Creating an Orchestral Room for a Virtual Orchestra Film music, John Williams Style By Graham Plowman of Cthulhu Mythos Music Version 2.0 (September 2022) Overview The purpose of this guide The goal for this guide is to help provide the building blocks for creating an orchestral set up with virtual instruments that can produce a realistic room sound. My benchmark for this is the earlier recordings of John Williams film scores such as Raiders of the Lost Ark, The Empire Strikes Back and many others in the 80s and 90s, including the prequel Star Wars recordings. If the new modern Star Wars recordings such as The Force Awakens, with their cleaner digital sound and less ‘room’ quality is your preference, you can certainly use this guide to achieve that as well, it’s just a matter of adjusting depth and reverb to taste. The choice of room depth and many other decisions you’ll see is going to be ultimately up to you. There’s nothing in this guide that suggests this is the right way and other methods are wrong, it is the method I’ve had the best results with and is simply a guide to get you off the ground. All methods I use here can be tailored to your own preferences or dumped entirely (in which case I apologize in advance for wasting your time!) What you should already know and have…* TIP: There is likely going to be a lot of information you already know. It’s ultimately up to you if anything here is new, different or useful to you, and there is no suggestion that one method or another is wrong. I base my methods on results, not on what some may suggest is ‘supposed’ to be done. Remember, too, that this is a singular aim, a good room orchestral setup, not a template that will create any style of music for any situation. That isn’t how a template works. A digital audio workstation (DAW), such as Cubase, Ableton Live, Logic, Studio One, Reaper (any one of these or others would do). I will be using Ableton Live for this guide. Familiarity with how to route in your DAW and perform other tasks in it. General familiarity with sample libraries and how to use them in Kontakt (and other sample players) Orchestral Sample Libraries – an assumption is made that you have orchestral instruments available to you already. Strings, Brass, Woodwinds and Percussion. I do not make recommendations on which sample libraries you should buy as people's preferences (and the expense in purchasing these) means you should think hard about which ones suit you – a catch 22 because suitability comes from experience with a library. I do however have key expectations from sample libraries that are not met by all. I explain these later. Good MIDI programming is required with all sample libraries by the user. You can get great results from most libraries if you put enough time in programming them, choosing the right articulations, and have a well-trained ear. Some are easier to program than others. I have heard the best sample libraries sound poor, and the oldest sound good – skill of the user sets most of these apart. It’s worth nothing that despite this, certain libraries can only achieve so much, and others offer the possibility to reach a closer realism benchmark with less effort. *Despite some prior knowledge preferred, as noted above, this guide will still start with a lot of basics. Only because they are relevant to how I do things. I go through basic routing, mic use in sample libraries, and placement basics. I wanted to make sure this guide was covering everything needed to end up with a better result, even if starting with a blank project file. But I am using my ears? TIP: A commonly used phrase is ‘use your ears’. This is both a frustrating phrase to be told, but also a vital part of understanding issues. If your ear is untrained the phrase doesn’t help much. You will need to develop the sense that what you are hearing is natural sounding or not. A shocking revelation, but a well-trained ear is required to know what sounds good or bad. Using real recordings as references and critical listening to these recordings is vital. The goal is NOT to match the sound of original recordings, the goal is to have your own natural room sound and orchestra. But how can this be achieved if you cannot recognise this sound naturally? It’s tricky because there are different needs for different situations. Remember the focus of this tutorial, a John Williams orchestral sound. I will attempt to address this by explaining some basic balance techniques between instruments later. Several factors come together to contribute to the aim of a realistic virtual orchestra…(these are the ones I consider to be the most important ) Realism Chart • The principle here is that good and balanced orchestration already accounts for solving a large portion of both mixing and programming needs, meaning less reliance on those to fix issues. It also won’t matter how good your mixing and programming is (or even your sample library) – if you arrange orchestral music poorly, you may not get good results. Mixing ability and Reverb 20% MIDI Programming Skill 20% Sample Library Quality 10% NOTE! Orchestration accounts for half the overall total here (which you may or may not agree with). • Skilled MIDI programming also chips away at the need for greater mixing needs. This means utilising everything in your arsenal to produce a ‘performance’. Choosing the right articulations for example and cc automation is vital. Orchestration 50% • Good sample library quality (with a host of articulations, mic options and a good sound out of the box) further reduces the need to heavily mix and rely on reverb. If you rely on reverb to fix problems, you’re already in trouble. The dry mix (no reverb tail) must be balanced as much as is possible. This is difficult when bringing together a series of libraries from different developers as they can have varying degrees of wetness (room) baked in. If you add a dry instrument into an orchestra that is already placed with close, mid, and far mics, the dry instrument is going to stick out until it is given placement treatment with panning, convolution reverb to simulate a hall, and perhaps EQ for further sculpting of the sound. Before you add a reverb tail to anything, it’s best to have all the instruments balanced and placed, dry or wet. Note the distinction here that I’d consider the use of convolution reverb to place an instrument not part of the actual reverb set up for the whole orchestra. It is for hall placement. A Typical John Williams Orchestra This is the typical score order. You can include additional patches as needed. For example, you might have a horn (solo), a 4 horn ensemble, then 6 horns, 12…or 4 horn solos, 3 trumpet solos, those choices are up to you, but they are not necessary. Woodwinds Flute 1 Flute 2 Piccolo Oboe 1 Oboe 2 English Horn (if you have one) Bb Clarinet 1 Bb Clarinet 2 Bass Clarinet (if you have one) Bassoon 1 Bassoon 2 Contrabassoon (if you have one) Hi Woodwind Ensemble Lo Woodwind Ensemble Brass Horn (solo) Horns 4 ensemble Trumpet (solo) Trumpet 3 ensemble Trombone (solo) Trombones 3 ensemble Bass Trombone (solo) Tuba (solo) Percussion Timpani (hits and rolls) Bass Drum (hits and rolls) Snare Drum (hits and rolls) Toms Suspended Cymbals Large Cymbal Med Cymbal Small Cymbal Piatti Tam Tam Triangle Mark Tree Orchestral ‘Toys’ (Woodblocks, shakers, etc.) Crotales Xylophone Marimba Glockenspiel Vibraphone Celeste Harp Piano Strings Full String Ensemble (1) Full String Ensemble (2) Violins 1 Violins 2 Violas Cellos Basses Add solo and con sord patches if needed. TIP: You can add or remove any instruments you want. It’s your set up in the end and your choice whether to use certain instruments over others. This guide will show the most typically used concert performance instruments and is useful for many orchestral pieces, not just John Williams music. Sample Library Basics Though I won’t recommend any one library over another, I do have my own preferences for what makes a sample library useful. MIC options (Close, Mid, Far, etc.) – this helps you get a sound that has a room quality around it without even touching reverb. The natural ambience around a recorded instrument is very useful as a starting point to creating space. ‘Dry’ recorded instruments are very useful for smaller ensemble, and studio like mixes, for a more intimate sound. Take note of my distinction between a Close Mic and a Dry recorded instrument (see box out). More on making dry instruments work in the mix later… Performance ability: very hard to quantify until you spend time working with a library. Though a good range of demos can give you a sense of the performance ability of a library, they are often written in a way to avoid their shortcomings. Still, if you hear performance quality in the demos then it’s a useful indicator the library can do certain techniques. Things to look for are: • Good range of articulations (a range of shorts, true legato, longs, marcato, tremolo, and extended techniques). I list the most useful articulations on the next page. • Listen for fast repetitions - short repeated notes, particularly in the same pitch. If demos avoid this, see if you can find out how it sounds through a walkthrough or other means. Even if a library has many round robins, it does not necessarily mean it won’t sound odd with fast repeated notes. • Ability to change or shorten set-length articulations. For example, marcato in brass is a more aggressive mid-to short note, and if these are a pre-set length, it greatly limits your ability to use it. If you hold down a marcato and it stops when you lift the key, this is much more versatile. Pre-baked lengths are tricky to compose around but not impossible. TIP: Mic use in a sample library better represents how an orchestra is recorded in a real setting. There’s also a certain real room quality to the behaviour of instruments in these rooms that may not be replicated well using a dry close instrument with convolution reverb. I don’t consider ‘close mics’ to be the same as dry recorded instruments! I avoid using just close mics. I use all the mics available. NOTE! Close Mic versus a Dry Instrument. A dry instrument is recorded in a dry environment. A close mic’ed instrument is usually recorded in a hall, with a mic close to it (for adding definition to a recording). I consider the close mic only sound to be a thin sound and thus not a good representation of that instrument’s timbre. With this in mind: • For a dry sound, use dry instruments recorded in a dry environment. • For an orchestral sound, use close mics to add some definition to an instrument in addition to the mid and far mics. This is my list of the most useful articulations and instrument sections – it is not an exhaustive list of all possible articulations or instruments, and to you, perhaps one I haven’t listed might be more useful to you than I. I’ve not included Choir. Woodwinds Flute (solo) Flute 2 or ensemble* Piccolo (solo) Oboe (solo) Oboe 2 or ensemble Clarinet (solo) Clarinet 2 or ensemble Bass Clarinet (solo) Bassoon (solo) Bassoon 2 or ensemble Contrabassoon (solo) • • • • • • • • Legato Sustains Staccatissimo Staccato Sforzando Portato Marcato Trills * It is useful to have Flute 1 and Flute 2 if possible (and repeat for each woodwind), but either option works. Brass Horn (solo) Horns 4 ensemble Trumpet (solo) Trumpet 3 ensemble Trombone (solo) Trombones 3 ensemble Bass Trombone (solo) Tuba (solo) • • • • • • • Legato Sustains Staccatissimo Staccato Sforzando Marcato Mutes Percussion* Timpani (hits and rolls) Bass Drum (hits and rolls) Snare Drum (hits and rolls) Toms Suspended Cymbals Piatti Tam Tam Triangle Mark Tree Strings* Violins 1 Violins 2 Violas Cellos Basses Legato Sustains Spiccato Staccatissimo Staccato Sforzando Pizzicato Harmonics Tremolo Trills Marcato Con sordino Fast runs Harp (including glissandos) Piano • • • • • • • • • • • • • * Most percussion libraries will include this and much more, there’s a lot of choice available. These are what I use most. * I haven’t included solo strings, but you can if you wish. Orchestral ‘Toys’ (Woodblocks, shakers, etc.) Xylophone Marimba Glockenspiel Vibraphone Celeste Isn’t it better to have 4 solo horns? 4 Horn patches Some people prefer the versatile nature of having 4 solo horns, 3 trumpets, etc. This is more conducive to natural writing; but the resulting sound can be a little off because your 4 separate horn patches may not sound like they are playing back in a room together. But also, it may not matter too much as good writing/orchestration can overcome this. What happens in a real room environment when 4 horns play together is not what happens in your DAW with reverb on 4 separate horn tracks. Ensemble patches exist for the more natural joining of the instruments into a single line – but it’s also fair to say it is far more cost effective to record ensembles than to record many single instruments multiple times. = 4 Horn Ensemble You could have 4 solo horns and have a 4 horn ensemble. Having that flexibility is good, but it’s not required. If you are limited to ensemble patches, then you can still get great results if you are careful with the sound balance by using a 4 horn ensemble patch to write a 4-part horn chord. My main advice here is to ensure the top voice is a little louder than the lower ones. This is easy to do with velocity for shorts, but for longs you may need to duplicate the 4 horn patch and MIDI track and use less mod-wheel (or reduce actual volume) of the lower 3 voices on 1 track, and on the track for the top voice, have it a little louder. So, two tracks in total of the 4 horn ensemble. One is the single upper voice, and the other is the 3 lower voices. Careful listening to the balance is key. This theory can be used on any instruments, trumpets, trombones, etc. Warning – do not double the same solo instrument patch over itself to make it sound like 2 instruments are playing the same phrase. This causes phase issues because you are playing the same samples directly over itself (an example of something that would not happen in a real recording environment) Signal Flow The basics of routing in your DAW to set up the Orchestral Template. DAW Basics Here I will go through the initial setup and routing of tracks. You can skip this part if you have your own set up or preference to how you do this. However, if you are a beginner then this section could help. It’s also possible your current setup isn’t working well for you, if that is the case perhaps check some of these notes and see what suits you better. There are various ways to set up routing and this is my method. I will also explain the other methods that are popular. In my final template, the Instrument Patches are routed to Audio Tracks from Kontakt (or any other sample player such as SINE, Spitfire Audio Player, Play Engine, etc.) I use Kontakt to hold multiple patches. Each patch goes to a MIDI channel. Each patch is also Output from within Kontakt to an audio track. If you wish, you can instead use one instance of Kontakt per instrument and route that track to an audio output. It doesn’t matter which method you use. I also use a single MIDI track per instrument to hold all articulations (where possible). I prefer to use Keyswitching on a single track. For example, Flute 1 MIDI track can play shorts and longs, legato, etc. If you wish, you can use one MIDI track per articulation. A common approach is to divide the instrument into LONGS and SHORTS. Route all LONGS to one set of audio tracks and route all SHORTS to another. TIP: There are various personal choices in setting up a template. And some are choices based on delivery of audio to others. Also, different DAWs have different methods for routing. The principles are the same regardless of the software used, and so will translate across different DAWS and sample players (e.g., Kontakt, SINE, Spitfire Audio Player, Play Engine, etc.) Setting Up Instruments – (all articulation patches) 1. To start I load Flute 1 and Flute 2 into Kontakt. 2. The MIDI channels are set up to receive Flute 1 and MIDI channel #1, and Flute 2 on channel #2. 3. Kontakt is set up that the Outputs (st.2)* thru to (st.16) will output each patch on that channel so that my audio tracks can be set up the same way. Flute 1 audio and Flute 2 audio. 4. Steps 1 to 3 are repeated for the remaining woodwind section. If your woodwind section would fit into a single instance Kontakt (16 patches), you can continue adding patches. Or open a new Kontakt instance per instrument type. Alternate Method: Some find it much simpler to load one instrument only into Kontakt and create a new instance of Kontakt for each instrument. Then, simply route the audio from that track to an associated audio track. This method is useful if you wish to load up different articulations (shorts and longs) as different patches, and route the shorts to one audio track, and the longs to another. I just like using as few instances of a sample player as I can, so I prefer to group them. The choice is totally up to you! Kontakt Flute 1 (all articulations) Output: st. 2 Midi Chn:: [A] 1 Flute 2 (all articulations) Output: st. 3 Midi Chn:: [A] 2 * I find that st.1 output is used by the actual Kontakt instance, so only outputs st.2 thru to st.16 can be routed. Depending on what DAW you use, this limitation might not be present. Signal Flow – different types There are various ways to route the signal to a bus (If delivering multiple stems, you usually need a reverb set up for each bus) Delivering all individual tracks to a mixer (whether they want reverb or not should be discussed) Delivering stems (combined groups of woods being 1 stem, strings being another stem, etc.) – multiple reverb setup required Delivering stems further divided so that short articulations collect in a bus, and long articulations collect in another bus (also divided by each section, woodwinds, strings, etc. – multiple reverb setup required OR The simplest set up which I use most of the time is to have a single reverb setup for the orchestra. This is when producing and mixing music for my own productions where I avoid the need for multiple busses with multiple reverb setups. I generally have my set up ready to be routed to the stems, but if not required I turn them off and disable unneeded processing. Note! Over the next few pages, I explain the different signal flow options. What you choose (or already do) is entirely up to you. The results of the room sound will be exactly the same regardless of the routing method chosen above. Simplified Signal Flow - if you don’t need to deliver stems Most of the time, if producing music for my own productions and finished releases, I can greatly simplify the reverb setup by only using one signal flow. Audio channels send % of signal to reverb, reverb signal goes to Sub Master. Less taxing on the system, so if resources for you are tight, this set up can work. Kontakt Woodwinds Audio Channels Kontakt Brass TIP: ‘Sub Master’ is used as a track to put my final plug-ins on. The reason I have this track is in case I want a certain track (film sound effects, dialog) to bypass this processing. I can send any track I want to bypass the Sub Master processing and go direct to the Master instead. MIDI Channels % Sent to Reverb MIDI Channels Audio Channels Kontakt Percussion MIDI Channels Audio Channels Kontakt Strings MIDI Channels Audio Channels Reverb Tail Aux Sub Master Master BUS Signal Flow The process for every single instrument is the same, regardless of sample library, dry or wet. A percentage of each audio signal is sent to a Reverb Aux channel (for each bus) (more on this later so ignore for now) Kontakt Woodwinds MIDI Channels Audio BUS Audio Channels Woodwinds BUS Audio Channels % Sent to Reverb Reverb Tail Aux One for each BUS Kontakt Brass MIDI Channels Audio Channels Kontakt Percussion Strings Brass BUS MIDI Channels Audio Channels Kontakt Each BUS Sub Master Percussion BUS MIDI Channels Audio Channels Strings BUS Master How that may look in a DAW Combined Stem tracks – all woodwind audio is grouped here, including reverb, then sent to SUB MASTER Your DAW may look different, but the idea is the very same. Refer to your DAWs manual if unsure how to do routing in it. Group your sections together, and colour code them however you like. Naming and grouping tracks is very important for a faster workflow. Grouped MIDI Channels Grouped Audio Channels Send the audio to the grouped BUS Stem - Woods Track with Kontakt on it Read MIDI Channel 1 from Kontakt Combined Mix BUS Mix (Sub) Master Read st.2 Output from Kontakt % Sent to Reverb Master Reverb Tail Aux Further Busses - Dividing Short and Long Articulations If you wish to have the flexibility to delivering to a TV or Film mixer, then dividing short and long articulations is often expected so that they can be delivered as separate tracks. You can also keep splitting into more groups, increasing the granularity. Percussion can be drums, metals, melodic, and so on. This also greatly increases the requirement to have multiple reverb instances. 1 or more for each BUS. Woodwind - Longs Woodwind - Shorts Brass - Shorts MIDI Channels Audio Channels Audio Channels Percussion - Melodic Strings - Shorts Strings - Longs Woodwinds BUS Shorts MIDI Channels Brass BUS Longs % Sent to Reverb Reverb Tail Aux One for each BUS Each BUS MIDI Channels Audio Channels Percussion - Hits Woodwinds BUS Longs MIDI Channels Audio Channels Brass - Longs Audio Channels Brass BUS Shorts MIDI Channels Audio Channels Percussion BUS Hits MIDI Channels Audio Channels Percussion BUS Melodic MIDI Channels Audio Channels Strings BUS Longs MIDI Channels Audio Channels Strings BUS Shorts TIP: Remember that a % of audio signal will go to reverb, and then back to the bus for that audio group. This means that you’ll need a reverb set up for EVERY BUS you have. In this example it’d be 8 reverb set ups. 1 for each bus. If this taxes your system, you can work around this by routing 1 reverb set up to each bus in turn, bouncing each stem, route the reverb, bounce the stem, and so on. If you’re not bouncing stems, you can route the reverb to the Sub Master instead. Microphone Options Using the available mic options in sample libraries. Seating Arrangements On the right is a rough layout guide for orchestra seating. A sample library is often recorded in seating positions like these. You will hear this in the mid and far mics mainly as those mics represent the orchestra hall placements. Timpani and other Percussion If your chosen library is not recorded in seated positions (or is a fully dry library), it is likely centered, and as such would then need to be placed with both panning and reverb. I’ll show you how to do this later. I don’t recommend to use only close mics (when multiple mics are available), and don’t try to place the instrument with just that close mic. You can if you want to, I just don’t recommend it. Close mics (or spot mics) are for adding definition, and not intended to be used as the sole instrument sound. They simply don’t represent the instrument at its best (often sounds thin/harsh). There are some limited exceptions to this, but if the other mics are available, I use them in addition to the close mic. If you want a dry intimate sound, you should ideally be using dry recorded instruments recorded in a small dry room – which can then be given placement treatment later. French Horns Trombones Trumpets Harp Clarinets Piano Celeste Flutes/Piccolo 2nd Violins 1st Violins Bassoons Tuba Oboes Contra Basses Violas Cellos A basic guide only. There are many variations on this layout, and it changes from orchestra to orchestra. However, this would be a ‘typical’ seating arrangement. Note that the percussion would more likely cover the width of the orchestra across the back. Mic Options Mic options in a library vary by maker, as do the names they give the mics. Close (or spot mic) – right next to the instrument. For adding definition to a recording or bringing out a solo. Mid (or Decca tree) – above the conductor’s position. For adding placement and room sound. Far (or ambient mic) – at the back of room. Adds the most room sound. Full Mix – usually this is a combination of other provided mics. A default mix created by the developer. Lowers the RAM footprint because it’s the other mics combined. If using the 3 other mics, that’s 3 times the RAM amount. For this reason, I tend to favour the Full Mix if available. There may be even more mics available depending on the library: Different brands of mics recorded Spill mic (captures an instrument through the OTHER mics . for those other instruments) Additional close mics. Kontakt Flute 1 Output: st. 2 Midi Chn:: [A] 1 0db 0db OR Close Mid Far Full MIX Using the Different Mics for Depth Preference of mic balance is up to you. If you want a close mic to be the highest and only have a hint of mid and far that’s fine. Some people prefer drier mixes. However, remember that we’re looking to create a room here and that is best done starting with the room present in the samples. If you don’t like the provided full mix (if available), you can sculpt the sound you want from the available mics. You’re looking to hear a sense of depth, but retain a clear sound. Try to keep the relative volume levels the same as when the patch was loaded. Libraries recorded in a similar way: The depth you produce with a single section should be a set benchmark to position the remaining sections around. Start with strings, the closest section. You want these to sound full, with some sense of the room, but not distant. Then, you can balance your woodwinds to this to ensure the depth you set with the mic options available has the woodwinds just behind the strings. Brass behind the woodwinds, and finally percussion at the back. With orchestra sections from a single developer (given they are recorded in the same way or part of a single big orchestra package), this is easy enough to do. Usually nothing more than setting the same levels across all. Remember: they are recorded in position in most cases. This means the woodwinds are automatically more distant from the ‘mid’ (Decca tree) than the strings are, and closer than the brass are. So, setting the level the same, could work and should be the starting point. However, you should still do this by ear – because if from different developers, the room sizes aren’t likely the same – the distances aren’t consistent. Libraries recorded in different halls or by different developers: If you have 4 different sections by 4 different developers, woodwinds, brass, percussion, strings, the depth balance (and your mic options) might be all over the place. The best you can do here is strike a depth balance without washing out any one section. Even if it means with no other plugins used, the woodwinds (for example) are now sitting in front of the strings, this will have to do (for now). We’ll push them back later with plugins to fix this. It’s important to continue to balance the orchestra with no plugins yet. Just watch you do not have a badly defined section sound (too distant) because you turned off a close mic, or only use the far mic to get it behind the strings. That is doing more harm than good. 0db Close Mid Strings Far Close 0db 0db 0db Mid Woodwinds Far Close Mid Brass Far Close Mid Percussion Far Orchestral Balance Adjusting the levels to balance the orchestra sections. Bringing Balance to the Orchestra 1 Ok, you’ve loaded all your instrument patches, and routed the tracks. What now? The plan is to balance these instruments against each other so that they are blending with each other in the same way they would be during a performance. For example, how a flute playing ff would sound against violins playing ff . The aim is for a balanced playback sound. If you are using an orchestra from a single source, let’s say the BBC Symphony Orchestra from Spitfire, you should already have a balanced orchestra (not that it is perfect) – but it would be a good start that might need only a few adjustments – perhaps the same adjustment each section just to bring the levels down to mixing volume. If you have a collection of libraries from different developers (like I do) then different balancing levels of the orchestra may be required. This is to achieve a general balance across the 4 sections, but wouldn’t necessarily negate the need for volume automation during the composing process. That’s normal - a mixer with a live recording doesn’t just leave the faders idle! There’s no exact science to this, but let’s look at this in more detail on the next page. 0db 0db -14 db Strings Bus 0db 0db -12 db -20 db Woodwinds Bus Brass Bus -12 db Percussion Bus Bringing Balance to the Orchestra 2 These are my starting points when I was setting up the template and are a basic guide only. There’s always variance with what you might be using that could change the results but the main goal is having headroom of around 6db on the master bus when a lot of the louder instruments are playing together at max dynamic. Percussion: Play a timpani hit at max velocity (or a roll at top dynamic). Try a G2 note (taking C3 as middle C). Note where it hits on the meter of the percussion bus. Get this to hit at its loudest around -12db by adjusting the percussion bus volume. The percussion bus, not the timpani track. Your instrument tracks should remain at 0db. This -12db on your percussion bus is your base volume level. You haven’t made the timpani quieter (and must not balance all other percussion against it), what has happened here is you have made the whole percussion section quieter – but maintained its own balance within itself. The only caveat here is if you are bringing in other percussion patches from other developers. You may need to check how they sound. But again, this is difficult unless writing a piece (where it can become clear that something is out of whack volume wise) – if so, you adjust that one track to balance it with the rest). But first, let’s just balance the busses to create headroom, before getting into the tracks. Brass: Play a chord in the trumpets, say a high G (G4, B4, D5), max cc1 (or velocity – but I find sustains are much easier to read. On the Brass bus get this to a max of -12db. Strings: Take the cellos and the basses together and play a low sustained E1. Get this volume to around -14db Woodwinds: This is harder to gauge as it is more difficult to get the woodwinds to project. What I do is I play a flute and violins 1 patch together at max dynamic. I want to only just about hear the flute blending with the strings once you get past G4 and up. And it will probably be heard a little more as you go up the register, becoming a little more obvious there is a flute playing along. So, this means you’re adjusting the volume of the WOODWINDS bus, not the strings, to get this blend. There’s no telling what the woodwind bus will need to be adjusted to here. Maybe it doesn’t even need to be adjusted at all. In my setup, with no adjustments, the flute (and other woodwinds) hit around -20db with no adjustments, so I left it. You can then also play a clarinet along with the viola section, and ideally, they should blend as well now with no adjustments. If you find the clarinet is too loud, it’s time to just adjust the clarinet track down instead (the bus is already set via the flute and strings). It may not be perfect, but it’s a start. You might even find these figures don’t work for you. If using different patches from different developers, then you may need to repeat balancing steps adjusting another instrument against the one in the same section. I suggest adjusting the other woodwinds track levels this time (not the bus) to sit well with the flute and do not change the flute (as you balanced that against the strings already by adjusting the bus). Ignore this step completely if you are using woodwinds from a single library. This principle also applies to the brass but note that trumpets are going to be weak in their lower register, trombones are going to be very strong in the same low register the trumpet can play, so these can’t really be judged and balanced together. You must use your ears. 0db 0db -14 db Strings Bus 0db 0db -12 db -20 db Woodwinds Bus Brass Bus -12 db Percussion Bus Bringing Balance to the Orchestra 3 Taking C3 as middle C, this is my rough guide to blending different instruments. There’s no exact science, and differences between libraries and their own dynamic ranges makes this more difficult to judge in people's personal setups. Some of these balances can be adjusted in mixing, but I try to have them close to this while writing. The needs of the music I’m writing will dictate if I simply disregard a suggestion here due to what I’m looking to achieve in a phrase. Blend with: Balance Dynamic – with legato* Register Loudest Instrument Flute 1 with Violins 1 At highest dynamic have the flute blending under the violins C3 to G4 I do not expect to hear the flute - you should start to hear the flute a little more as you rise through the scale approaching G4 and beyond. Violins 1 Clarinet 1 with Violas At highest dynamic have the clarinet blending under the violas Around middle C I do not expect to hear the clarinet clearly, but you should know it’s there underneath. Violas French Horns with Cellos Mid dynamic they should blend together, it’s a great sound. Higher dynamic, expect the horns to start taking over the sound more. Around C2 at mid dynamic, I want to hear the cellos a little more, the horns are adding a blending tone. As you rise in dynamic and the horns become more powerful, they should be easier to hear – maybe even becoming dominant. Equal at mid dynamic Horns high dynamic Highest dynamic. Low trumpets (below and around C3) are not very powerful, and violins should lead here. As the register increases the trumpets become more powerful and should become more obvious (but still blending) as you start hitting E4 and up. Violins low register High register – both blend and trumpet is clearer. No clear winner here, but both should be obvious enough they are playing. Blending in low and mid. Then timpani takes over in higher. Trumpets with Violins Tuba with Basses Highest dynamic. A solo tuba against the basses section – doesn’t seem fair. Tuba should be clear enough through all the dynamics and register. Not dominating, but clear that there’s a tuba playing as well. Overall, the tuba is one instrument I’m always adjusting during a composition. Timpani with Cellos and Basses Use HITS/SHORTS for this test. All dynamics. Timpani and the strings here should be clearly matching each other in low dynamics, but as the dynamics rise, I expect the timpani to start dominating. You will often see timpani scored at forte, when strings are at double forte. You get the idea here, and the balancing should ideally be done on the busses, because the sections should be balanced within themselves – but if you find there’s a disparity between some instruments, then you can use the track fader instead. Again, if you want to swap out your trumpets for another library, you need to balance those new trumpets by adjusting its levels against the old trumpets, don’t reach for the bus again. *Legato - In my experience, many instrument patches in legato mode tend to represent a mid to high level dynamic rather than full raw power. This is ideal for gauging the balances above because in most cases, legato playing will usually be a blend, instead of one instrument completely dominating another. Shorts and other articulations aren’t really suited to this, because a short high trumpet has that immediate loud attack that would be heard over just about anything else. Some Basic Instrument Characteristics General guidelines and basic roles for the instruments. Woodwinds – Basic Instrument Roles Starting with the Woodwinds, these are some generalizations on the use of those instruments in the orchestra. Mostly related to common use from John Williams (as a starting point). There are so many more uses - this is just some basic beginner use from an orchestration point of view. Instrument Common Role Common Doubling or Unisons Flute Scale runs (often as colour, over sustains, gaps in the melody and as transitions). Solo melodies, and adding colour with high piercing flurries. Violins, piccolo, and other woodwinds, piano (runs). Piccolo Scale runs (often as colour, over sustains, gaps in the melody and as transitions). Adding colour with high piercing flurries. Violins and flutes, piano (runs). Oboe Solo melody. Scale runs in the middle register and rhythmic support. Trumpets, other woodwinds, strings. English Horn Solo melody. Scale runs in the middle register and rhythmic support. Horns, other woodwinds, strings. Clarinet Solo melody. Scale runs. Harmonic support (often two clarinets used to form a triad or other harmonic interval). Rhythmic support. Horns. Violas, other woodwinds. Bass Clarinet Bass support, rhythmic support. Trombones, basses. Bassoon Solo melody. Bass rhythmic support. Cellos, basses, trombones. Contra Bassoon Bass rhythmic support. Bassoons, cellos, basses, tuba, trombones. Doubling – same notes but supported at a different pitch. Usually, an octave above or below. If you double the flutes with the violins, you are deciding that one of them will be an octave above the other (or two). Unison – same notes, and same pitch. This represents a single idea or phrase, across multiple instruments. If you have the flutes in unison with the violins, you are deciding that they are playing the same pitch. Brass - Basic Instrument Roles Some general characteristic use of the brass section. Instrument Common Role Common Doubling or Unisons French Horns Melody, solo and in unison. Rhythmic support (often in chords), sustained harmony support in chords (low dynamics for soft passages). Cellos in melody (great sound). Trumpets (low melody). Clarinets. Trumpets Melody, solo and in unison. Rhythmic support. Rhythmic motifs and support during melody breaks and transitions. Violins, horns, trombones (low melody). Trombones Rhythmic support. Rhythmic motifs and support during melody breaks and transitions. Sustained builds for tension. Harmonic support in chords (low dynamics for soft passages). Cellos, basses, trumpets (low melody), bassoons, contra bassoon, and horns. Bass Trombone Useful low support for the tuba and trombones. Can often play the role of trombone number 3, as the lower voice. Trombones (part of the section), tuba, basses, cellos, bassoons. Tuba Bass support with sustains and rhythmic support. Often to punctuate an idea the tuba is often resting for moments before joining the brass section to thicken the bass. Trombones, basses, cellos, contra bassoon. Percussion - Basic Instrument Roles Some general characteristic use of the percussion section. Instrument Common Role Common Doubling or Unisons Timpani Bass accents, rolls for transitions, dramatic builds, orchestral pulse (repeated note to drive a beat). Solo flourishes. Low instrument accents, basses and tuba, etc. Bass Drum Bass accents, rolls for transitions, dramatic builds. Very impactful when used to accent 2 nd beat (not first). Low instrument accents, basses and tuba, etc. Snare Drum Traditionally military and march, and can also provide drive and colour. Various snare sizes can add different colours. A snare and bass drum hit is a good combination. Concert Toms The varying pitches creates interesting fill moments. Very attack heavy and less resonance. Suspended Cymbals Colour, texture, suspended rolls for build up and transitions. Often used in combination with a timpani roll. Piatti Crash cymbals for accents, colour and rhymical support. Sometimes used in combination with a bass drum. Tam Tam Rolls for transitions, dramatic builds. Big booming sound. Triangle The triangle’s distinctive ring can add top end accents and some sparkle, especially if used as a tremolo. (Such as the opening of Star Wars) Mark Tree The magical sparkle you hear in many fantasy and mystery orchestral pieces – mostly for adding colour and can be used for transitions. Toys Woodblocks, shakers, vibraslap, whip, and more. Lots of percussion colour options. Best used sparingly. Harp, Keys and Mallets Some general characteristic use of the harp, piano and other keys, and the mallets. Instrument Common Role Common Doubling or Unisons Xylophone Often used to add a percussive element and colour to high accents in the woodwinds. Often with high woodwinds. Marimba Can add a slightly exotic flavour and melody to a piece. Glockenspiel High ringing attack can help bring out a melody even with the whole orchestra playing. Shorter passing notes might be ignored, and the harmonic main notes of a melody doubled. Commonly with brass. Often doubles the main notes of a melody (not every note of a phrase, the most important ones). Vibraphone Often used in quieter moments, to add mystery (commonly as a two-note harmony). With motor on, adds a slight sustained movement to a chord. Common with harp, piano or celeste. Celeste Colour, texture and adding mystery or sparkle (in the high end). Famously used for the Harry Potter main theme. With harp or piano. Harp Chords, and arpeggiated chords for harmonic support, glissandos for transitions and key changes, and sometimes a main melody. Commonly with the celeste and the piano. Piano Bass support (low notes add a lot of percussive weight), melody, harmony support (chords or arpeggios). Runs (often doubles the woodwinds for this). Percussive accents. Often with the harp, and woodwinds (runs) or percussive accents. Strings Strings are the backbone of the orchestra and fulfil many roles. Every section can be used for harmony, melody, rhythmic drive, and so on. The best use of the string section is to utilize their registers in your orchestration and avoid simply copying one part into the next and avoid doing constant tight triads (playing it like a piano). Instrument Common Role Common Doubling or Unisons Violins 1 Around 16 players. Harmonic and rhythmic support, melody or unison with both violas and violins. Violins 2, violas (often unison in melody), flutes. Violins 2 Around 14 players. Harmonic and rhythmic support, melody in double or unison with both violas and violins 1. Violins 1, violas, flutes. Violas Around 12 players. Great for rhythmic support and drive, filling middle harmony, and supporting big melodies in unison with the violins. Violins 1 and 2, cellos, clarinets. Cellos Around 12 to 10 players. Great for melody with the horns around middle C. Big arpeggios for movement in sweeping melodic passages. Also great for supporting big melodies. Violas, basses, horns, bassoons. Basses Around 8 players. Sounds an octave below the cellos. Aside from standard bass support, great for pedal tones, pizzicato notes for light rhythmic accents and movement. Bassoons, contrabassoon, tuba. Placement for the Orchestra Sections How to achieve depth and placement Depth and Orchestra Section Placement Let’s say you’ve got your main orchestral sections set up, mic placement chosen (if available) and the volume levels reasonably balanced. We’ll now need to tackle a couple of problems that could be present: One or more orchestral sections is too close (note, if a section is too far at this stage, then there’s something wrong in the earlier setup, using too much far mic, or volume is too low – that needs fixing in the earlier stages, not here). If you are using a dry recorded solo instrument it’s going to sound out of place with an orchestra, particularly if using close/mid/far mics for your sections. We’ll adjust this instrument placement after the rest of the orchestra is placed. I recommend you place entire sections before you start placing solo performance instruments. Let’s start with getting sections to sound placed and help push back those that are either too present and up front - or are fully dry recorded sections/instruments and so must be placed. Percussion Brass Woodwinds Strings Front Help me, my Woodwinds are invading my string section… Let’s explore this topic with an example that can be applied to any of the sections or individual instruments if needed. After using your mic placement, let’s say you have a woodwind library that sounds too close, even with the mic options it has. For the moment, I’m not counting this as being a fully dry set of woodwinds yet – I’ll explain what I do for that situation after this. Here, I don’t want my woodwinds to sound washed out, but also, I would prefer if they sounded a little further back where they would usually be, middle and centre. It’s likely if I ended up with this situation, the woodwinds can’t even sound washed out. They might have been recorded close and present (even if in a hall). I always recommend trying to judge the placement from the strings point of view. However close you have the strings sounding, you want the woods and the other sections to be a little further away. For this – it’s time to talk plug-ins… Timpani and other Percussion French Horns Trombones Trumpets Harp Tuba Piano Celeste 2nd Violins 1st Violins Clarinets Flutes/Piccolo Contra Basses Violas Oboes Bassoons Cellos The woodwinds have infiltrated my string section, and if I know my music theory, they’ll eat the string players unless we do something about it. Fixing the Woodwind Section Placement (1) It was possible until now to avoid being specific about using certain plugins. In fact, that’s still possible as the principle of how I use the following plugin is not much different to other plugins that are designed to help add depth in a 3D space. I just find this one works well with no fuss. Panagement 2! Woodwinds MIDI Channel Audio Channel Sub Master Woodwinds BUS I place this on the Woodwinds BUS, so that all of the woodwinds are going to be positioned together as a section. This is based on my using a consistent library of woodwinds – meaning - I don’t have a rogue flute for example from another source that doesn’t match the tone and levels of the rest of the woodwinds. If I did, I might need to consider treating that flute on its audio channel before it hits the BUS. Anyway, back to the now. On left is Panagement, default settings before I’ve touched anything. I am going to need to do a number of tweaks here to get the woodwinds ‘seated’ correctly. Let’s try that now… Master Fixing the Woodwind Section Placement (2) The woodwind section sits a little better in the centre when the stereo width has been narrowed – just a little. Very important – turn OFF the Reverb here, I want to use my own reverb later. This is the main placement setting. By dragging this circle around, you can hear the effect of the positioning in a 3D space (hold down a woodwind note while moving it around and you’ll see). This is where I’m trying to place this icon in such a way that I hear the woodwind section is now placed just behind the strings. Some trial and error is involved. To compensate with the woodwinds ‘fading’ as they are pushed back, you may need to adjust the volume output. Here, for example, I’ve pushed this up to compensate. I don’t want to lose my volume balance, I want them to sound seated back, but still balanced with the other sections. What if I’m Using a Completely Dry Instrument? Ok, our string section is now saved from cannibalistic woodwind players, but what if we have that solo instrument we want for a violin concerto? The completely bone-dry one that doesn’t belong? It’s very common to have a solo violin recorded as a dry instrument. The principle behind adding a completely dry instrument into what is hopefully now a well balanced and positioned orchestra is basically the same regardless of the instrument. To get a dry solo violin to sit well with the orchestra can take a number of treatment steps specifically just for that instrument. This means we’ll be adding plugins directly on the audio channel for that instrument. SOLO Violin MIDI Channel Audio Channel EQ Convolution Reverb Panagement 2 Strings BUS The main must have plugins here would be EQ and the Convolution Reverb. I’ve added Panagement 2 at times as well, because it’s very useful to add a final touch of placement to an instrument. Using the Plugins to Place a Dry Instrument (1) EQ The need to EQ is dependent on the instrument. Some principles can apply to all though, and one of those is rolling off the higher frequencies to simulate some distance. You may not need to do this. Use those ears! Another slight issue I had with the violin was the sound was a little harsh (sometimes a symptom of a close dry recording). To better make it sit in the hall, I EQ’ed out some of the harsher frequencies. I apply the EQ first because if I want to EQ out some harsh tones, then I want that to be taken out before it goes into the convolution reverb. Otherwise, I would then have to deal with the reverb's treatment of those frequencies. Opinions on this may differ. Convolution Reverb Next is a convolution reverb. This is to simulate the room and give a sense that the instrument has ‘mic’ positions added to it. I leave the early reflections AND the tail on. I found turning the tail off creates some sort of ‘disconnect’ with the room. Your experience might be different. Remember that this is placed directly on the dry instrument, so it’s likely to sound very washed out and ‘dream-like’. You must lower the MIX level of the reverb, so that it is not 100% wet. I found only 25% wet was good enough to get a dry violin to sound good with the orchestra. The % used is going to be dependent on what it takes to get that instrument sitting well with your other sections. Either, you are trying to get a dry instrument to match a section (that rogue flute I mentioned), or, you are trying to add a solo performance instrument (usually playing at the front of the orchestra) but still sound like it’s in the hall with them. Using the Plugins to Place a Dry Instrument (2) Panagement 2 Finally, and only because I felt it was useful, I added Panagement 2. I again turned off the built-in reverb and reduced the stereo width by a good bit (it’s a single instrument so I want it to be narrow). I played around with the positioning grid until I heard a sound and placement I liked – in this case it was slightly to the left and forward. Of course, none of this matters without testing how it sits with the orchestra, so when writing a piece, it’s always possible you might need to revisit and tweak something in this chain of EQ/Convolution Reverb/Panagement to get the result sounding like it’s naturally in the same ‘space’ as the orchestra. Unfortunately, it’s not always a case of ‘set and don’t touch again’. SOLO Violin MIDI Channel Audio Channel EQ Convolution Reverb Panagement 2 Strings BUS Final Reverb Settings The final glue… Reverb Tail (1) After successful placement of the orchestra, it’s possible to just stop there, and some people do – letting the natural mic placements and ‘room’ sound in the samples be the reverb. Even live musician recordings get digital reverb added, however, and there’s something very off-putting to me to hear a sampled instrument just…end. It doesn’t even seem to sound like a real instrument note finishing (perhaps a result of how release samples are programmed – the room sound is in the samples, but is the natural room tail sometimes cut off in the processing of the note? I’m not sure.) To mask this issue that I hear, I put a reverb tail on the end, and it can be as short or as long as you personally like. Once it masks that odd dead air cut off at the end of the samples it’s good enough for me. I’m demonstrating this using the simplified audio routing method (on page 14). Audio signal goes direct to the sub master channel A percentage of this audio signal goes to the Reverb Tail (first passing through an EQ, then into the reverb), then on to the sub master Which brings me to the very important point about EQ’ing the reverb tail signal…let’s look at this more closely. MIDI Channel Woodwinds Audio Channel % Sent to EQ/Reverb EQ Reverb Tail Aux Sub Master Master Reverb Tail (2) Audio Channel EQ The EQ here is taking the percentage of the signal I send and shaping it before it hits the reverb. I’m using a rough version of what is known as the ‘Abbey Road Curve’. Cleans the reverb signal by quite a lot, by removing the low end (rolling off from around 300hz and down (you can tweak this) and the high end (rolling off around 6khz region). The reverb tail is entirely up to you – adjust to taste. Must be 100% wet. Reverb Tail Aux Reverb Tail (3) To summarize how I use reverb: For the Reverb tail I use an algorithmic reverb. I find this is best for a bright clear tail. Convolution Reverb is used only to place dry instruments into the hall so they sound part of the orchestra. I want to avoid that odd ‘sucking’ note drop-off effect, so the amount of signal I send from an instrument into this reverb tail is designed to eliminate that. This means the amount sent will likely differ per section (or instrument), but also note here this doesn’t necessarily mean that just because a trumpet is meant to be further away from a violin that it should get more reverb tail. It doesn’t – you’ve already distanced or spaced the orchestra. Adding tail is a global reverb ‘icing on the cake’ – don’t apply the same distance rules here. Do what sounds right, not what might seem correct on paper. Also note that this is the ONLY reverb tail I have for the orchestra (ignoring the convolution reverb that might have been used to place a dry instrument – remember I consider that placement, NOT specifically reverb in the sense of a tail). There’s no multiple reverbs here for close, mid, and far tails – just this one single reverb. I have found results using multiple tails for distance rather pointless. It doesn’t do anything at all that I can personally hear. I would suggest don’t bother doing this – it’s another one of those ‘seems good on paper’ ideas that brings nothing worthwhile to the table. I’ve tired this and after a time I realized I was just creating additional resource use by using multiple reverbs tails with different settings for close, mid and far, and when I removed them it made no difference (or I could even argue it cleared things up by removing it and made placing the instruments much easier). The Final Mix and Master Let’s say you’ve finished composing your masterpiece! You have a good balanced orchestral sound, mixed with automation where needed and you want to produce the final master yourself. Most people recommend if you’re 100% happy with the MIDI side of things, bounce your audio stems out and make the final mixing decisions here. I’m not going into this area as it’d be a document twice this size of this already, and you know what? There are people far far better than I for explaining how to go about creating that superb mix and even into the dark art of mastering. I will however just drop some closing notes on what I do: My combined submaster track is used to place my final plugins on. These consist of (in order): A simple gain plug-in that I can use to boost (or reduce) the overall volume of the piece. I usually ride this while listening back through the piece to find the peaks and dips I want the piece to have. EQ – corrective EQ that might be adjusted based on the needs of the track. Too bass heavy, not enough top end? It depends on the track and is adjusted as needed. Tape saturation - there’s something about tape saturation that adds a small layer of realism over a whole orchestral mix for me. I notice it when I turn it off, there’s a slight dulling of the sound that I don’t want. Be careful that you don’t over do it. Like with many plugins that change the colour or tone, you stop ‘hearing’ the effect after a while mixing, so you must not respond to this by adjusting it again. What you heard and liked when you first turned it on and tweaked it is what you wanted – now leave it alone. Finally – a limiter. A simple limiter to stop any rogue peaks clipping. And that, as they say, is that. I hope you found some useful information in here (even if you disagree with some of it (or all) – that’s ok). It’s how I’ve done and enjoyed doing things for some time now. A simple workflow (all things considered). Some Parting Words I hope there was something useful in this guide for your orchestral needs. There are a few ways to approach things, and opinions vary on the usefulness of those approaches. This is just my method, honed after years of trial and error. Whatever approach you use, I just want to say, happy composing! Graham Plowman from the Virtual Orchestration Facebook Group