Orchestrating with Sampled Strings
Modern string libraries have ushered in astonishing levels of realism for computer-based orchestration. Mark Cousins unlocks the secrets of the virtual string section.
Of all the instruments in the orchestral palette, it seems that strings command the most continual fascination and use in modern music. In truth, there are few styles of music that don’t benefit from the strategic addition of a good string arrangement, whether it’s a chugging cello in an indie track, some soaring violins in a ballad or an aggressive pizzicato punctuating a hip hop loop. While most of us would love to have access to a real string section to achieve this, many musicians and composers – particularly those who work only at home – now turn to virtual solutions as a means of re-creating the vibrant sound of a string orchestra in full flow.
However, having access to the best samples libraries in the world doesn’t necessarily guarantee that you’ll produce a realistic and believable output. Working effectively with sampled strings requires detailed knowledge across several disciplines, from musical decisions that you make while piecing the arrangement together through to MIDI sequencing and sample techniques – and, of course, the art of mixing. Although good production skills should always be transferable, there is a range of techniques and practices worth learning that relate specifically to the world of sampled strings.
Finding your voice
Before getting too engaged with your samples collection, it’s worth taking some time to think about how the strings will work in your given track, as well as forming a basic understanding of the voicing. In this respect, one of the biggest mistakes is simply loading up a generic string patch (like a sustained string sound, for example, or some pizzicato) and playing this sound just like any other keyboard part in your song. Rather, you need to start thinking like a group of individual players – violin I, violin II, viola, cello and double bass – with each line having something meaningful to contribute to the overall effect.
As an example, let’s consider some simple chordal movement in the strings. Once you’ve played the sequence, take a look at it in your DAW’s piano roll editor. Rather than see it as a collection of chords, look towards visualising your sequence as a series of single monophonic voices. As an initial consideration, therefore, check that you use a consistent number of voices for any given part of the track. You might, for example, use just three voices (violin I, violin II and viola) for the introduction, adding in the final two voices (cello and double bass) once the chorus kicks in.
As well as the number of voices used, you also need to look closely at the movement within each voice. Again, the tendency is simply to play a series of block chords, but this will immediately reveal the keyboard-based origins of your string section rather than sounding like a group of real players. Avoid so-called ‘parallel motion’, whereby two or more voices move in the same direction using the same interval; instead, start to think of each voice having its own direction and rhythm – holding notes across bars, for example, or even playing melodic fragments.
Becoming articulate
Sketching out the foundations of your voicing and arrangement shouldn’t require more than a basic string patch or even a piano to get you started. Keeping your sound palette limited at this point will allow you to get the basics right – elements such as the melodic movement, voicing or harmonic progression – rather than being distracted by myriad further sonic possibilities. However, once we’ve formed the basic arrangement ideas we need to start considering how our virtual string orchestra will articulate its performance.
One of the key characteristics of string writing is the use of multiple different playing styles to add colour and expression to the music. To cater for this, most string libraries include a number of different sampled articulations, split between long articulations – including legato, tremolo and trills – and short articulations – covering playing styles such as staccato, spiccato, pizzicato and col legno. In short, the more articulations you have access to the greater the degree of expression and colour you’ll be able to bring to your score.
From a technical perspective, the issue with articulations is how to move between them as part of your MIDI sequence. In the case of a written score, all you need to do is notate the appropriate playing style (whether it’s legato, staccato or pizzicato, for example) and the player will effortlessly move from one articulation to the next. In the virtual universe, however, we need to consider ways in which we can move between the different articulations within our MIDI arrangement – something that may require banks of virtual samplers and a range of MIDI-based switching techniques.
Switching places
One of the simplest techniques for articulation switching is to load separate articulations onto individual track lanes – for example, a cello legato would be on track 1, a cello spiccato on track 2, and a cello pizzicato on track 3. Switching between the articulations simply requires the appropriate MIDI region being assigned to the corresponding track lane. As you’d expect, there are advantages and disadvantages to each approach. In the case of multiple track lanes, it’s easy to see how the various articulations can soon consume your screen real estate; as an upside, at least it’s easy to see how the articulations are being used.
As a variation of the multiple track lanes approach, you can always use the multi-timbral capabilities of your software sampler and simply change channel assignments within your MIDI sequence. This approach certainly makes your arrangement easier to see and also keeps the related notes of a musical passage together as part of a single MIDI region (especially if the part is moving quickly between staccato and legato articulations, for example).
Of course, the other principal means of switching between articulations is to use key switching, which is a technique preferred by many of the leading string libraries. Key switching uses either the lowest or highest keys on a MIDI keyboard to move between two or more articulations within the same patch. The clear advantage with key switching is the ability to move between articulations on-the-fly, either as you’re recording the passage or as part of a second pass. However, visualising the articulation switching can be tricky, although a solution is to run a second track lane with colour-coded regions to show the key switching taking place on ‘arrange level’.
The long and short of it
As well as the broad articulation switching techniques already discussed, we also need to consider some important differences between long and short articulations. One of the big problems with short articulations is the so-called ‘machine-gun’ effect, whereby rapid retriggering of the same sample can lead to the sequence sounding mechanical and repetitive. To eliminate the machine-gun effect, most libraries provide alternate takes of short articulations so that no two successive notes trigger the same sample. Switching between versions is performed either manually, with key switching or, in most cases, using a ‘round robin’ that automatically switches between the variations on-the-fly.
The main sequencing consideration in the case of long articulations is the use of a ‘true legato’ on sustained strings patches. A true legato patch analyses the MIDI sequence as it’s playing and identifies the interval movements that occur within the performance. As well as playing the straight sustained string sound, the legato patch also inserts a small ‘transitional’ sample taken from the real players moving between the two notes (from D to E, for example, or D to G). In theory, the result is a more natural legato sound, mimicking the way in which a human player glides from one note to the next.
Getting a legato passage to sound right can be tricky, although the results more than justify the extra effort involved. The key is to make sure that the MIDI sequence contains the right mix of played legato – where the duration of the last note ‘bleeds’ into the next – as well as notes that are distinct and separate. Some libraries also enable you to change the speed of the legato, moving between a small transitional ‘blend’ and a more pronounced gliss effect. In the case of LA Scoring Strings – which uses this transitional legato to good effect – the speed of the legato is controlled by velocity, which can take a while to get used to.
Getting dynamic
The final piece of the sequencing puzzle is the consideration of dynamics. Although most engineers seem to be intent of removing all dynamic range from a piece of music, it’s important to note just how vital dynamics are to a good string arrangement. On a simple level this might involve contrast between pianissimo (quiet) or fortissimo (loud) sections of your arrangement, or the so-called ‘micro’ dynamics in respect to the shape of an individual note (whether it gradually dies away, for example, or begins quiet and then crescendos). Adding these dynamic movements are just as important as articulation switching, and can form a fundamental part of the overall musicality of the end result.
Controlling the dynamics of short articulations is easy, with note velocity either controlling the overall volume of a note, or moving between pianissimo and fortissimo variations of the current articulation. Indeed, if you’re using short articulations it’s well worth looking at note accents and how these affect the musicality of the results. For example, you could choose to accent the main beats of the bar, or create a more unusual result from accenting a syncopated pattern, such as two groups of three followed by a group of two.
The dynamics of long articulations can be controlled in a number of ways. As well as velocity – which sets the initial dynamic – you might also find that your patch features some form of mod-wheel crossfade, using the mod wheel to move between different dynamic levels. A so-called mod-wheel crossfade is a great way of adding natural expression, whether it’s small amounts of note contouring or more dramatic crescendos and diminuendos. Failing that, don’t rule out some basic automation moves, shaping the notes using changes in volume, as this can be almost as effective as modulation crossfades if applied correctly.
The orchestral soundstage
Although many of the string libraries sound great ‘out-of-the-box’, it’s still worth using a variety of mixing tools to help your overall arrangement gel effectively. The first point to consider is the stereo soundstage, ideally panning your various samples so that they form a natural and coherent soundstage. If you’re following the conventions of a traditional orchestral layout, the first violins should sit stage left, the second violins around 11 o’clock, violas at 1 o’clock, cellos stage-right and the basses slightly behind the cellos. Some libraries will present the samples ‘in-place’, although you might need to perform a degree of tweaking to shape the soundstage as you see fit.
Both EQ and compression can help you to shape the overall sound in a variety of ways. Adding top-end ‘air’ (12kHz or so) can help bring out bow movement, although you’ll need to take care so that the sound doesn’t become too harsh. Control the low-mids on the cellos/basses at around 200Hz and consider a small amount of bass boost around 80Hz if you need to give the low end a bit more power.
Using some strategic compression can help to glue the various articulations together, arguably using a soft ratio (around 2:1) and low threshold so that the strings are massaged together with just a few decibels of gain reduction. Using harder compression (4:1 or so) on some of the short articulations can also help them to sit in the mix in a more consistent way – particularly useful on fast-moving spiccato lines that can occasionally lose notes in a busy mix.
The crowing glory is, of course, a sumptuous concert hall reverb for the strings to sit in. When it comes to strings you’ll find that one good reverb (applied across the strings en masse) usually does the job. As you’d expect, convolution reverbs deliver the most lifelike results, although the ‘colour’ of the reverb’s output can occasionally fight against what you’ve achieved with the strings. My favourite is an algorithmic reverb like Lexicon’s PCM Native (or, preferably, dedicated hardware like the Bricasti M7).
String theory
Achieving a realistic virtual string sound takes time, but given the quality of today’s libraries and the speed of computers, you can create great results with relatively minimal outlay. Along the way you’ll stretch your DAW’s MIDI sequencing, sampling and mixing capabilities in ways you didn’t think were possible, and you’ll also gain a valuable insight into a musical tradition that has developed over hundreds of years. Despite all the sounds that the modern-day studio can deliver, strings still possess an elusive charm that few musicians, engineers and producers fail to resist.
This feature first appeared in Music Tech Magazine issue 102
Filed under
Features
Sign in to download this article
New users, please register here
See also...
 MTM 107 |
25 Tips for Mix Processing |
MTM 107 |
Control for Live Performance |
|
MTM 106 |
25 Pro Tips for Kick-starting Composition |
MTM 106 |
Contemporary Mastering |
|
MTM 105 |
Re-creating the Sound of Old-skool Dance |
MTM 105 |
25 Tips for Live Sound |
|
MTM 105 |
Mixing Tips from the Industry Pros |
MTM 104 |
Creative Distortion |
