Kenneth Wilkinson

saxophonist | composer | artistic researcher

Home » Writing & Music » Audio Levels in Classical Saxophone Recording

Audio Levels in Classical Saxophone Recording


Over the last decade I’ve become more involved in recording saxophones or the production of electroacoustic music. I thought it might be useful to list a few things that I had to understand in the beginning. Absolutely basic stuff for a properly trained audio or recording engineer, but for instrumentalists perhaps not.

Recording Saxophone (particularly Classical)

You can easily find recommendations for the best microphones to record acoustic instruments. Microphones can colour the recorded sound and be weighted towards particular frequency ranges (similar to speakers). I chose two diaphragm microphones with a particular recording pattern that produced a neutral sound with little added warmth, and also weighted to pick-up high frequencies; they’re incredibly sensitive. I’m comfortable that they seem to afford a pretty good representation of acoustic sound across the whole family of saxophones. 

I use Focusrite audio interfaces with Cubase Pro. I find that recording to a peak level of around -12 dBFS works for me, you’ll have to experiment. The timbre of the recorded sound changes with different microphone positions. I’ve experienced the ‘microphone in your bell’ position in various ensembles, which produces a completely different (and unnatural?) sound to a more distant position, especially for classical saxophone.

Levels within your DAW

  • I use 32 or 24 bits rather than 16 bits to give me a larger dynamic range and recording at lower levels.
  • I try to record with a peak level -18 to -12 dBFS to allow for later mixing and other processing. 
  • Gain-staging is the way to go I think, establishing good mono levels before ‘messing around’.
  • Don’t attempt to end up with a mix that peaks at 0 dBFS, that was the old days of ‘loudness wars’.
  • I usually set the target levels on meters to match the required streaming/broadcast levels.
  • I avoid normalising when creating audio files.

Creating an Audio File of Your Mix

National and International organisations have agreed a set of protocols relating to audio delivered over different platforms such as YouTube, Spotify, iTunes etc. If you don’t comply to the rules, they have algorithms in place to alter you audio to ensure you do.

Rather than focusing just on peak normalisation, it’s good to also look at Loudness Normalisation.

  • LUFS Target Level. Different systems (streaming, radio etc.) has different requirements regarding loudness. The European level is -23LUFS, America -24LUFS and streaming services around -16LUFS or -14dBFS.
  • If an audio file exceeds the target level, the audio is raised or lowered until it meets the rules. This means that if you have produced a highly compressed audio file at a high LUFS, it will be lowered to meet the rules, but sound worse because of lack of transients and dynamic range. 
  • Keeping the peak of your mix below -2.0 dB on a true peak meter will avoid clipping and distortion during any transcoding processes, or adjustment by a streaming service.
  • Audio with a larger dynamic range will usually sound better than more compressed audio.

Target Levels for Streaming Services

Apple Music-16 LUFS-1.0 dBTP
Spotify-14 LUFS-1.0 dBTP
Amazon-14 LUFS-2.0 dBTP
YouTube-14 LUFS-1.0 dBTP
CD-9 LUFS-0.1 dBTP
EU Broadcast-23 LUFS
US Broadcast-24 LUFS

Useful Terms used in DAWs like Logic Pro or Cubase


1 LU equals 1dB  


Integrated Loudness, the average loudness. Use to check if your level adheres to the streaming or broadcasting rules of the service you’re using.  

True Peak  

True Peak Meters detect inter-sample peaks to give a more accurate reading of peak levels.  

Peak to Loudness Ratio  

This is the difference between the average loudness (LUFS) and the True Peak level.  

LU or LRA  

Loudness Range. Shows the dynamics of the audio material. This is not measured from lowest to highest point, rather an average dynamic range. I prefer as wide a range as possible for recording classical music.

I set my target levels in meter plugins to get a sense of audio levels as I work. I aim to get a good dynamic range for acoustic recordings, avoiding compression as much as possible.

Some great sources about recording:

Bartlett, B. and Bartlett, J. (2009) Practical recording techniques: the step-by-step approach to professional audio recording. 5th ed. Burlington, MA: Focal Press.

King, R. (2017) Recording orchestra and other classical music ensembles.

LinkedIn Learning: There are so many great courses giving an insight into basic recording and editing, to mixing and on to mastering. Sometimes the courses are free if you are involved in a university.

Kenneth is a saxophonist, composer and artistic researcher based in Europe. His current focus is an exploration of embodiment, action-perception, and cognitive processes when using cane (organic) saxophone reeds.

Latest Posts

Clockwork – a simple piece for one or more saxophones

Clockwork – a simple piece for one or more saxophones

'Somewhere, distant, the ancient mechanism of an old clock can only just be heard'. This piece portrays a clockwork mechanism, relying on the performer/s creating an almost inaudible texture of soft sounds. It is most effective performed by an ensemble, as a greater...

Nine Pieces for Five Players

Nine Pieces for Five Players

New recording by Quirk saxophone quartet. Nine Pieces for Five Players by renowned composer and saxophonist Richard Ingham was originally commissioned and written for the Plume Quintet. The original wind quintet version was later rewritten as a saxophone quintet for...

Fragile Dyads – solo alto saxophone

Fragile Dyads – solo alto saxophone

Fragile Dyads: small fragile units of sound building up into a larger moving texture. Narrow bands of controlled instability, exhibiting fragility and flaws giving a natural life span. Reflecting the fragility and complexity of life, the dyads (two sounds played...