efficient.

[RDBOIS]

Originally Posted by sjs94704

The part that I am struggling with is, just as the pre-VST volume of the tracks should be at or very close to -18, what should the POST VST peak meter level be of each track including the MASTER track?
.

If I understand the question properly, for individual tracks, my answer is -18db.

The Pre-Master -6, and the Master (assuming you are mastering the song yourself in Reaper) = -1 to -0.5db

I think it is wise to keep the post-volume coming out the VST the same as the pre-volume coming in the plugin. Why? Because you want to know what the friggin VST is actually doing! This is only possible if the volume is the same going in and out. If you don't do it this way you will never really learn how to dial in various effects efficiently. Some effects are really subtle, much more subtle than perception of volume change.

Volume coming IN the Premaster can be -18 to -6. In the Premaster you have the final chain of mixing effects (normally a bit of compression, a bit of limiting if needed, etc.. What comes out of the Premaster is normally the FINAL MIX that goes to the Mastering processes; it should have headroom (e.g. -6) unless you want to piss off the mastering engineer.

If you want to Master the song yourself, just raise raise the volume in the mastering stage so your peaks don't go above -1db, with perhaps a ceiling limiter at -0.5 db to catch the interpeaks. In the Mastering stage you will listen to you song A/B with other known songs to learn more about your mix. You will listen to your song in headphone, in crappy speakers, in the car, in mono, etc, while comparing with known songs. This is where you have a chance to make some final tweaks to make the song better on all plateforms.

[Bri1]

FINAL MIX that goes to the Mastering processes; it should have headroom (e.g. -6) unless you want to piss off the mastering engineer.

^heh-you know,i actually do not think it pisses these people off-- they just lower the input levels for their 'mastering chains'- and are usually working in floating point anyways..(or consoles)
if artists can send mastering house a 32bit file (or even 64bit!) --all is well_er... lots of *frequency specific* gain can be added,or subtracted,to suit any final 16/24 mastered file.

some plugins i find to respond much better with hotter signals going in>>>> then smaller tweaking increments can make noticable changes.
depends on instrument choices,genre+ personal dynamic 'tastes'. =)

question="what is the lowest volume 1 can hear?" people have their own hearing 'range'.

[JohnnyMusic]

Hello All,
I was reading this thread and it is very similar or directly related to what I am trying to figure out regarding levels in mixing and mastering the get the best sound out of the plugins in the mastering chain. It would be really helpful if any of you could clarify a few things:

-When mixing, for each track, when you -18 going in is a good level, is that peak or rms or lufs? I read that serr says it doesn't matter post plugin chain level as long as you aren't clipping.

-Are you saying that -6 dbfs (peak?) is a good level for a mix going into the mastering chain? And is there an optimal LUFS (integrated loudness) guideline that would correspond to this for optimal level to send into the mastering chain? Do mastering plugins want to see a certain level to work optimally? If so what level (dbfs or LUFS etc)?

-If there is extra headroom left in you mix that you want to make use of (to explain, the mix is peaking below 0 dbfs whatever amount, lets say it is peaking at -6 dbfs and LUFS of -18 , I don't if the numbers are realistic, its just an example), do you want to turn it up with a gain plugin at the beginning of the mastering chain, or if not, where would that gain be made up? In the mastering plugins? At the final limiter after the mastering plugins so that you aren't feeding them a suboptimal levels?

Hopefully these questions make sense, and not trying to hijack, my clarifications seem to be directly related to what you are talking about.
Thanks for any input!
John

[Bri1]

@ JohnnyMusic- m8,let me try to ease your mind while working with reaper..
try this 1 thing ok,then report back your 'findings' if you will, please+ty>

ok:make a mix of your choice,make sure it clips at over 6db on master and render file to 24bit wav.
render same project @32bit wav,same overshoot of db--and reimport both files..
inspect both files--- bring down both item volumes-compare files again.
scratch head--consider 32bit > superior!

point: just totally focus on getting great recording to begin,then a great mix--render to 32bit-and no 'masters' can really complain about levels..

[ashcat_lt]

-18dbFS is usually pretty close to what would be 0dbVU in the analog world. VU meters are usually more like RMS meters. -18dbFS RMS is therefore often used as "nominal" level for plugins where that kind of thing matters.

Most of those plugins are going to have input gain controls, though, and like all things, you should really just turn the knobs until it sounds good. What's the worst that can happen if the levels too high? You'll get distortion. Do you hear distortion? Do you like how that distortion sounds? If not, turn it down. If so, try turning it up and see if you like it even more.

[Bri1]

If not, turn it down. If so, try turning it up and see if you like it even more.

^sticky ! worx4me!

[JohnnyMusic]

I believe you on the 32 bit floating preventing clipping within reaper, so I am not going to do your experiment.
So I don't have to worry about levels, even going into plugins, unless they are designed to see a certain level.
my main question though is this:
So lets say a I have a plug on my mastering chain that wants to see -18 dbfs rms or whatever so I give it that, then it comes out and its still got some headroom left, do I want to gain it up with the output gain before it hits the limiter (either with the output control on the plug, or a gain plug, or the input of the limiter, does it matter how I handle that?
Don't know if that question makes any sense?
thanks!

[Bri1]

^ you have all the answers inside-look there<
people worry so much about so little sometimes--some say listen with your ears-- i advise listen with all the senses.
just do your best to get a balance of instruments/vocals--make sure you do not clip master renders @16/24bit,simplez.
try to enjoy reaper without worries mayyng.
even really distorted and flattened dynamics can be reshaped by a semi_decent process afterwards.. some peaks can even be reconstructed after apparent oblivion !
as we all have a listening range--it's upto the mixer wether we use that whole range,or not.
these days a lot of artists try todo everything themselves,which is fine--but there are well trained peoples in all areas of productions--specialist mixers for eg:

[karbomusic]

-When mixing, for each track, when you -18 going in is a good level, is that peak or rms or lufs?

Roughly RMS and it is that average suggestion because...

1. Most 24 bit sound cards will show 'around' -18 dBFS RMS at their analog zero dbVU point as ashcat pointed out - so the incoming track is already where it should be assuming it used sensible recording levels. There is additional reasoning for this range/number, we can discuss that as needed.

2. You have ample headroom overall for mixing work. At this point there's not much to be concerned with it's just wiggle room, as tracks themselves will sum and mixing will tend to add volume as you go. The main exception is non-linear plugins *and* someone wants a preset to sound like the preset was intended (likely academically) and maybe the vendor said use -18 for that, otherwise, who cares, turn the knobs till you like what you hear.

Honestly, I'd not think much about what to do between -18 and 0 other than don't clip the master. It is somewhat of a self-correcting issue, you make choices, you mop yourself into a corner by clipping the master, you readjust - these are just suggestions to keep us from being too close to the ceiling before we even start (no wiggle room). When you are done, if your are say 6dB short of zero, simply turn up the output at/near the end of the chain. Here's how I do it...

At the very end of my chain is a limiter, the limiter has an input gain knob. If my final mix is peaking around -6 dbFS then I can just raise the gain knob to somewhere close to +6, which means I'm now peaking close to 0 dbFS and if a stray peak happens the limiter will catch it. The amount chosen will be more/at/less than +6 depending on whether I want any additional limiting going on other than overs protection. The input gain can be replaced with any plugin that can increase volume and be placed before the limiter and so on.

[ashcat_lt]

Originally Posted by JohnnyMusic

So lets say a I have a plug on my mastering chain that wants to see -18 dbfs rms or whatever so I give it that, then it comes out and its still got some headroom left, do I want to gain it up with the output gain before it hits the limiter (either with the output control on the plug, or a gain plug, or the input of the limiter, does it matter how I handle that?

It doesn't matter unless it does, and if it actually does, you'll hear it. In most cases, it probably won't.