Normalizing VSTs Across a Set

[Motorist]

I need to normalize a dozen or more VST instruments throughout a set of music. There are dozens of tunes and several part transitions within each of them. Most of the instruments are in Kontakt and I can edit the load-time volume level, amplitude, velocity and even add more gain via dynamics. I'm trying to avoid making constant adjustments using my volume pedal and the sound engineer doesn't have time to micro-manage my sound.

This should be a typical problem for keyboardists who are playing live using a wide variety of different software libraries. I'd like to follow a repeatable process to achieve success. In the past, I've edited gain ranges on a hardware synthesizer, with uneven results. I've been using a sound meter and PA to measure results. For software, the only measurement tool I have is Spectralissime by VB-Audio. In addition to Kontakt 7 and other sound libraries, I have Cubase 11. My final performance software is VST Live Pro 2.0, which lets me set a level per song part. However, if an instrument already lacks headroom it's too late to fix it at that point without adding a dynamic insert that then creates a difference I have to replicate as needed. I'm trying to keep the keyboard controller and VST Live usage minimized so I don't have to twist knobs during a performance or spend time inserting software dynamics on a per-tune basis before the performance.

Before editing the instrument output ranges, should differentiate the solo instruments from the ensemble or section instruments and have different target volume ranges?
1. Some of the solo instruments can only be pushed so far before you begin hearing a lot of white noise.
2. Some of the ensemble collections, like the brass, might be used mainly in the background behind vocals.
3. How might I measure or evaluate signal-to-noise so I know when to stop?

I play a lot of split keyboard combinations with strings and piano or strings and horns. If I need to use a software compressor on the strings or oboe so it can always be heard, where do I stop as it starts to affect its dynamic range?

If I use a software limiter as the last insert before output to the mains I can stop random transient loud notes from escaping, but then I have to watch how my volume pedal is affected by the limiter, right?

Thanks

[MarkOne]

What Software host are you using (presumably this is a live set using plug-ins instead of hardware keyboards)

I'm on a Mac using MainStage as the host to manage my shows and I either balance things our with the MainStage mixer or have gain plugins on each software instrument channel.

I then check my full set-list is well balanced and has a reasonable headroom going into the mixbus

I know others use Cantable on their windows PC to host their VSTis , and I'm sure you can do the same in that host with either a gain plugin, or the set mixer

[The Elf]

I suspect you're overthinking this.

I've never for one moment considered 'normalising' my sounds across a live set. I set up my levels in rehearsal and I go with it - and I use Cantabile under Windows - which makes it simple to set levels for each instrument (in reality I created my own mixer rack in Cantabile, but that's just by-the-by).

I keep all my levels conservative, such that I have headroom to turn up if the (extremely) rare need arises. I keep my internal levels to -10dBFS peak and I present FOH with an isolator box sporting line and mic outputs - the rest is up to them. I give them a -10dBFS signal for them to set their peak and it's job done - curtains can open.

I would never use a limiter in my live output chain. Again, this is up to FOH. Even if I did use a limiter I would be horrified if by normal use of my expression pedal I could go anywhere even approaching its threshold.

I never use compressors either, since live I want all the dynamics I can muster, but I can imagine many would, and that's fine.

[Tomás Mulcahy]

I am with the Elf on this. Do it by ear in rehearsals. You will probably tweak it a little in the first few shows and the settle in to it. No point in measuring it since every place you play will change that. I learned a long time ago that a limiter is a very bad idea for live. Keep the dynamics, let FOH take care of that. In 2015 I did a festival show with no sound engineer (the tiniest of stages ). Mainstage was brilliant. Focsrite Red compressor plugin on the bass, guitar and vocals. Dialled it in for rehearsals. Only thing I assigned to the controller was master volume since we did need to adjust the overall song levels in the set. Worked brilliantly. There was a limiter on the speaker processor, I don't know if we ever tipped it but phone recordings from out front sounded well balanced.

[sonics]

I suspect you're overthinking this.

Do it by ear in rehearsals.

Yes and yes. Find your balance during rehearsals. It should be obvious where there are problems, and that's your chance to sort them out. I would caution against controlling the dynamics too much; let the FOH engineer do that. Needs will change with the venue.

If you have/know a good FOH engineer you can get some advice from them. I've invited one along to a gig in situations where I can't monitor the balance myself at a soundcheck.

Unlike Elf, I do like to use some limiting and level control tricks. They are part of the sound I want, and the limiter helps avoid those situations where an engineer sees a peak and turns your whole feed down, thereby ruining the balance.

[Motorist]

What Software host are you using (presumably this is a live set using plug-ins instead of hardware keyboards)

Steinberg VST Live Pro 2.0 - It's still a little buggy, but comprehensive and intuitive.

I'm on a Mac using MainStage as the host to manage my shows and I either balance things our with the MainStage mixer or have gain plugins on each software instrument channel.

Yes, each sound has a separate meter in each song part in the Steinberg product. However, when I edit each patch at the source - directly in the library, I always find real problems I can fix there. Each instrument sample set has unique issues. I'm prepared to resolve these issues at the source so I don't have to than add gain inserts across dozens of mixer instances (one per song part.)

Edited to fix quotes --ED

[Motorist]

I suspect you're overthinking this.

I've never for one moment considered 'normalising' my sounds across a live set. I set up my levels in rehearsal and I go with it - and I use Cantabile under Windows - which makes it simple to set levels for each instrument (in reality I created my own mixer rack in Cantabile, but that's just by-the-by).

First rehearsal, I didn't bring a powered monitor and couldn't come through the PA much because the entire mixer sub-mix section was fried and not useable to send the singers their own mix. When they turned me up in the PA, it killed the singers. I'll see how it goes next rehearsal when I bring my own monitors.

I keep all my levels conservative, such that I have headroom to turn up if the (extremely) rare need arises. I keep my internal levels to -10dBFS peak and I present FOH with an isolator box sporting line and mic outputs - the rest is up to them. I give them a -10dBFS signal for them to set their peak and it's job done - curtains can open.

I have a Radial direct box. Left out to the mixer and Right output will now go to my powered monitor.

Okay, -10dBFS seems like a useful threshold at the final stage, but my problems are resident in the VST libraries themselves as I have some patches that can't compete among themselves. For example, I've set the minimum MIDI velocity on a bari sax to 30/127, otherwise any note will be gone before it is heard, but that's just a quick shortcut. As a group, the low horns need special treatment and I now know just enough about sound design to resolve certain issues, but it's been like trying to learn to fly a helicopter.

I would never use a limiter in my live output chain. Again, this is up to FOH. Even if I did use a limiter I would be horrified if by normal use of my expression pedal I could go anywhere even approaching its threshold.

Okay, they can do that for me. They will anyway because they want to protect their equipment.

I never use compressors either, since live I want all the dynamics I can muster, but I can imagine many would, and that's fine.

I just now read about EBU R 128 and improved tools for measuring loudness. At this time, Cubase 13 comes with Supervision which is one implementation for measuring instruments and mixes.

[Motorist]

I am with the Elf on this. Do it by ear in rehearsals. You will probably tweak it a little in the first few shows and the settle in to it. No point in measuring it since every place you play will change that. I learned a long time ago that a limiter is a very bad idea for live. Keep the dynamics, let FOH take care of that. In 2015 I did a festival show with no sound engineer (the tiniest of stages ). Mainstage was brilliant. Focsrite Red compressor plugin on the bass, guitar and vocals. Dialled it in for rehearsals. Only thing I assigned to the controller was master volume since we did need to adjust the overall song levels in the set. Worked brilliantly. There was a limiter on the speaker processor, I don't know if we ever tipped it but phone recordings from out front sounded well balanced.

I think I'll get rid of the volume pedal on the floor as mine is hyperactive and therefore dangerous. I shouldn't have to adjust volume during tunes, and a controller slider would be more precise.

[The Elf]

It would be worth considering in-ear monitors - even a humble pair of domestic ear-buds. Just a smidgeon of your own signal into them might be enough to help you hear your parts above the melee. For this purpose you don't need anything fancy, or perfectly fitting - just a little signal to help you pick yourself out of what's going in around you.

VST levels are a lottery. Many VSTs are set to blast out at 0dBFS, while others are more conservative. But dedicated performance software should be able to handle levels for you and switch gain in/out as necessary. I can't speak for other software, but in Cantabile I press one button and my entire rig can change in an instant - all hammered out during rehearsal.

It sounds to me that you're using Cubase for live performance? Much as I love Cubase, I wouldn't do that myself - it's just not the tool for the job. You should at least take a look at other options, and maybe add a small control surface to your live rig, unless your controller keyboard has a few faders and buttons of its own.

[sonics]

It sounds to me that you're using Cubase for live performance?

VST Live is Steinberg's live performance host and control software. Released in 2022, and now in V2.0. As a Cubase user I'd have thought you knew that, but perhaps you don't read the newsletters?
I know you love your Cantabile, too.

[Motorist]

It would be worth considering in-ear monitors - even a humble pair of domestic ear-buds. Just a smidgeon of your own signal into them might be enough to help you hear your parts above the melee. For this purpose you don't need anything fancy, or perfectly fitting - just a little signal to help you pick yourself out of what's going in around you.

IEMs require a sub-mix and the rehearsl space mixer is broken for all submixing. The place is a dump, but I'm the only one with a problem. Therefore I get to set up two personal monitors, one for the band and one for me.

VST levels are a lottery. Many VSTs are set to blast out at 0dBFS, while others are more conservative. But dedicated performance software should be able to handle levels for you and switch gain in/out as necessary. I can't speak for other software, but in Cantabile I press one button and my entire rig can change in an instant - all hammered out during rehearsal.

Steinberg VST Live is complete regarding gain options. I foot pedal through a song and each song part has it's own mixer. It's just that some instruments are out of range for the musical context. If I fix these at the source, in the library itself, I'll only have to use these song/part sliders. I won't need to add additional inserts to the performance software, which are then only recorded in that single project file. When I start a new project, I don't need to remember which instruments needed inserts for dynamics in the last project.

I'm looking at Voxengo SPAN and the Steinberg Supervision feate in Cubase to re-balance the patches. I like the science behind the newer loudness stardards and R128. Perhaps more sound library designers could get in line with those.

It sounds to me that you're using Cubase for live performance? Much as I love Cubase, I wouldn't do that myself - it's just not the tool for the job. You should at least take a look at other options, and maybe add a small control surface to your live rig, unless your controller keyboard has a few faders and buttons of its own.

So far, I've managed to use nothing on the Akai MPK261, except the pedal ports. Steinberg VST Live Pro 2.0 is easily worth $100. Their support has been responsive - they actually find and fix bugs quickly.

Edited to fix quotes --ED

[Tomás Mulcahy]

R128 will cause you two problems- first is it cannot deal with the component parts of the mix, second is it cannot account for the venue acoustics. It's a tool for broadcast, not live work. And even in broadcast it is not the ideal solution, various organisations are working on improvements such as object based streaming. Sound designers don't use it because again, it is for measuring the entire mix not component parts. The sounds will form part of a mix and there is no predicting how they might be used, so R128 is not appropriate. If you set them all at that standard they would be too loud in a mix. But we all know that making patches without headroom is a widespread problem. When I do patches (I've done them for Arturia and Audiotent among others) I try to level them by ear and leave some headroom, but not everyone does unfortunately.

Pedaling levels during the song sounds like a total distraction to me. Pedal is for doing expression not mixing. But if you can mix while playing, more power to you. I could not.

[Motorist]

R128 will cause you two problems- first is it cannot deal with the component parts of the mix, second is it cannot account for the venue acoustics. It's a tool for broadcast, not live work. And even in broadcast it is not the ideal solution, various organisations are working on improvements such as object based streaming. Sound designers don't use it because again, it is for measuring the entire mix not component parts. The sounds will form part of a mix and there is no predicting how they might be used, so R128 is not appropriate. If you set them all at that standard they would be too loud in a mix. But we all know that making patches without headroom is a widespread problem. When I do patches (I've done them for Arturia and Audiotent among others) I try to level them by ear and leave some headroom, but not everyone does unfortunately.

Pedaling levels during the song sounds like a total distraction to me. Pedal is for doing expression not mixing. But if you can mix while playing, more power to you. I could not.

I'm not using volume or expression pedals, just switches to move forward to the next song part. This is for hands free progression through the songs and parts. I did have a volume pedal on the floor but can now get by with a slider on the controller. Before I used software I had a Liquid Foot controller. Software is easier.
I've started reading a book, "Audio Metering" by Eddy Brixen.

[The Elf]

It sounds to me that you're using Cubase for live performance?

VST Live is Steinberg's live performance host and control software. Released in 2022, and now in V2.0. As a Cubase user I'd have thought you knew that, but perhaps you don't read the newsletters?
I know you love your Cantabile, too.

I do know about VST Live, and have tried it. I found it obscure and unhelpful. It really didn't address the way I approach setting up my live rig at all. It seemed to be all about the backing material. Stuff I find simple to achieve in Cantabile was either not possible, or so inflexible as to be no help.

But that's just me. I'm sure it suits others.

[sonics]

I do know about VST Live, and have tried it. I found it obscure and unhelpful.

Interesting. Maybe they did not research the market fully, or it's aimed at a different type of user? I use Gig Performer and the Studio One Show Page myself.