School Project

[carousel182]

This year in school i have to do a project using the scientific method. The topic of the project is, Which Electric Guitar Bridge Gives the Most Sustain. I'm basically using three bridges for my experiment a hardtail stringthrough, a hardtail non string through, and a tremolo, and putting them on this body/neck. What im wondering though, is how i can research the experiment; does anyone know of any website or book that i can read to get information on the topic. Does anyone know why a certain bridge gives more sustain? Any information would be helpful on this subject.

Thanks,

Carousel

[RGman]

You could look into the different materials used to build the bridges and their affects on sustain?

[Mickguard]

I imagine you'll want to do the trem measurements last of all, since you'll have to route the guitar for it.

But if the guitar's already routed for the trem, then there's no way you can get true measurements for the other two bridges (at least in the way they're supposed to be used, that is, on top of a solid body, not a trem routed guitar).

Although, once you've routed for the trem, you won't be working with exactly the same guitar anymore (although that's not a big deal, since this type of bridge requires that much of a route).

Not trying to discourage you-- I think it's an interesting project. It'd be nice to get some real measurements for this type of thing, so I just wanted to point out some of the variables.

[carousel182]

Yea, I'm doing the trem last, right now the hardtail non string through is on.

[dh7892]

I think it's great that someone is doing this. It will be good to have some actual data to refer to rather than just people's anecdotal "evidence".

How are you going to pluck the string? Obviously you'll need to try to make each pluck the same. Are you going to do any type of frequency analysis on the notes too? How about the sustain of different notes since different frequencies will transmit to the bridge/body differently.

Make sure you post your results here as I for one will be very inrerested

[carousel182]

I'm gonna make a string plucker that drops down so that its nearly the same each time. If i do about 10-20 trials each bridge it should be pretty accurate.

I'm gonna plug my guitar into my computer and record all of the notes and see how long they last before they die.

Does anyone have any ideas on how i could do this experiment more accurately.

[dh7892]

So the strings will be struck a bit like a piano? Like you say, you'll need to do a few repeats of each pluck.

How will the plucker fall? Will it be moving towards the pickups or across the face of the guitar?. Perhaps it would be easier to pluck by gripping the string and pulling it out a fixed distance and releasing it. This way you only have to worry about the release mechanism to make sure it's a clean release. I think this is more repeatable than trying to hit the string in the same way each time with a falling object.

[aidlook]

I got some ideas...But I'm at school right now. I can make some illustrations of my idea later.

[carousel182]

thanks everyone anyone know any sites that i can research this at?

[thegarehanman]

Just make a pendulum to pluck it. Make the pendulum long enough so as to nearly simulate the action of a hand plucking a string. 4' radius should do the trick. Just have it set up so you're releasing it from the same height every time. For something like this, it would be wise to use a soft pick to avoid popping a string. You could even incorporate a mechanism that catches the pendulum so it doesn't go back and pluck the string again. This would free your hands up to pay attention to the computer. Make sure that you figure out what frequencies and amplitudes are audible and what aren't. Because while one bridge may have the longest physical sustain, it may not actually have the longest audible sustain. I know I've heard of a study that was done on les pauls vs strats (although the person could very well have been bs'ing) where a les paul was said to have longer audible sustain, while the strat had longer physical sustain.

I think a computer is going to be the most accurate recording instrument you're going to get without spending quite a bit of money. I could be wrong though, I'm no recording guru.

peace,

russ

[carousel182]

hey thanks man, do you know how i could determine what frequencies are audible and which arent?

[carousel182]

Does anyone know reasons why some bridges would sustain more than other?

[Mickguard]

Does anyone know reasons why some bridges would sustain more than other?

Different types of metal, different grades of metal, different amount of metal.

Different attachment method, different amount of torque used on attachment screws, different saddle materials, different saddle position/notch.

Etc. Etc.

Big project!

[thegarehanman]

I'm sure some diligent searching on the internet, compliments of google or yahoo, should give you a range of audible sound waves. One thing you'll have to determine is what the sensitivity of your microphone, recording program and sound card are. Remember that just because your computer doesn't tell you a certain frequency is present, doesn't mean it really isn't. Then again, anything your computer isn't sensitive enough to hear, your ears probably couldn't hear anyways. Anyhow, these are some things to look into and consider. Knowing the limits of your testing methods is just as important as knowing what you're testing for in the first place.

peace,

russ

[Muzz]

Fun research, how were you going to measure the sustain?

I guess you could record the guitar in a program like Audacity that displays the sound signal versus time and measure how long the signal takes to reach 10% of initial volume.

[Mickguard]

I use an Intellitouch tuner for my acoustic instruments --seems like a similar device would be helpful.

Sustain is a weird thing though....I remember reading someone else's experiment (can't find the link anymore) where he measured the difference in sustain between a Stratocaster and a Les Paul. The Strat had measurably longer sustain in terms of vibrating strings. But the Les Paul's sustain was more audible for longer.

[carousel182]

To measure the sustain i was going to put it into a program like auacity and check how long it took until the sound stopped.

does anyone know of a better program to use that will measure the sound better

[carousel182]

thats how the experiment is set up as of now. the guitar is wired one volume and an output which goes to my digitech rp250 which goes to my computer

the pendulum thing is on top of the guitar, works good.

[Scrappy_Squirrel]

I'm sure this will come across as harsh but it probably needs to be said so I'll take a flame or two...

MickGuard is correct in that there are many, many variables and many types of bridges to test. Unfortunately your results would be meaningless because you are changing more than one variable with each model in your experiment. For a true experiment you would want multiple copies of the same bridge: One bridge would serve as your control and the copies would each have one variance from the control. For instance: You might have four tom bridges of the same brand, size, shape, style and method of attachment into the same guitar but each bridge would be made from a different metal. You can only change one variable at a time with a true experiment otherwise your confounds are too numerous to gather any real information. In fact, in order to establish a true control, prior to introducing variables, you would need to measure the sustain of six, or more, of the exact same bridges, in the same guitar, even before introducing any variables into your bridge experiment. Real science is a pita but going through the motions is what yields real results. I'm not trying to burst your bubble but I just wanted it to be stated for the record in case it makes a difference in your grade. I'd hate to see you get a bad grade.

[thegarehanman]

This is a highschool project, not the basis for a thesis, and I think, while the results may not be entirely useful or all-inclusive, how he presents those results could very well still get him a fine grade. Just say, x bridge is this metal, attaches to the body this way, and the strings go over/mount to it this way. y bridge differs from x bridge this way. And so on. If he doesn't say "the difference in mounting techniques and not the difference in metal is the definite reason for a tonal difference," then I don't see why he couldn't get a good grade.

Just be as consistent with testing methods, and as objective with results as possible.

Me thinks the teacher wants to see the quality of the research, not the quality or usefullness of the results. If the former is the case, than this should be a good project. You are the only one that will know for sure though.

peace,

russ

[Muzz]

I see the point that the results are not likely to get published in New Scientist, but these dudes get away without independent repeats or analysis of variance all the time. Should be cool for school. I am interested to see how it goes.

[carousel182]

This is a highschool project, not the basis for a thesis, and I think, while the results may not be entirely useful or all-inclusive, how he presents those results could very well still get him a fine grade. Just say, x bridge is this metal, attaches to the body this way, and the strings go over/mount to it this way. y bridge differs from x bridge this way. And so on. If he doesn't say "the difference in mounting techniques and not the difference in metal is the definite reason for a tonal difference," then I don't see why he couldn't get a good grade.

Just be as consistent with testing methods, and as objective with results as possible.

Me thinks the teacher wants to see the quality of the research, not the quality or usefullness of the results. If the former is the case, than this should be a good project. You are the only one that will know for sure though.

peace,

russ

yea... its a highschool project. I don't think how accurate the results are matter as much as how well done the experiment is. my teacher probably has no idea about anything having to do with guitars so i honestly don't think it will be thought into that much.

[dh7892]

The original post states that the aim of this work is to demonstrate use of the scientific method. Here’s what I think the scientific method is:

Decide a Hypothesis (and a contrary null hypothesis) and devise an experiment that aims to test (to a level of statistical significance) which of these two hypotheses is true. If your hypothesis is something like: “Bridge A provides longer sustain than bridge B” then your experiment is fine.

If you were trying to provide a comprehensive investigation into all possible types of bridge/material etc. then you’ll need to do more tests. But you know this. You also need to make sure that you account for any (significant) factors outside your control. You also know this which is why you’re repeating your pluck because you know that there’s likely to be a relatively large random error in these. You’re also trying it with different strings because you know that “sustain”, as you are interested in it, might be different across a range of frequencies.

Changing more than one thing at a time most certainly does NOT mean that your results are meaningless. It might mean that you can’t say exactly what causes the sustain (is it bridge design or material?) but that’s not what your trying to investigate so that’s fine. By the way, in big experiments (like pharmaceutical/farming trials) they deliberately vary several factors at each trial and use statistical methods to back out the importance of each variable. It’s called Design of Experiment technique and you really wouldn’t be able to make much progress without it.

Sorry for the scientific high-ground taken here but I’m a research scientist and I think the scientific method gets abused all the time and it’s a shame because it’s really the only way to provide insight into anything. Science is not a PITA, it’s about deciding what you want to know and devising a way of satisfying yourself about the result. Yes you have to be clear and precise in terms of your aims and methods but that doesn’t mean it has to be onerous. I think you’re doing a good job here and I’ll be waiting with interest to see your results.

[WezV]

Sorry for the scientific high-ground taken here but I’m a research scientist and I think the scientific method gets abused all the time and it’s a shame because it’s really the only way to provide insight into anything. .

maybe if he manages to do what the course requires then one day he can be a research scientist too.

what we (people who teach science based methods at highschool age) tend to do do is give them an introduction to science, if they decide to carry on with it in life they will learn the points you are making but it is very hard to grasp till you have tried science the high school way. That means changing one variable, measuring another one and controlling the rest - It doesnt have to be perfect science at this level, it has just has to teach them the basics

I have just been doing the a'level (17-18 year olds) coursework with my students and they struggle with two main variables and they really struggle with statistical tests and analysing data.

i am not sure any of them would be ready for a test involving changes in multiple variables and i know they would not be able to do the statistics - even with a powerfull statistical processor to help them

[thegarehanman]

Glad we're seeing eye to eye here, dh7892. As a high school project, he's doing exactly what he should be doing. Hell, he's doing it all on the same guitar, which is more of a constant than most "luthiers" use when think they try to compare guitar parts' effect on tone.

I do a lot of design and fabrication with composites for my job and for school. I'm with you 100% on the importance of a methodical application of the scientific method. However, I fear you're going to scare the kid away from science because he'll be afraid of turning into a scientific snob! haha

peace,

russ