Note: I created this tutorial in order to learn myself – any improvements to my method will be gratefully received.
Tutorial 2 controlling levels using Cubase and GPO
Ok so we assume that you have set up a ‘GarritanBasic’ template in Cubase as in my tutorial 1.
In this second tutorial, we are going to look at how to set up levels in Cubase/GPO, and how to balance musical sounds. Before we proceed let’s take a quick look at the issues.
Samples versus real instruments:
With the real acoustic instrument we can vary the amplitude and timbre of the sound in a variety of ways, our GPO samples are a cut down version of all the possibilities an acoustic instrument offers.
Secondly, a single instrument’s timbre can vary widely across its range – a flute, for example, will be soft and bulbous in it’s lower register, but shrill and piercing in its higher octaves.
Also have an effect on the overall level of the sound. Thus, matching sounds for timbre and amplitude cannot be done on a once and for all basis, at the beginning of a track. Each note is important.
For a real player Vibrato is different for every note he/she plays. Every note is a universe. You can handcraft/simulate attack and vibrato for any given note – especially long note solos. See my post ‘Real Vibrato and crafting phrases’.
Sound engineers tweakings
After the sound engineer records a sample of say a pianissimo note on a flute, they may bring up the volume of the recording – this standardises the sound viz or viz other flute notes. Use your ears to hear whether this is the case. Occasionally a little bright EQ is added to the higher register of notes (I believe Tom mentioned this in a post).
At the level of the phrase it may be that you want a crescendo or a diminuendo when playing a real instrument. It may be that you want to use a couple of staccato notes and make the rest legato.
Bowings and embouchures
Some notes might be louder than others or be produced with a varying embouchure or bowing. All these factors contribute to giving the phrase a unique identity – minimising boredom. All have to be understood before they can be controlled. Lets first list the possible ways to control GPO levels:
The sound stage
An insight I gained from reading Alan Belkin’s excellent tutorials is that when we process music in our minds we are not concerned primarily with levels, but with prominence i.e. foreground/background or degrees of. Any musical passage can become more prominent by many other means, such as ‘novelty’, ‘complexity’ or timbral or registral change. Alan’s site is very deep, but its well worth spending many hours digesting. There are three virtual books to start with! Here is a link to the discussion of foreground and background:
OK so now we examine the actual controls offered by Cubase and GPO.
Volume Control in Kontakt.
This can be used to control the volume level for instruments as a whole. It effects all samples in the instrument. This means that if you use the same instance of the instrument form the same GPO slot in the same Kontakt player, on two different tracks, raising or lowering this control will change the level on both tracks.
Can be controlled from keyboard and from Key Editor screen in SX. This is best reserved for affecting the volume level of individual notes. Use in tandem with Velocity.
Velocity does not really control volume in GPO it controls the sample used (for most instruments). For example you might have three samples for middle C on a particular trumpet. The first is the sample when the instrument is played softly, the second; sample is the instrument mf, and the third; is a forte sample. These three sounds will differ widely in wave structure – its not just a question of raising the amplitude, or getting louder, the shape of the sound changes too. When we recognise a trumpet sound in our brain we recognise a suite of sounds which we (only) associate to the same object. (Aside: For those of you who understand linguistics think of the relationship between phones phonemes and allophones). Velocity settings control which sample is used. In this example there would be three ranges of velocity settings – 1-30 might be the soft sample, 30-100 might be the mf sample and 100-128 might be the loud sample. Use the velocity settings in tandem with Modulation settings, this way you wont get a soft trumpet sound playing at loud amplitude.
I don’t think that the actual settings for velocity sample changing are published for the individual instruments in GPO, lets hope that’s part of the new manual.
Volume Control in Cubase MIDI Tracks
I am talking MIDI tracks only here. I do NOT recommend using these track controls (see below) but we enumerate them so that we can avoid them. All the following faders and knobs work together on a given midi track – raise one, raise them all; lower one, reduce them all:
1. Volume control slider in the MIDI track’s inspector;
2. Volume control in the mixer’s MIDI channel strip;
3. Volume control in the Key Editor window, whilst the relevant MIDI track is highlighted.
4. Volume control in event display for the MIDI track – shown as a line-gradient under SX’s midi track display.
5. The channel strip in the MIDI track’s Inspector, if displayed (controller click ‘Channel’).
Think of all these volume controls in this group as a single volume control that you can access via different graphical interfaces. My advice is to identify them and then avoid them – use the ‘Stereo Out’ pairs instead – see below. See my first tutorial for reasons why.
Using the ‘Stereo Out’ pairs loaded into the eight GPO outputs
This is only possible if you followed my first tutorial and set up the ‘GarritanBasic’ template. Read this tutorial first. This is my preferred method for control volume for a whole track. Load the GarritanBasic template, go to SX and bring down the channel outputs for the eight outputs of PersonalOrchestraVST. Highlight one channel within. These channels function basically like audio channels – giving you access to reverb and FX. You can now use any of the following methods to control the level for the given track.
1. Volume control slider in the Audio track’s inspector;
2. Volume control in the mixer’s Audio channel strip;
3. Volume control in event display for the Audio track – shown as a line-gradient under SX’s midi track display.
4. The channel strip in the Inspector, if displayed (controller click ‘Channel’ in the Inspector).
This list is the same as the above for the midi channels but it is critical to realise we are now we are using audio strips – this way we can also control reverb and fx using the same strip. We avoid midi channel strips in the mixer because this reduces confusion – again see tutorial 1 for a proper explanation.
Master Volume Control
There are two master volume controls. Firstly the one on your amplifier. This should be set to a comfortable listening level using a professionally produced CD. Be aware that you could be listening through speakers, or through headphones – what difference does this make in your system? If you change this volume setting what effect is this going to have on your recording? The second Volume Control is the ‘Stereo Out’ strip in the mixer – this should be set to 0.00 in most cases, at least to start with.
Automated Volume Control
Lastly, it should be remembered that Volume Controls for a given channel strip can be automated in SX – perhaps to produce a crescendo or diminuendo for a track as a whole (these slopes can also be drawn). Here is a method for automation in short.
1. Go to the mixer. Select the channel strip you want to automate. Highlight the button marked ‘W’ the one with the tooltip ‘Write Automation’ (left of the fader).
2. Record your track whilst altering the volume fader.
3. Deselect Write Automation and select ‘R’ button to ‘Read Automation’. The volume levels change in accordance with your last recording. If you want to suspend the volume changes just deselect ‘Read Automation’ this will give your original levels back.
Automating volume is good for bringing the levels of a section up.
1. First think about your sounds and how they will blend – Will an oboe blend better with a clarinet or with a flute? Test them across there ranges by writing scales a third apart (thanks Mr Belkin). Create test chords. Try to think of a section at a time.
2. Be familiar with the idiosyncrasies of each type of instrument - the effects of the mechanical operations of moving through a register – these constraints effect the character of the instrument. Adler’s book - The study of orchestration is thorough. Paul Gilreath’s ‘The Guide to Midi Orchestration’ which is more sample-orientated is almost as comprehensive.
3. Master Volume Control on your amplifier: Set this first to comfortable listening level – remember to mentally compensate when using headphones
4. Volume Control in Kontakt: Use this to control the volume of the loaded instrument on all tracks it is used.
5. Modulation Wheel and Velocity: Use these in tandem with each other to craft individual notes and phrase dynamics
6. Volume Control in Cubase MIDI Tracks: Do not use – reduce complexity.
7. Audio outs from GPO channels (in SX)– Use these to control levels for each track
8. Automation – useful for writing in fader control data that changes over time.
9. Master Stereo Out in mixer: Not normally touched –except perhaps to fade out the whole orchestra with automation.
OK so it’s a lot to take in first time round. Lets look at how we approach creating and balancing a string section pad chord. Hang on a minute – maybe that’s a third tutorial! We could also look at using group channels in Cubase, to create advanced templates.
Ok so I leave that topic for now.
Feedback appreciated, if you disagree with the approach I would love to know.
Thanks for taking the time to write these tutorials. They will be a great benefit to GPO and Cubase users. When I get back from the NAMM show I would like to make theese available on the GPO tutorial and Cubase pages.
I know I'm seeing these tutorials almost a year after they were put up, but I wanted to give "thumbs up" for these helpful posts.
There was this one paragraph from the 2nd tutorial which I felt could be misleading, regarding mixer automation in Cubase:
"3. Deselect Write Automation and select ‘R’ button to ‘Read Automation’. The volume levels change in accordance with your last recording. If you want to suspend the volume changes just deselect ‘Read Automation’ this will give your original levels back."
Keeping both "Read" and "Write" on at all times is much more convenient, but that's the not as important an issue as:
Deselecting "Read" doesn't actually return the data to what it was before the automation recording, it just bypasses it. I don't think you meant to indicate that deselecting "Read" would erase data, but how this paragraph was written could possibly be taken that way.
One needs to un-do the last action/recording in order to truly get rid of a bad slider move, or if there's even older automation data that needs to be edited, it's most accurately done in the linear track view of the data, so you have access to the nodes, moving or erasing them as needed.
Of course you can record new automation data over already existing data, and that's the first step towards correcting an automated mix--but when you let go of a slider, finishing your new pass at the section you're working on, you may find the slider snapping back to an old position, producing an unwanted jump in volume. That's when hand editing of the linear track fixes things up best.