Skip to main content

Cubase 11. Kontakt 6.

Orientation: I am in the process of building a very large, mostly orchestral template in Cubase 11 with Kontakt 6, Halions, EW PLAY, Spectrasonics GUIs and more. for use is a large scale project. The Template has about a thousand tracks and will be used to pull ready made presets for largely (but not exclusively), orchestral projects, that will be part of gthe same production and require similar levels. Live in theaters in London. I am not composing with the intent to go to a live orchestra the tracks will be the end result.
The template is largely finished. I am now in the process of balancing the instruments. The structure of the template is folderized and grouped. Each instrument type has its own folder. Flutes for example number 16 in count. All flutes outputs are then sent to a Flutes Cubase group track, which is then sent to Woodwinds group and then to the orchestral out etc etc.

In order to make a pro job of balancing the instruments in the template, I am seeking, an objective way of balancing the loudness of each track - first instrument against instrument (flute against flute), then instrument against other instruments in the orchestral sections, then sections across the orchestra.
instrument group.
I could use Cubase's control room to measure soloed loudness of tracks 1 to 1 against each other, but the issue for me is all instruments of the orchestra have different natural levels of loudness, so a triangle has less DB than say a sustained French Horn. Not sure what levels to aI'm for.

In order to understand how to professionally balance this template, I have concluded that I must understand my Cubase 11 meters.I am more musician than engineer.

Let me say now I am NOT seeking to maximize loudness of my tracks, use limiters and crush them with compressors. I want a broad dynamic range and plenty of headroom for partials in these orchestral mixes.

Reading up, I have learnt the following:

LUFS (or Loudness) metering is taking over from Peak level (dBs) metering because the human ear perceives the volume of different frequencies differently. LUFS compensates for this and gives us a more "human" measurement of loudness. The purpose of the loudness scale as far as I understand it is to compensate for facts like a sound at 100 hz of X Db is subjectively perceived as quieter than a sound at 1500 Hz also at X dB . In short two sounds of different frequencies, of the same dB, are perceived differently in terms of loudness. This is because we are wired for sounds in the range of human speech, unlike whales.

Q1.

I am told that 1dB = I LU. This makes no sense to me as if it were true it would seem to me that the two scales would be identical (Db and LU scales) and that would be pointless.

So, would it not be the case that a LU scale would change so that 1 dB would NOT be equal to 1 LU at certain frequencies?

Q2
Engineering Boffins (and that excludes me), talk about various Loudness measurements:

1] Integrated or "relative" loudness. The describe as: "Average loudness that is measured over the whole track in LUFS (Loudness Unit, referenced to Full Scale."

What is this Full Scale? Google does not know and utube sweeps over the topic.

They go on to talk about:

Short-Term Loudness
Loudness that is measured every second on an audio block of 3 seconds. This gives information about the loudest audio passages.

OK that's easy to understand

Momentary Loudness
Maximum value of all momentary loudness values that are measured every 100 ms in an audio range of 400 ms.

OK so that's super-short loudness

Then they talk about Loudness "range"

In this is a single LU number that measures the dynamic range over the whole title in LU. There are a few caveats in my definition of this measurement, but it's a LU number which represents the distance between the softest perceived sound, and the loudest perceived sound - sort of.

This is what I am learning:

"A loudness range between 6 LU to 12 LU shows that a track has a considerable difference in loudness between the various sections." ( I would think for classical pieces often this could be toward the max.

"Tracks with a loudness range below 4 LU could be considered rather static in loudness." Probably this is some kind of crap rave track that is supposed to bypass the brain and appeal to your anatomy.

OK so this is important information - ignore it at your peril when mixing and creating tracks. Realize that Cubase mixer tracks use DBs whereas the control room, which everyone should use, can meter in either loudness or peak DB values.

Bringing all this back down to composing. It's clear that different manufacturers provide instruments sampled at widely different volumes. So if one is going to have a professionally balanced template, where one can easily swap out a ridiculously soft Berlin piccolo with a loud EWest piccolo, then adjustments need to be done, either in Kontakt or in the pregains of Cubase.

But (I have now actually got to my second question! ) to what level? Obviously different instruments can create different levels of volume.

If one does nothing, then one has a very lumpy unequal template. If one levels all instruments to one level, then one could ruin the template. If one does this " subjectively", by ear, then its so easy for mission creep to set in, by which I mean in one session balancing the flutes, you set levels at X, the next day the clarinets get level Y, the strings level Z and again the whole process goes out of kilter

So, that is why I think an objective measure of some sort is required, so as to avoid mission creep. But what measure? I suppose they will be loudness measures but which ones?

Comments

kmetal Tue, 12/08/2020 - 16:03

You can't really mix with your eyes to any great result. Even if all the meters were exactly even, the sound would not be balanced because we don't hear flat, and different instruments will have different frequency content, and speakers project sound unevenly, and every room has a different frequncy response and sound decay curve/time.

Your best bet is to get as accurate a listening room/ speakers as possible.

With your template you can spend time getting a good balance that translates, then use that as a default set of settings for other compositions using the same settings. From there it can be minor adjustments since your in the ballpark.

It might make sense to hire a pro studio and engineer to get your mix right at first, and then you will have your starting point.

Eventually after a while you will probably notice similar mixer settings being used when using similar instruments, music style, and gain settings. But you will always have to use your ears.

KurtFoster Tue, 12/08/2020 - 16:13

kmetal, post: 466203, member: 37533 wrote: You can't really mix with your eyes to any great result. Even if all the meters were exactly even, the sound would not be balanced because we don't hear flat, and different instruments will have different frequency content, and speakers project sound unevenly, and every room has a different frequncy response and sound decay curve/time.

Your best bet is to get as accurate a listening room/ speakers as possible.

With your template you can spend time getting a good balance that translates, then use that as a default set of settings for other compositions using the same settings. From there it can be minor adjustments since your in the ballpark.

It might make sense to hire a pro studio and engineer to get your mix right at first, and then you will have your starting point.

Eventually after a while you will probably notice similar mixer settings being used when using similar instruments, music style, and gain settings. But you will always have to use your ears.

+1(y)

pcrecord Wed, 12/09/2020 - 08:05

I found the whole idea over complicated.. then I'm not into classical music at all..
The only thing, it got me to question the level buildup.
if you record 20 tracks with louder parts of -18dbfs, what how many db will hit the master buss if you let the faders at nominal.
Of course I know it's content dependent, lower frequencies will affect the loudness reading more that high frequencies. but in general ?
Is there a calculator or a equation we can use ?
Why am I asking this ? Simple, If I know that 40 tracks will be too loud at -18db of input gain, I could decide to record at -25db instead and push the preamps a little bit less.. no ?

audiokid Wed, 12/09/2020 - 10:36

Hi ZeroZero , welcome to the forum. You sound like a deep thinker!

Your concept sounds a lot like orchestral keyboard patches> cross fading, midi mapping etc without dynamics/ velocity controls to me. Are you are thinking this concept would apply the same way for mixing large tracks? Maybe I'm not quite understanding this?
Being a musician myself first, engineer second can give us some great advantages as an engineer but it can also make us over-think the process of mixing.
First and foremost... when it comes to mixing music... if our mixing room isn't tuned perfect all your templates with volume presets will be off in every song over and over.
Other than building great templates for large track counts, it sounds like you are going to see this as a big waste of time because nothing about mixing real music is exactly the same all the time. Thus templates don't really work much past the basics.
Templates are great to set up the basics but its sounds like you are going to be endlessly adjusting presets to compensate for dynamics which is exactly what you are trying to avoid by doing this in the first place. In other words, what you are hoping to achieve is only going to prove to complicate your mixing.
The best template for recording, mixing and mastering music is actually a great room. This is the best kept secret to awesome sounding music.

pcrecord Wed, 12/09/2020 - 11:08

I agree with Chris, templates will limit you over time.
The only template I use is my drum recording tracks with the right input already setup. (because my drumset is always ready to record with the mics installed..)
Yes it takes time, but there is something organic when starting from scratch and it forces you to analyze and do things differently from time to time..

kmetal Wed, 12/09/2020 - 17:10

pcrecord, post: 466205, member: 46460 wrote: I found the whole idea over complicated.. then I'm not into classical music at all..
The only thing, it got me to question the level buildup.
if you record 20 tracks with louder parts of -18dbfs, what how many db will hit the master buss if you let the faders at nominal.
Of course I know it's content dependent, lower frequencies will affect the loudness reading more that high frequencies. but in general ?
Is there a calculator or a equation we can use ?
Why am I asking this ? Simple, If I know that 40 tracks will be too loud at -18db of input gain, I could decide to record at -25db instead and push the preamps a little bit less.. no ?

I think the place to adjust an overloaded mix bus would be at the track fader/vca or the mix bus itself. Protools mix bus is pre insert so when you adjust the mix bus you mess with the levels going to the inserts on it. I don't think every daw handles the mix bus inserts that way.

The recording level is set to optimize signal to noise, and dynamic range of the source.

Dan Zellman, tech for Sear Sound, Abbey Road's tape machines, and all around great engineer and dude, told me he tracks kick drums digitally at like -35 or -40. Something like that. He explained that he does so because thats the typical dynamic range of a kick drum.

Some converters are calibrated to -22 or -24dfs = 0db on a vu meter, so the -18dbfs average level is just a good generalized practice.

As far as mix tenplates i think it makes sense to have some auxes, vsti, and re-amp/replacement type things set up and routed, even if the effects are disabled. I find menu surfing a complete buzzkill. Im too experimental and not busy to need to have complete mix templates setup with tons of inserts pre loaded like some of the big dogs have.

Although for writing, it is nice to have a a channel strip or chain pre loaded for quick eq/compression mix console style.

Its been a while since i worked on anything so my new system may include a very big template with lots of stuff spread out on the slave computers, to simulate the "room full of synths" that some people have in their sonic playground. It should help keep many things a click or two away. I also will probably set up certain chains of efx particularly for busses/sub mixes to easily route things to them, again keeping sub mix and parellell options a click away.

Having adopted some standard mix bus effects 1/8 note delay, vocal doubler, ect I've had them in my template for a while.

One thing i want to experiment with is the sonic quality of having all those things engaged in a session. DP7 used to hit a wall where at a certain point when enough effects were on it would suddenly get flat and a bit grainy and foggy, it was strange but happened routinely. And it seemed not to revert back even if the processing was removed. This could be dp or the computer hitting some threshold, but it was real.

If Samplitude (or any daw) has that same quirk i can see a case for "add as you go" for all but the most commonly used things.

Regardless, when it comes to final mix down i always remove any unused, or disabled effects and channels. I like to bounce from the most clean possible session, free of edits, and vsti ect. Full lenght audio regions.

It would be interesting to experiment with printing the inserts on the tracks before bouncing to stereo so the only thing live is the busses/sub mix. I would be curious to see if there was an audible difference for better or worse.

paulears Thu, 12/10/2020 - 01:30

Classical music is our thing - oddly too much signal is rarely the problem, quite the reverse. I find that I'm often having to lift tracks to get them audible. For those who don't do classical stuff much, the typical classical instruments come in numerous playing styles. I'll use one of my favourites to explain. Pick a flute sound and it has indicated styles available from pp to ff, quiet to loud, but the reality is these are ranges, not absolute levels, so when you play it on the keys, you can play the quiet sound at MIDI velocity 27 and it will be just 'chattering' that sort of volume a flute can play when the wind pressure is just about enough to form the note and not be just wind. At 127 it will fully formed quiet. If you pick the other end of the scale (the ff) and play at 127 levels it will be full in your face screaming with overtones and noise. Orchestral music might be doing Ride of the Valkyries and every line you have going is going nuts but real orchestras rarely do that. When you record a live big orchestra, thank God for digital dynamic range because while you will have set maximum in rehearsals, for the majority of the time, you are nowhere near maximum. I use compressors quite a lot for woodwinds because their texture means they often play quiet, so the compressor slope doesn't change but I use the make up gain function to get me a fader setting that works.

This is also why I gave up templates. They were just becoming my favourites and I used X violin and Y cello and Z flute EVERY time, because it came up in the orchestral template. Now if I want a flute, I still have favourites, but buying the Kontakt keyboard last year changed things. Kontrol coming up as a browser on the keyboard revealed just how many sounds I had! If you select bowed strings, for example, you can audition little samples of each one, and I did not realise I had so many - dozens and dozens, from all the instruments in the different libraries. So now, If I want a cello for a certain piece, I go to Kontrol and step through each one and when I hear a sound that might work, I press the button and an instrument opens up that I would never have thought of. This is why I have stopped using templates - my choice is better. I've been doing a series of electronic classical stuff, inspired by Tomita from the 70s, and I wish it was as easy for synth sounds. The makers just have really rubbish names and descriptions.

My stuff probably rarely exceeds 30 tracks. I like section instruments where the 6 violins, or 3 flutes have already been blended, so I don't think I have ever overloaded a bus. In Cubase even the busy mixes only need a small drop of the the master to get the output back to roughly where it should be. When I have a busy mix, I probably create groups, or VCA groups, and these can get the louder clusters sorted without red lights popping up. I have no idea how Cubase handles bussing, in level terms, but I just know I don't personally have issues with this at all. It just works. My favourite composers who use Cubase do seem to like templates, they just don't do much for me.

pcrecord Thu, 12/10/2020 - 06:13

kmetal, post: 466209, member: 37533 wrote: I think the place to adjust an overloaded mix bus would be at the track fader/vca or the mix bus itself. Protools mix bus is pre insert so when you adjust the mix bus you mess with the levels going to the inserts on it. I don't think every daw handles the mix bus inserts that way.

Interesting.. I try to keep the track faders near nominal because they are more precise in that area. (this habit came from FOH analog work) I think it still make sens to me when I start to automate volumes. Of course I could always compensate with the gain setting on each tracks..
I was just curious if anyone has some kind of calculation, for exemple, each adding tracks at -18db will add 3db to the master buss ..

ZeroZero, post: 466202, member: 52141 wrote: The Template has about a thousand tracks and will be used to pull ready made presets for largely (but not exclusively), orchestral projects, that will be part of gthe same production and require similar levels.

A thousand tracks ? I'm curious if any DAW wouldn't crash with that many VSTi or even audio tracks.
Also, has any real orchestra ever counted a thoushand musicians in the same room ?? That would be Epic ! ;)

paulears Thu, 12/10/2020 - 07:36

The tracks actually don't stress the machine too much, apart from if the template actually loads an instrument. If you have say 60 packages you could have them appear in the instrument list, then drag them to the main screen to use them, or you could have them in empty tracks. I don't find a full screen easier to navigate, so not having it set up this way doesn't really matter. Having even 50 all with big sample instruments all loaded up would be, er, interesting.