Sign up Calendar Latest Topics Chat

  Author   Comment  

Posts: 100
Reply with quote  #1 
My apologies to readers of this long-winded post but unfortunately this subject deserves this. Also keep in mind that what I mention here is not entirely hypothetical. I have designed a sarod which has been built by Dave Nugent a luthier, a pattern maker and a CNC milling machinist, also a sarod player and student of Pt. Buddhadev DasGupta. I have also made extensive modifications to my sarod and the conclusions here are a result of my actual experience and also discussions with a lot of folks including Dave, Pt. DasGupta and a host of other luthiers. My own background is in structural dynamics and after all musical instruments are vibrating structures. This does not mean that I may not have overlooked something or am not open to suggestions. I just feel that these things must be thought through before attempting work on ones instrument.

The subject of action has come up several times over several posts. It seems that high sarod action is the bane of all sarods in general. It makes playing more difficult, reduces sustain, makes the position of the notes different than what you would expect. e.g. Pressed Pa on Ma string is not directly underneath Pressed Re on Sa string (while this is not a major difference, this difference gets more exaggerated the farther you move away from the nut). Also, ekhara taans are a lot harder to play when the action is high. So I thought I would consolidate all of my thoughts into one post.

There are several things to consider that have interplay related to action and attempting to correct it on the sarod. These are as follows (with detailed descriptions below):

1. Buzzing (High action = less buzzing, low action = more buzzing…the primary issue for which action is usually made high)
2. Sarod volume (High action = High volume, Low action = low volume but sustain is also affected)
3. Sustain (Lower action = more sustain and vice versa with caveats)
4. Sarod plate relief (Detailed discussion below along with other parameters)
5. Sarod plate curvature (Greater curvature = less buzzing, lesser curvature = more buzzing but only to some extent)
6. Sarod plate irregularities (Irregularities are bad…all around…they cause buzzing and also dead spots and loss of sound during meends and thus reduced sustain)

There are two distinct types of buzzing.

The first is caused when action is low and all other parameters are correct. This buzzing occurs immediately adjacent to your nail which is pressed on the fingerboard….simply because the action is not high enough. In this scenario there will be buzzing, regardless of where you press the string, starting from some point near the nut and all the way up to some such point on the plate as you proceed towards the bridge where the angle of the string taking off from the plate increases sufficiently to prevent buzzing. You will also note that this buzzing is the worst at the nut and improves as you proceed to the “good” point.

The second type of buzzing happens only at certain spots on the string and not all spots and is caused by plate irregularities.

So let us study the actual mechanism of buzzing. The string in its most primary mode of vibration can be imagined to be like a skipping rope. Imagine it to be vibrating not up and down but side to side with a small up and down component. Looking at it by sighting down the plate, the cross section of the vibrating string would trace an elliptical motion with the longer axis parallel to the plate and the shorter axis perpendicular to the plate. This shorter component is what causes the string to smack against the plate and buzz. Also consider that the total size of this ellipse is large at the center of the vibrating string length and smaller at the ends, down to zero. Obviously as you proceed from the point of contact of the fingernail at the plate towards the bridge, this ellipse is lifting away from the plate. Buzzing will then be caused if the ellipse is large enough at any point to make contact with the plate. The harder the string is plucked, the larger the ellipse, therefore a softly plucked string may not buzz but a strongly plucked one might for any given action.

So in the first buzzing scenario, the part of the string immediately adjacent to the nail is vibrating with enough amplitude as to touch the plate. In the second scenario, the action is high enough and the first kind of buzzing does not happen but due to an irregularity in the plate, there is a bump and as you approach this bump, the string makes contact with the top of the bump and buzzes. In a situation where the string is plucked, the buzzing will be immediate whereas in the case of a meend, the buzzing will be absent but as you approach the point, if the string has sufficiently lost its motion, you will not hear buzzing but the sound will suddenly get killed and the meend will not continue to be heard.

Given the above discussion, one can deduce that on a sitar, or any fretted instrument, the action can be lower because by the time the string approaches the next possible buzzing point (next higher note fret), it is high enough to not buzz. You can get away with fairly low action. But on non-fretted instruments such as the violin, sarangi, sarod, etc., you do not have this benefit, the next possible buzzing point is immediately adjacent to the point where you press with your nail. The last 3 parameters I mention above will affect this and this is discussed later on in the post.

Sarod Volume
If you read any luthier manuals you will see that they always stress that the angle of the string from the bridge to the attachment point (the short non-plucked part of the string) is important for volume. The reason for this is as follows: Imagine that a string is suspended between two points, the nut at one end and the far point of attachment (near your right forearm) at the other end with the bridge being non-existent. Then imagine if the bridge is slowly raised from below until it just contacts the string but does not cause it to bend. In this scenario, the string is exerting very little weight back onto the bridge, they are just barely in contact. If you now pluck the string, the string will vibrate but very little of this vibration will be transferred to the bridge. Since very little is transferred to the bridge, the skin and thus the air chamber in the sarod (or any instrument) will not be set into motion as much…ergo less volume. Now as the bridge is raised (ie raising the action), more and more effort is needed because the string exerts a lot of weight on the bridge. If you pluck the string now, a whole bunch of the motion will be transferred to the bridge, ergo more volume. Also, since the entire system is inter-related, if the skin causes damping (not wet damp…I mean damping as is meant by damping of a body in harmonic motion) the sound quickly dies out thus lesser sustain. If the bridge is not too heavy and the skin is thin and nice and springy and the action is right, you will not only get good volume but also good sustain for meends. The shape of the lip of the bowl has considerable impact on the damping properties of the skin but that is a topic for another discussion.

The sustain is affected by the action as seen from the previous discussion but also for another reason. The nail itself causes the string to get dampened. If the action is high, more force is required to hold the string down and thus there is a greater dampening effect, thus causing reduced meends. Try this, see the amount of sliding meend you can accomplish closer to the bridge as opposed to closer to the nut. Since the “take-off” angle of the string adjacent to the point of nail contact for a given fixed bridge height is greater near the bridge than near the nut, more of the damping effect is exerted by the nail near the bridge (corrected...I had said "nut" in my post but I have edited it now...sorry for the error and any resulting confusion) since it absorbs more of the energy of the vibrating string.

Sarod Plate Relief
Relief as applied to string instruments refers to the deliberate bow introduced into the neck to allow for lower action. While a lot of people consider this a black art, it is actually not. In the latter part of this post, you will see exactly how relief effects action and why relief is necessary only towards the nut and not towards the bridge. Relief has to be accurate, too much relief is as bad as too little relief. Also, guitars can have their relief adjusted with the truss rod while sarods have no recourse for this. My ideas about how this could be done will be presented in a later post.

Sarod plate curvature
This is the curvature of the plate perpendicular to the strings. Normally, this is referred to as radius in guitar luthier references. Guitars generally have a constant radius whereas the sarod has a radius that changes, being sharper and more curved near the nut than towards the bridge. Going back to the discussion of the path of the vibrating string being an ellipse, one should visualize the following. If the sarod plate is flat, it would be parallel to the long axis of the ellipse and below this axis. The possibility of buzzing is represented by the part of the ellipse below the sarod plate. If this flat part were curved downward, a significant portion of the ellipse would come out above it. As the curvature increased, more and more of the ellipse would be above rather than below but the maximum gain is in the beginning as you start curving. There is a law of diminishing returns. Thus the plate is not curved too much.

Sarod plate irregularities
Already discussed in the buzzing section…Of course, all this stuff is inter-related so it is hard to discuss one thing without discussing the other. One way to tell if you have irregularities is to find a reference edge. The reference edge may not necessarily be dead straight but straighter the better...e.g. the edge of a wall, etc. Look at this edge reflected off your fingerplate. Turn the sarod as you see this edge, causing the edge to move from one edge of the plate to the other. This reflection should be parallel to the strings and move from tarab side to M string side. If the edge undulates, you will know you have irregularities.

How to achieve the best action, volume and sustain:
This is a hypothetical scenario. You will always sacrifice one thing or the other and ultimately come up with the best compromise. To understand how to achieve this, one should go back and read the vibrating string discussion in the buzzing section above.

Let us say that the angle of the string to the plate is defined as alpha. (Consider the plate as one line and the string as another line and the two lines intersecting at the point of contact where the nail presses down). Then alpha is lowest near the nut and highest near the bridge since bridge height is constant.

Let us say that the vibrating string in its fundamental mode forms the arc of a circle, and the angle between the string in its resting position (horizontal) and the tangent to the circle at the point of contact with the nail is defined as beta.

Whenever beta is greater than alpha, there will be buzzing and whenever not, there will be no buzzing.
Beta is larger when the string is hit harder.
Beta reduces as you approach the bridge since a longer string will vibrate with greater amplitude than a shorter string. (This is straight physics, no assumptions here, the frequency of the shorter string is higher because it is stiffer…and because it is stiffer, its amplitude of motion is lower) (Note: Stiffness is a proper engineering term used to define the ability of a given shape to deflect. Stiffness is dependent on the strength of a material AND its shape. Thus a long beam is less stiff than a short beam of the same cross section. Similarly a long string under a given tension is less stiff than a short string under the same tension.)

You can see from these points above that the issue of buzzing is much lesser near the bridge than it is towards the nut. Of course you knew this inherently based on experience and feel but now there is a scientific basis for it and for thinking about it.

Now imagine that you made the action very very low, you would not buzz near the bridge but as you went backward, toward the nut, at some point, the buzzing would start. At this point in time, you have two options, the first is to raise the action and the second one…get this…is to change the angle of the sarod plate itself by angling it upwards towards the nut….thus, behind your nail where the string is not vibrating. In front of the nail, the sarod plate is flat but behind this point, towards the nut, the plate angles upward (In actuality, you want the plate to start angling up, slightly earlier, ie in front of your nail). Now, as you proceed towards the nut, you would have to keep angling the plate upward more and more, so ultimately what the plate should look like is as if it is going flat for a while and then taking off and that too getting steeper and steeper as you go towards the nut. In reality what you have done is introduce a complex curve to the plate in the direction of the strings. This curve should in reality be barely perceptible. You can see now that as the plate starts to cuve upward, the final resting point of the string at the nut is going to be higher than where it would have been if the plate was dead flat in the direction of the string. The final height so to speak should not be so much that it is more than the height of the bridge itself (as measured from a point of intersection of a straight line from the flat part of the plate and extending to the bridge). If this happens, you have guaranteed buzzing at a different location but the geometry of that is too complicated to discuss here. Suffice it to say that this gradually increasing rise should ultimately lead the final location of the plate at the nut to be no more than 1/64th of an inch or so higher (This part, I have no actual basis for...I will have to experiment and fix my plate before I can provide more actual figures). I have an excel spreadsheet and will mail it to whoever wishes to contact me for it. You can play with the various parameters and see what curves you end up with.

To have good volume and still have low action, the way to achieve the lower action is not by lowering the bridge but by raising the plate at the end that is closer to the skin. Look at cellos and violins and you will immediately see what I mean. The fingerplate is not parallel to the upper body of these instruments. The sarod as a fretless instrument should be the same way but it is not and this results in a lot of problems.

What problem would be caused by raising the plate?….aha…the tarabs would have to be raised since they come off the plate. Thus, as the plate is angled, the tarab holes would have to be raised and would end up being too close to the main strings. There is a solution for this also but it is not easy to implement.

My ultimate conclusion on the search for low action is that a hammered or rolled sarod plate is just not the way to go. It needs to be cnc milled completely or after it has been bent and rolled. This is a project I intend to pursue with the University of Cincinnati.

When I modified my sarod last, I did raise the plate and inserted a curved maple piece that I carved at the location where the plate meets the skin. The action on my sarod is now very very low (around quarter inch) but I do have buzzing. The buzzing is from plate irregularities though and only at 3 or 4 spots and I will write in my next post as to how I fixed those. (Straight edge, sandpaper, flashlight and feeler gauges will be my friends).

Keep in mind that as action gets lower, sustain increases to a certain point and then starts reducing. The optimal action for the sarod is around 3/16 of an inch. Rarely achieved but I will try and do this on mine.

Please comment so that we can get this board alive again.

Posts: 100
Reply with quote  #2 
For those of you who have already read this post, I have made some changes. I am a retard and should have added the edits in the form of replies but I directly edited my own post. Sorry...You may have to re-read.

Posts: 163
Reply with quote  #3 
WOW! Thanks for going into such detail. Are you using a Morse taper for the pegs? The photos you posted are great.


Posts: 100
Reply with quote  #4 
Do not know if it is a Morse taper or a matter of fact, I do not know what a Morse taper is. The reamer and shaper is available as a standard set. They have different sizes for Cello vs. violin, etc. I used the ones they make for cellos.

Previous Topic | Next Topic

Quick Navigation:

Easily create a Forum Website with Website Toolbox.