question about the -18 dB level

[Giusmex]

hi guys,

given that i recently purchased slate digital vcc, and given than now i want to be "wiser" in my songs, i decided to pull down the volume of my tracks; slate digital vcc for example is calibrated so that its 0 level corresponds to your -18 dB fader volume ( in general)

Now, my question about that is:

Has -18 dB to be to loudest peak in a track or does it simply represents the value at which you should posiiton your fader meter?

because in the first case i imagine that the volume of that track shoud be controlled by the output volume of the virtual instrument itself..

and, after that, the master output volume? how is that related to the "-18 dB" thing?

could you please shed some light to this subject for me please?

thanx

[fhunktion]

I don't own slate but -18db is probably an rms (average) level so you should aim to be hitting -18 on average - this gives headroom for peaks. search for rms on google

[naturestoned]

This -18 dB thing is about AVERAGE levels (18 dB RMS), so typical peaks for
non-percussive material could be around -8/9dB with maybe some transients up to -6dB at max. The best way is to track already with this right level. Lots of the better AD-converters are also calibrated to this 0dBVU= -18dB RMS.

If you or others tracked hotter which also might be on purpose because of the specific sound of your devices, TRIM those levels down in the DAW to -18 dB RMS already in the insert FX chain BEFORE all following FX plugs and leave the fader at 0 (e.g. with tools like VUMT from Klanghelm or the free MonoChannel/StereoChannel from SleepyTime Records). Then during mixing also the fader will be in a range with much better resolution for finer adjustments.

You can check the levels also within VCC with the VU-Meter on top (shoot around 0dB, the beginning of the red) slight corrections to the input level (-6dB - +6dB) could be made with the INPUT knob without further plugs if you plan to use VCC Channel on all tracks. After that I would recommend that you use the DRIVE knob to consciously drive the consoles that VCC models, this works with gain compensation so you really only hear the sound difference and not just a louder signal, which always sounds better.

In the end your sound will be much better, more open transparent.

Google or search for "Gain Staging" if you want to know more about that whole concept.

A lot of other plugins are also calibrated to this -18dB RMS level but attention, some are not... they expect a different level to sound right or are also equipped with settings to calibrate them (e.g. Thrillseeker LA, Slate VTM, etc..). This is especially relevant if there is some kind of input gain dependend behaviour in the plugin like saturation.

[Compyfox]

Read: my KVRmarks

[pekadan]

@naturestoned: Wow, that's one excellent answer!

[AudioGuy720]

hi guys,

given that i recently purchased slate digital vcc, and given than now i want to be "wiser" in my songs, i decided to pull down the volume of my tracks; slate digital vcc for example is calibrated so that its 0 level corresponds to your -18 dB fader volume ( in general)

Now, my question about that is:

Has -18 dB to be to loudest peak in a track or does it simply represents the value at which you should posiiton your fader meter?

because in the first case i imagine that the volume of that track shoud be controlled by the output volume of the virtual instrument itself..

and, after that, the master output volume? how is that related to the "-18 dB" thing?

could you please shed some light to this subject for me please?

thanx

In the future RECORD at -18 dB RMS levels as well instead of reducing them after the fact. It'll make your recordings sound better, especially if you're using budget gear.

[Giusmex]

hi guys,

given that i recently purchased slate digital vcc, and given than now i want to be "wiser" in my songs, i decided to pull down the volume of my tracks; slate digital vcc for example is calibrated so that its 0 level corresponds to your -18 dB fader volume ( in general)

Now, my question about that is:

Has -18 dB to be to loudest peak in a track or does it simply represents the value at which you should posiiton your fader meter?

because in the first case i imagine that the volume of that track shoud be controlled by the output volume of the virtual instrument itself..

and, after that, the master output volume? how is that related to the "-18 dB" thing?

could you please shed some light to this subject for me please?

thanx

In the future RECORD at -18 dB RMS levels as well instead of reducing them after the fact. It'll make your recordings sound better, especially if you're using budget gear.

hi audioguy thanx for the answer

well i only use virtual instruments so no problem for the record level

[Giusmex]

This -18 dB thing is about AVERAGE levels (18 dB RMS), so typical peaks for
non-percussive material could be around -8/9dB with maybe some transients up to -6dB at max. The best way is to track already with this right level. Lots of the better AD-converters are also calibrated to this 0dBVU= -18dB RMS.

If you or others tracked hotter which also might be on purpose because of the specific sound of your devices, TRIM those levels down in the DAW to -18 dB RMS already in the insert FX chain BEFORE all following FX plugs and leave the fader at 0 (e.g. with tools like VUMT from Klanghelm or the free MonoChannel/StereoChannel from SleepyTime Records). Then during mixing also the fader will be in a range with much better resolution for finer adjustments.

You can check the levels also within VCC with the VU-Meter on top (shoot around 0dB, the beginning of the red) slight corrections to the input level (-6dB - +6dB) could be made with the INPUT knob without further plugs if you plan to use VCC Channel on all tracks. After that I would recommend that you use the DRIVE knob to consciously drive the consoles that VCC models, this works with gain compensation so you really only hear the sound difference and not just a louder signal, which always sounds better.

In the end your sound will be much better, more open transparent.

Google or search for "Gain Staging" if you want to know more about that whole concept.

A lot of other plugins are also calibrated to this -18dB RMS level but attention, some are not... they expect a different level to sound right or are also equipped with settings to calibrate them (e.g. Thrillseeker LA, Slate VTM, etc..). This is especially relevant if there is some kind of input gain dependend behaviour in the plugin like saturation.

thanx a lot naturestoned for your answer


unfortunately, i am kind of newbie so for me it is difficult to understand what you exactly mean; taht is to say:

1) my "normal " cubase faders are dBVU, not RMS, right? so you are saying that i should set the faders' levels at o dB because it corresponds to -18 dB RMS ?

2) in this case, which are the maximum peaks that i should reach with drums tracks? you referred to -6dB maximum for non drum parts; you mean -6 dB RMS or -6 dB VU ? i dont catch this point

3) if , lets say, i use slate digital vcc on EVERY track, i can use it to check for the 0 VU level i can do that with th input knob, but can i also do that simply with the mixer fader ?; so, every track should not exceed 0 VU on the meter of vcc?


excuse me but i am a beginner in computer music


thanx a lot

[Compyfox]

1) my "normal " cubase faders are dBVU, not RMS, right? so you are saying that i should set the faders' levels at o dB because it corresponds to -18 dB RMS ?

No, your digital faders are digital faders. The meters in your DAW usually correspond to dBFS. Stop confusing dB VU (dB in relation to VU) with RMS, and with digital full scale meters.

A VU (Voltage Unit) needs specific settings: ballistics, reference level, specific tagging of the GUI. Same with the RMS meter. Though both(!) are usually setup to 300ms response time (rise and fall of the meter needle/bargraph).

I wrote tons of posts in terms of how to use your host properly with a reference level of -18dBFS. One is the one with the Slate VCC in my KVRmarks.

2) in this case, which are the maximum peaks that i should reach with non -drums tracks? you referred to -6dB maximum for non drum parts; you mean -6 dB RMS or -6 dB VU ? i dont catch this point

Again, as RMS is basically VU - stop confusing yourself.
We talk about VU/RMS compared to dBFS (digital full scale peak).

Since I don't think that ANYONE will dig into my archives, I have to write it again...


Let's assume our VU meter (or RMS meter) is setup to -18dBFS as reference level, and has a ballistics of 300ms (VCC should have this at default).

You should level in your signal so that the average level of your signal (bass intensive material) should not(!) exceed 0VU. The digital peak of your signal should however not exceed -9dBFS on a per-channel basis.

Again in short:
Bass intensive material: not higher than 0VU
Transient intensive material: not higher than -9dBFS digital

Faders at "unity" (default position).

A combination of both a VU meter and digital meter while recording and mixing is essential.



I think it's about time I write/finish a series with the very same topic on my techblog.

3) if , lets say, i use slate digital vcc on EVERY track, i can use it to check for the 0 VU level i can do that with th input knob, but can i also do that simply with the mixer fader ?; so, every track should not exceed 0 VU on the meter of vcc?

VCC is usually setup "pre-fader" (an insert pre fader). So in order to level in your signal properly, you should use an input gain knob (Cubase) or use a gain plugin prior to VCC (Logic) to compensate the only +/- 6dB range of VCC.

excuse me but i am a beginner in computer music

Nothing to say against that - but this topic is covered over and over by now.

[Acrobat]

In the future RECORD at -18 dB RMS levels as well instead of reducing them after the fact. It'll make your recordings sound better, especially if you're using budget gear.

why this sounds so wrong to me?

[Giusmex]

hi compyfox thanz for the answers

you say :

Let's assume our VU meter (or RMS meter) is setup to -18dBFS as reference level, and has a ballistics of 300ms (VCC should have this at default).

You should level in your signal so that the average level of your signal (bass intensive material) should not(!) exceed 0VU. The digital peak of your signal should however not exceed -9dBFS on a per-channel basis.

Again in short:
Bass intensive material: not higher than 0VU
Transient intensive material: not higher than -9dBFS digital

Faders at "unity" (default position). "


i have read some of your interestinf KVR remarks, some things i understood, some others not ( of course, otherwise i wouldnt be a beginner)

ione thing i understood is that i need a VU meter and a RMS meter, only this way i could correctly compare the two worlds; do you suggest any plug in that do that?

and why do you "separate" the bass intensive material from the transient material ( in your above example), and furthermore you refer to 0 VU for the first and to dBFS in the second?

and faders at unity what it means? 0 dBFS?

[Compyfox]

ione thing i understood is that i need a VU meter and a RMS meter, only this way i could correctly compare the two worlds; do you suggest any plug in that do that?

A VU with the same setup as a RMS meter is a RMS meter. Both measure the average loudness of a signal - just one uses a needle, the other one a bargraph. A digital meter measures the maximum peak of a signal.

While recording and mixing, I always suggest going for a VU/Digital Peak meter combo.

and why do you "separate" the bass intensive material from the transient material ( in your above example), and furthermore you refer to 0 VU for the first and to dBFS in the second?

The RMS meter measures "average" loudness of a signal. 300ms responds stronger to bass intensive signals (example: bass and pads) than transient intensive material (example: snare drum).

You pretty much use two different measurement tools to setup your signal, rather than trusting one measurement tool alone.


If I talk about 0 VU, I usually always mean -18dBFS as reference level. So a bass intensive signal should hover around 0VU (measured with a VU!), or -18dBFS digital for that matter. A transient intensive signal should not exceed -9dBFS digital.

and faders at unity what it means? 0 dBFS?

"Faders at Unity" (or "unity gain" for that matter) mean "default position". If your fader (DAW) is at default 0dB, then it's at the unity position.

[Giusmex]

ok thanx compyfox now i am beginning to understand.

YOu say : "A VU with the same setup as a RMS meter is a RMS meter. Both measure the average loudness of a signal - just one uses a needle, the other one a bargraph. A digital meter measures the maximum peak of a signal.

While recording and mixing, I always suggest going for a VU/Digital Peak meter combo. "

in this moment i only have slate digital vcc as vu meter; does it behave ALREADY as a RMS meter or it needs some adjustments to behave as an RMS meter? and in this case, which rms meter do you suggest? i only have izotope alloy 2's RMS meter, but it is embedded within the plug in itself, so it is not suitabale for the initial part of gain staging, is it?


You also say:

" If I talk about 0 VU, I usually always mean -18dBFS as reference level. So a bass intensive signal should hover around 0VU (measured with a VU!), or -18dBFS digital for that matter. A transient intensive signal should not exceed -9dBFS digital. "

In this case, to which value does -9dBFS correspond in terms of VU meter?
and, mor eimportantly, it means that a snare, or a kick can exceed 0 VU?




And , anyway, even if i calibrate this values for bass intensive or transient intensive material in my song, as you say i should do that in the iNITIAL part of gain staging, such as using slate digital vcc as prefader; but, if i then add other plug ins in the same channel ( for instance...1) slate digital vcc; 2) izotope alloy 2 ; 3 ) a compressor...and so on....

at the end of this process, i imagine that the reference levels you mentioned should still be respected; if these plugins ( inserted AFTER slate digital vcc) someway alter ( increase..or decrease ) the VU value of that track, what to to then?


thanx and excuse me for the many questions, but now im getting more and more interested into that

[Compyfox]

ok thanx compyfox now i am beginning to understand.

It's not hard to wrap your head around it, once you understand the difference between measuring average signal strength and absolute (digital) signal strength.

in this moment i only have slate digital vcc as vu meter; does it behave ALREADY as a RMS meter or it needs some adjustments to behave as an RMS meter? and in this case, which rms meter do you suggest? i only have izotope alloy 2's RMS meter, but it is embedded within the plug in itself, so it is not suitabale for the initial part of gain staging, is it?

The VU is already properly setup in VCC. On the setup panel, you can do minor adjustments, but default is more than suitable.

If you want to use an easy to understand VU, I recommend to take a closer look at SleepyTimeRecords or Klanghelm. RMS meter, pretty much everything works that can be setup to your neds.

In this case, to which value does -9dBFS correspond in terms of VU meter?

I throw in another meter, just to confuse you.

Nah, seriously... the -18dBFS reference level and -9dBFS headroom limit come from old Peak Programme Meters that only worked with 10ms or even 5ms response time compared to digital meters which react on a sample basis. It's a savety headroom for measurements.

Want to read more about it, go to Wikipedia and read up on Peak Program Meters and VU Meters.


If I talk about -9dBFS as absolute maximum, I mean the digital meter. The VU wouldn't be fast enough to measure that.

and, mor eimportantly, it means that a snare, or a kick can exceed 0 VU?

In theory, yes. But in reality it won't happen. At least not for a snare drum, since a kick drum can have intesive lowend which is easy to level in at 0VU. Unless you strip it off all this information.

A snare drum however wouldn't barely reach 0VU on the readout of the VU, but on the digital meter, you can clearly see that it can (and will) exceed -18dBFS. In this case, level in at -9dBFS maximum (or lower) and you should be good to go.

And , anyway, even if i calibrate this values for bass intensive or transient intensive material in my song, as you say i should do that in the iNITIAL part of gain staging, such as using slate digital vcc as prefader; but, if i then add other plug ins in the same channel ( for instance...1) slate digital vcc; 2) izotope alloy 2 ; 3 ) a compressor...and so on....

at the end of this process, i imagine that the reference levels you mentioned should still be respected; if these plugins ( inserted AFTER slate digital vcc) someway alter ( increase..or decrease ) the VU value of that track, what to to then?

I compensate the input/output gain of each plugin.
Plain and simple.


Example:
Signal -> Input gain (cubase) -> Klanghelm VUMT (since I trust that one more) -> Slate VCC -> EQ -> compressor -> Fader


My signal is analysed by Klanghelm's VUMT and Cubase digital meter, so that the average signal strength doesn't exceed 0VU or -9dBFS digital, depending on the provided signal. It's then ran through VCC, no need to to change anything, the signal strenth is still the same.

I then run the signal into an EQ - here I do my sound shaping, but do no volume compensation. Don't feel the need actually. It's different with the compressor however. Here I try to shoot for the same signal strenth on output, as it was applied on input. This is something I do by ear.

Now the signal is routed to the fader/pan poti. All I do here is place the signal where I want it to, and attenuate it to fit to the rest of the mix, while not exceeding 0VU or -6dBFS on the summing bus.

thanx and excuse me for the many questions, but now im getting more and more interested into that

Dive a bit through Wikipedia, RANE Pro Audio Reference and GearSlutz (the Skip Burrows' "why ITB mixes don't sound as good anymore" thread). Maybe even read a manual of an external mixing console. It should come to you eventually.

[Keith99]

I too have been trying to get my head around this and thanks to Compyfox posts I am getting there. Sorry to butt in but just a couple of questions:

1. You say dont exceed -9dBFS on transient rich material but then later you say to not exceed -6dBFS on the summing bus. This is the first I had heard of -6dBFS. Do you aim for a max -6 on all buses?

2. I have tried following these guides in Cubase 7 with the new control room meter and I find it incredibly difficult to get RMS up while keeping the peaks contained. I guess this is what mastering engineers get paid to do. Are there any tricks to getting a high RMS and still reduce in the peaks? I know compression does that but I would need to slam sounds to achieve the correct RMS and contain the peaks.