StringSection Blog

Posts Tagged ‘overdubbing strings’

Overdubbing a 30 piece String Orchestra

Wednesday, June 18th, 2014

In May this year we took on one of our biggest ever jobs: namely, to record a 30-piece string orchestra for 12 songs. The client is a well known music producer and songwriter and he had contracted the services of string arranger Samuel Kimuli. Samuel originally hails from Uganda and is a wonderful pianist/organist, jazz improviser, composer and orchestrator. The challenge was to overlayer and create the effect of a 30 piece orchestra using only a handful of players (violins, violas, cellos and a double bassist).

 The string parts were very rich and a main feature of all of the tracks so our first job was mapping out the size and proportions of the strings. We settled on 6 first violins, 6 second violins, 6 violas, 6 cellos and 3 double basses. Normally tracks need more instruments in the higher registers (as the effect is naturally thinner), but as we recorded the violas and cellos simultaneously (in adjacent full soundproofed booths, meaning that there was no ‘spill’ from one to the other) we decided to do 6 cello overlayers at the same time as the 6 viola overlayers.

Tuning is must be absolutely impeccable when recording this way. When producers or songwriters have the luxury of 30 orchestral musicians simultaneously playing and recording, these musicians will make constant subtle adjustments to their tuning in order to ensure that the overall intonation remains very tight. When different parts are overlayered separately, the danger is that small deviations in pitch start to accumulate and by increments, the overall effect starts to sound out of tune of fuzzy. This can actually be exacerbated by many recent developments in recording techniques that have meant a use of autotune to regulate and flatten out pitch. One of the joys of listening to an acoustic ensemble such as a string quartet is that the four instruments aren’t quite perfectly in tune in an equal temperament. It is the minute and constant adjustment of intervals (by widening and narrowing within the emotional context of the music) that makes these performers sound so ‘soulful’ and tonally expressive. In the more uniform world of pop records, it is incredibly important to keep the tuning totally pristine, otherwise once it is placed within the track, even fractional errors will be noticeable. If the vibrato of the strings is narrow enough, autotune may be applied in post-production in order to align slight errors in tuning, but if the natural vibrato is thicker, this can disturb any application of autotune, causing the notes to ‘wobble’.

The whole process of overdubbing requires immense concentration and a rigorous analysis of each take before moving on. Just as when we perform as a full section, we had to ensure that the strings sounded really warm and lush, with natural phrasing, tight rhythm, tuning and a feel which many former clients have referred to as ‘heart and soul’. This is the nicest compliment to receive as when hiring professional session musicians, all the other qualities are taken for granted. This after all, is why people often reject samples and look for real strings: they are capable of giving a track an extra dimension of emotion and class. 

I’m very excited at the prospect of hearing the end result as the vocalist in the tracks had a fabulous, powerful voice, the songs were really brilliant and production values high.


Recording Strings at the Premises Studio, London

Thursday, September 19th, 2013

On Sunday we were hired to play for a recording session for Jim Perkins (of Bigo & Twigetti) at The Premises Studios in Hackney. It was the second time that String Section have recorded for Jim and it involved laying down strings for three tracks. Two were compositions written by him (one, an atmospheric piece for a film soundtrack) and the third consisting of strings he had arranged for a pop song written and produced by Tony Holland. The only two musicians present were myself (on violins and violas) and Tony Woollard on cellos. 

The first piece was written for double string quartet (so, 2 first violins, 2 second violins, 2 violas and 2 cellos). This is a particularly exposed combination as any discrepancies in tuning or timing become very obvious. Also, the tone quality on each of the two unison parts needs to be well matched. When overdubbing strings in this type of scenario it’s always best to keep the vibrato fairly narrow on the first layer in order that the second layer has something very clear to play with. 

The second track required a larger sound and had about six layers on each part. Some sections were written with pizzicato and here the challenge when overlayering is to remain scrupulously accurate to the click. This can be achieved by internally subdividing (this is where the musician counts in his or her head in a smaller note value than the beat, so instead of counting four crotchet beats in every bar, the player counts say, eight quaver beats). Mental focus is important to avoid wasting any valuable studio time. The track also contained very high notes on the first violin part: some which were two octaves above the top ‘E’ string. Stratospheric stuff indeed, but this is a region of the violin that can be highly effective when writing for a string section (particularly in film music). 

The third track was the pop track and here the strings added a fullness and warmth which hopefully will complement the track well. As a guide we had a piano and vocal in our headphones and as the strings built up we asked Ollie the engineer to change the balance in the headphones so that we increasingly listened to more strings and less of the track (this is good practice as once the foundations of the string parts have been laid, it’s good to be able to blend with the string sound as session musicians record each subsequent layer). 

It was great recording at The Premises as many of the recording rooms are the perfect size for studio strings: large enough for up to eight players to record simultaneously but not quite large enough for a whole string orchestra.

Recording the upper strings at the Premises Recording Studios

Recording the upper strings at the Premises Recording Studios

Recording violin and cello parts at the Premises Recording Studio

Recording violin and cello parts at the Premises Recording Studio


Recording a String Quartet at a studio in Herefordshire

Monday, August 5th, 2013

A couple of Mondays ago, myself and Julia (cello) recorded session strings for a talented independent singer/songwriter called Dave Roberts who had written all of the strings for four of his tracks and initially hired me to notate his completed midi files. From the beginning, Dave was a pleasure to work with as his instructions were very clear and he had a real attention to detail which resulted in well written string parts that complemented his songs really well.

 The recording sessions took place in a rural area of Herefordshire in a first class recording room. This must rank as the most idyllic location we have recorded in thus far, and the spacious, wood paneled rooms had thick curtains on adjacent walls which created an cleverly controlled sound. Both Dave and studio engineer Adam presided over the session, which involved a solo string quartet (with myself playing the viola, second violin and first violin parts).

Recording violin and viola parts

Recording violin and viola parts

 As it’s always good practice to start with the lowest instrument (so that each subsequent layer can build on the accurate tuning of the bottom part upwards), Julia recorded the cello part first for all four tracks. In situations like this, it’s also good to record each instrument separately (and not simultaneously) so that the producer can then have the ability to mix and process each of the four instruments separately. Her rich sound filled the room and as always proved a hard act to follow. I then added a viola layer over her cello tracks, making sure the tuning was kept spot on. This is potentially a tricky way of doing it, as in a ‘live’ string quartet, players will be continually adjusting their intonation with both players above and below them. With overdubbing, if any notes are not exactly in tune, it could make it hard for the subsequent layers to blend well.

 After the two violin parts were added we then returned to one of the tracks which had tremolandos all of the way through (a rapid repeating of a single note which creates an atmosphere of tension in a piece of music). As these seemed better suited to a larger string section, we took the decision to over-layer four violas and eight violin parts all in all.

 Recording sessions often involve lots of playback and sonic adjustments, and therefore studio time can slip away easily. The whole session on this occasion was completed half an hour early and everyone was happy with a job well done. It just remained for myself and Julia to travel the 100 miles or so back home again!

Recording strings on a budget – when the cost of hiring a string orchestra is too high

Thursday, March 21st, 2013

A couple of weeks ago I was approached to add some strings to a track by a talented singer/songwriter who I had recorded for last year. He wanted to record 6 passes of first violins and 6 passes of second violin parts for two songs and the session took place in the Gravity Shack Studio in Tooting. As both tracks were already fairly full in the lower and mid ranges, he did not require viola or cello parts and had arranged the score solely for violins so that the higher register could be filled out.

As this was an independent project, the client was on a budget and certainly could not have justified paying over a thousand pounds for a small string orchestra to come in, let alone hiring a studio large enough to seat them all. For this reason (and as an alternative to him using synthesized strings which he was unhappy with) I overdubbed all of these violin layers myself and we were all done in a single three hour session.

So, in a situation where songwriters and producers just can’t live with the slightly unrealistic quality of VST – because they just don’t sound as good, they might imagine that the cost of hiring string players to provide the real thing could be prohibitive. But this particular client ended up with a very high standard of playing overall (in terms of tone, phrasing and tuning) and retained the richness and feel of a full violin section for only £200.

The overdubbing of orchestral instruments has been widely used, especially when recording strings for pop songs since the 1960s and experience has taught us that it only really works with very good players who are 100% accurate with tuning and take a meticulous approach to each take. Even a hint of a mistake or inconsistency could become magnified as further layers are recorded and on several occasions we have been called in to re-record strings for clients who had initially tried to keep costs down by trying this approach with amateur or less experienced string players. Session musicians who specialise in accurate overdubbing will also develop other studio routines such as staggering the bowing slightly and subtly varying their articulation between takes, to sound more like a string section made up of many players, each with their own style.

Recording at the Chiller in Surrey

Tuesday, November 8th, 2011

Last week, four of us travelled down to a recording studio called ‘The Chiller’ in Thursley, Surrey to record 10 pieces for an independent composer. Each piece was scored for string quartet and generally lasted between 30 seconds and two minutes. As none of the pieces were technically difficult, we were able to record quite a lot of music in four hours. The composer attended the session and as we weren’t playing to a click track, he conducted us in at the beginning of each piece so that the tempo was exactly what he had in mind.

Although beforehand, the idea of overdubbing an extra layer of strings was discussed, the client decided that the sound was full enough with just a string quartet. One danger of recording two layers of a string quartet can be that by having two string players in unison on any one part, the sound can end up thin with vibrato clashing. With an overdub of four players being neither a string quartet (with the beautiful sonority of individual instruments) nor a larger string section, this often doesn’t sound satisfactory. In the event, the sonority of a quartet gave the sound that the client was looking for and it suited the character of his music.

We’ve observed several times that recording one player per part (as in a quartet) can have a stronger sound with more impact  because the individuality of the musician comes to the fore and the sound is more soloistic.

Recording Strings for a Re-branding Project

Monday, March 21st, 2011

On Friday, five of us entered the studio to record real strings for the re-branding campaign of a large company. The music had been carefully scored for double basses, cellos, violas and violins and was already well notated, meaning that we could simply turn up and play without us doing any additional work on the parts or wasting any studio time. As such, the three hours of recording studio time were spent in an intensive fashion, with three tracks (ranging from 15 seconds to two minutes) being completed with several overlayers to thicken up the string sound. Although the composer and producer had envisaged our strings being added on top of some high quality sampled strings, it was a distinct possibility that if we managed to create a full enough sound using overdubbing in the studio, the live strings would suffice on their own. As live strings are always far more convincing and natural sounding than even the most expensive string samples, this was an option we were all aiming for.

The session went well and although there were a couple of tricky passages which were really challenging, by the time we were finished, the sound was really rich and full and the clients seemed to be very happy with the result. We look forward to working with them again on other projects in the future.

Sustained notes….

Sunday, January 30th, 2011

Over the last couple of weeks we have been recording strings for an album of healing music which has also featured oboe, cor anglais, flute, alto flute and harp – it has been quite a new experience.

On looking at the score, many of the notes are held for several bars and the pieces are very slow in tempo so it would sound like an easy task to record, compared with something faster or more lively. In practice, playing ultra-slowly is a task that requires patience and concentration as it is very easy to lose one’s place. As we were layering in order to give the impression of a larger body of players, it was also important to know where we had changed bow stroke so that on the following take, the bowing changes could be staggered (this means that when the bow changes direction, on each take it is done in a different place to give the impression of a seamless sound).

Another aspect is being aware of the overall sound needed when all of the strings have been recorded. A good example of this is where crescendos or diminuendos are written in all of the parts. If there are 8 overdubs for example, then the changes in dynamic can sound much bigger than would have been anticipated after each of the individual takes. The same is true for individual dynamics, whether soft or loud.

In any situation where overdubbing strings occurs, the overall tuning is another vital issue to get right from the very first take as even slight variations can become magnified as further layers are added, particularly if there is a spread of sound over four or five octaves.

In music of a very slow nature, the challenge is to keep a beautiful string sound that doesn’t break up, have any lumpiness to it or sound as if the bow is changing. This can only be achieved by a singing cantabile approach where one is thinking in phrases with a strong rhythmical pulse, despite the sustained nature of the music.

Recording Violin and Viola parts

Tuesday, November 2nd, 2010

On Friday, we had the pleasure of recording some violin and viola parts for a composer who had already recorded the cello parts himself. As he hadn’t heard his composition performed on real instruments before, it was a pleasure to hear the music unfolding – one layer at a time.

One of the skills require in overlayering parts to an existing track is keeping a very accurate rhythm. It would only require one part to be rhythmically unstable for the whole track to sound uneven, so it is vital to make sure that every take is absolutely perfect. In a real ensemble situation, players have lots of subtle non-verbal cues between them and when these are missing in a studio situation, it is a matter of remembering exactly where the phrasing and speed alter in the minutest detail.

In the case of this composition, there were accelerandos which occurred over several bars as well as many changes of tempo. It is possible to record by stopping at each tempo change and editing but as we only had an allotted time of two hours, working swiftly was of the essence and therefore we did each individual layer in a single take. This meant that the changes of tempo had to be smooth and precise.

Due to extraneous noise around the recording studio caused by some building work in the vicinity (which nobody could have predicted!) the whole project did run over time a little, but the composer was absolutely delighted with the end result of the real strings – which hopefully brought his composition to life.

Symphony Orchestra Size Sound

Saturday, February 6th, 2010

Occasionally we are approached by clients who are self financing a project or who are on a very restricted budget, yet are unhappy using samples and still would like to add real strings to their track. Often, it is a private project with a tight budget, which (without going overseas and potentially risking the quality of player) would make hiring a full string section and large enough recording studio out of the question. With modern technology, several options are open – either using a high quality sample string sound as a base and layering a few real players on top (to give a more realistic and convincing effect than samples alone), or hiring a handful of very good players who are able to overdub themselves accurately – subtly varying the bowing, level of vibrato and microphone positioning on each take to try and sound like different people.

Although we always recommend that when the budget is not an issue, as many players are used as possible to capture a natural sound – surprisingly overdubbing can work well as an alternative to synthesized strings – although requires great concentration and an attentive engineer to sound convincing. This week, we were asked to overdub to create the sound of  64 string players – which is more the scale of string section found in a large symphony orchestra. Because each section was divided into two separate harmonies (or lines of music), we had all in all 8 layers for each line of music. The parts were recorded to be one element of a track which had plenty of other instruments and effects going on, so the overall sound was more convincing than adding synthesized strings, yet didn’t break the bank.

Layering strings

Tuesday, November 17th, 2009

Some of the work that String section does involves providing strings for composers or bands financing their own projects on a limited budget. People believe in a track enough to invest in live strings (rather than samples), but simply cannot afford to hire a full sized string orchestra without going overseas and potentially compromising on quality.  Occasionally, in order to keep things within budget, or because only a few of us can fit into a smaller studio, we have overlayered the same players in order to create the impression of more musicians.

Although this might sound straightforward and the obvious answer to keeping costs down, it is actually a skillful process that only very good players are able to pull off convincingly. It is vital that the first string parts laid down are rhythmically perfect with 100% accurate tuning, as any slight discrepancies of timing or tuning can become exaggerated with a subsequent overdub. It’s often the case that this first layer takes the longest time to record and get right. If a handful of players are to successfully layer their sound a second or even third time, it can potentially end up sounding weird and slightly artificial – rather like a choir made up of the same few voices, without the variety of sounds produced by a large group. One way to overcome this is for the session musicians to add a little variety to each take, without disturbing the rhythm or intonation. This can be done by varying the speed of vibrato, changing the weight of sound and even playing a passage on different strings (so that some notes on the lower region of the violins ‘E’ string for instance could be played higher up on the ‘A’ string). A good engineer will help enormously and can subtly change the position of the microphones between takes to avoid ‘phasing’. ‘Phasing’ is where the identical frequencies are replicated or fractionally overlap, causing the sound to become sort of ‘fizzy’. When overdubbing, if particular care isn’t taken to avoid this, what started out as a high quality group of session string players can end up sounding  more artificial than samples – which completely defeats the object of hiring live session musicians in the first place.

When overdubbing is done badly, it’s easy to tell straight away what’s gone on, especially in exposed passages. However if over-layering string parts is done with care and attention, the sound can be quite flawless- but success really depends on having excellent players, a sensitive producer and some very precise ears for detail.