Friday, September 30, 2016

Carnatic Music - Audio Engineering Basics

NOTE:  The goal of this post is to give some general pointers to those managing audio in Carnatic music or Bharatanatyam programs and written for a non-mathematical, non-technical person.  All too often, the enjoyment of a program is really affected by audio problems. Hopefully, these comments will be helpful to some.  I have kept things at an elementary level and avoided jargon as much as one can.  I have to be cryptic to save on space.

1. Frequencies are measured in Hertz.  Also kH stands for kilo Hertz or 1000 Hertz, ie one thousand cycles of the signal wave per second.  Sharper, squeaky sounds are in the higher frequency range. Thus, for example, Balamularali Krishna's voice may traverse a frequency range that is lower than the one traversed by Musiri Subramanya Iyer (or a female vocalist like MS).  The Hz level changes continuously as the singer "goes up or down."    Understand the frequency ranges of voice and various Carnatic instruments.  This is important in setting the equalizer. [Male voice: 100Hz - 1800 Hz; Female voice: 150Hz-2500Hz; Violin: 350Hz - 4000 Hz; mrdangam: 100Hz-2000Hz; these are approximate].  Just to understand how these numbers relate to what you hear consider these: if you push up the higher frequencies, a male voice will start sounding like that of a female and a female voice will become very squeaky,  violin will become unbearably high pitched and squeaky, mrdangam right will start sounding metallic.  Push up the lower frequencies and you distort sound again - some examples: voice and mrdangam right (sollu) will become less clear; violin will boom in lower octaves; left side mrdangam will boom drowning most others.  Got it?  This also gives you a hint on what to change if you see these types of problems.  Play around with your home equalizer if you have one and if not with the treble (high frequency) and bass (low frequency) knobs.  Best is to try doing the experiment with only audio portions where only one musician is performing.   [Unfortunately, many systems today like Bose don't let you do anything; trust me, they provide less optimal sound than what someone who knows, even moderately, what to do with manual controls.  It is just like using your camera in Auto versus Manual. Rarely will a professional put the camera in the auto mode unless they know they won't have the time to compose their pictures!].

2. In terms of equalization thus for a male vocalist, I may set the equalizer with the frequency  around 1200 or so high and tapering down to the sides; for a violin may be centered around 2000 Hz; etc.
Make sure successive frequency knobs form a reasonably continuous curve.
Since you are not doing some fancy stuff with a separate equalizer for each instrument or musician, use the main person as your guide for this.  Things may have to be changed later depending on the auditorium, the size of the crowd (that alters sound characteristics ), etc. for which your ears are your best guide.  There is no one single set of numbers that work always.  So, treat these as starting points.
After setting the equalizer thus, adjust individual instruments or channels with the frequency knobs low, mid, high on your console which cover different frequency ranges.

3. Volume and Gain: One big cause of problems is setting the gain high.  Without getting technical, think of gain as the input volume level.  If gain is set high, a small movement towards the mic by the musician can amplify things very much.  Such abrupt moves will make them sound like they are shouting or screaming.  (Hear those awful squeaks in the auditorium? Most of these are due to high gain settings.)  A common mistake I see with most audio people in Indian programs, including so-called professionals, is that they treat the gain control as the main tool for increasing or lowering volume.  That is really asking for trouble.  Golden Rule 1: Try to make do with the minimum possible level of gain to avoid squeaks and overloads of your system.  How do you find it?  Here is a simple trick.  Set the volume control at 0 (i.e., at center), let the person play or sing and increase gain just to the level that is comfortable.  From then on change using only the volume knob!  Volume knob is what controls the output level.  Golden Rule 2: Move the knobs gently and in a continuous motion at a very slow rate; especially so with gain.

4.  If you hear the violin or the left side of the mrdangam booming, reduce the lower frequencies.  Some violinists and vocalists will ask for more "base" without realizing a little bit what it can do to how they sound.  Make them play the lower octaves and use that as a guide.   Pay attention to the right side of the mrdangam; is the sollu clear?  if not ,you don't have enough treble (higher frequencies) for that; does it sound too metallic (then you have too much treble or are using high frequency ranges too much).

5. Never work without a stage monitor.  Make sure you can increase or decrease the volume on the stage monitor without affecting the way things sound in the hall.  Don't run the stage monitor too high.  Then amplified sound enters the mic, gets amplified again, and this process repeats God knows how many times since things travel at speed of light, overloads your channel, and you get a terrible squeak; in the worst case, it blows your equipment (yes, I have seen that happen too!)

6. Unless you have undergone some serious professional training, stay away from things (especially done manually) like Effects (Reverb) which some musicians will ask for although they have no clue whatsoever what it means or how it affects sound.  [You may be able to get away with the automatic setting of Vocal-SmallHall or Vocal-LargeHall for most Carnatic programs giving good reverb but beware of hall reflection properties of sound.  If it feels like you are hearing multiple versions of the same sound, don't use it!]  Smile, nod your head, do what you have to do.  In my long years, the only musician I consider as knowing anything about sound engineering is Sangeethasagara Balamuralikrishna.  How I wish a day comes when no one is allowed to ascend the stage without some minimum training on using the mic and the essentials of audio!

Also, when it comes to recording, please set the peak level at about -6dB.  One common error I see especially from India is to set it at 0dB resulting in frequency clipping.

Some other golden rules: #3 Go  and listen from various places in the auditorium and readjust. #4 Dont respond to each and everyone who comes to you and asks you to change this or that.  #5 Have some trusted people (like those who you know can make a good judgment about sound quality) who can tell you. #6 Don't set up and walk away for good.  Unexpected exigencies do occur.  #7 Don't get too cocky; there is a heck of a lot to learn, and at any stage, what we know is a very small part of what is out there to learn.
The author Dr. V. Ramaswami is a former President and Secretary of CMANA (Carnatic Music Association of North America).  As a volunteer, he has managed the audio for CMANA and some other non-profits for many years.  As a researcher in the citadels of technology - Bell Labs, Bellcore, AT&T - he has also had the privilege of learning many technical aspects of audio and video from experts and sometimes the very inventors.  He himself holds several patents related to video distribution over the Internet and related communication technologies.
Some additional comments related to audio in  future posts.  Your comments are most welcome. Please post them here and only here if they relate to question, additional technical matters etc.  Kudos and criticisms you can place wherever you choose.  If you have not seen the page for my book, kindly visit
You may also wish to read



  1. Interesting perspective. These notes would be very useful to all Sabhas.

  2. Ram, thanks for the sound advice; no pun intended.


Your feedback is deeply appreciated. Comments will be moderated.