29 July 2013

Recording a Violin: Microphone placement

My name is Claire Curtis. I'm a violinmaker who lives and works in southern Maine, about an hour north of Boston. Understanding how the instrument works, and how both the player and the listener perceive the sound of the instrument, is essential is my work. Knowing how to record the instrument, in both concert and studio settings, is part of that understanding. 

This is Assignment 1 for Week 1 of the course "Introduction to Music Production", offered by Loudon Stearns of the Berklee College of Music, via the online educational venue, Coursera
I offer a tutorial on the suggested topic, How To: Recording an Acoustic Instrument.
I am specifically focusing on microphone placement to record a violin. 
How to Record a Violin: Microphone placement

A violin is itself a complex sound-generating structure. 
To make a sound, a bow is drawn across the strings. This creates a transverse oscillation in the string, producing a fundamental frequency with harmonics. The string vibrations cause the bridge to dance, which transmits those vibrations to the body of the violin. The violin body has its own complex set of vibrational modes, which selectively enhance and dampen specific harmonics to produce the timbre typical of a violin.  In the end, the violin converts all those vibrations to the patterns of air compression and rarification that we call sound.  

Why not just say the violin makes a sound? 

The problem is that the violin does not radiate sound evenly in all directions, so recording becomes problematical.

The sound that comes out of the f-holes (the sound holes) will be heavily biased towards the body cavity ("Helmholtz") resonance. Miking from over the player's shoulder will be different from miking from the front. Miking at or near the bridge will circumvent some of the body resonances, which also means it will record less of the typical violin timbre. 

This means that very close miking—a pickup, a bug, or a very close mic—will not sound like a natural violin. The engineer will have to provide "warmth" (corrective equalization), and probably some reverb. This is quite do-able, but not optimum. 

Furthermore, there is a distance effect. The different mode vibrations are separate simultaneous sounds close to the instrument. Further away, the harmonics, especially the upper harmonics, start to blend. A more distant mic will do justice to the sound of a good violin. Distance will also reduce bow noise. 

Another problem that must be considered is that most violinists sway or turn as they play. Since the violin sound itself is so directional, this swaying means the mic will pick up tonal variations, especially if the mic is fairly close. 

One solution is to place the mic at least 6 to 8 feet away. This means that the mic needs to be quite sensitive, especially at high frequencies (which determine that violin "timbre"). And it can't be too directional, or the player's sway will have too great an effect. A good quality condenser mic would work here. 

In a live concert setting, a good compromise would be to have both a distant mic and a pickup. The distant mic will provide the necessary ambiance and timbre, while the pickup will allow the player to move as much as desired. 

The tracks can then be tweaked as necessary. 

I joined the class late, and did not have time to make a video or even an illustrated lesson. I hope what I do present is reasonably clear, even if it doesn't have diagrams or a video. Did I need to better explain how the pattern of upper harmonics constitute timbre, the "sound" of a violin? 
Maybe I will have the opportunity later in the course to redo this. I just felt that, even though this assignment is optional, that I ought to do something, and this really is a topic near and dear to me. 

In any case, thank you for taking the time to read this.