String seating on bridge - bridge groove is unchanged here - still cut for "normal" violin strings - this is a close-up of the G on the bridge - clearly seen that it's sitting too high ( I knew it would be ) - and ( OK, it could be because I *want* to see it  but I suspect it's visible ) - if you view the fully expanded image, yes, you can ( I think ) see the winding separation being just marginally wider as the string curves over the bridge....

As I mentioned earlier - I'm not *overly worried* - I know what I'm doing here, and the "risk" I'm taking with potential damage to the string - the strings are currently only properly tensioned when I play the instrument.... so it's only an hour or so at a time....

Oh - BTW this is the G2 ( not the D3 I was talking about, but it is similar ) - clearly it is way too high on the bridge groove.

I'm on the task, weather has just been too good recently not to spend time in the garden (which needs more attention than my octave strings )

Octave strings - ongoing investigations.....

I finally got around to doing this - I was curious to investigate the fiddle's response, in particular to the low strings (the G2 and D3) which are of course below the normally expected frequency-range the (acoustic) fiddle is designed for.

Many will be familiar with the concept of a graphical equalizer on your stereo, even smart phones these days, where you can select specific frequencies to accentuate or attenuate....

Here's a 200Hz to 4kHz equalization curve which cuts out low-ish frequencies (below 200Hz) and higher frequencies (above 4000Hz) and amplifies (lifts) those in between - 

Now - by its very construction - a violin has some very specific resonances, and has a response curve which is no way as simple as the example above - but the principle is the same - there are plate and body resonances which effectively respond in sympathy with various notes being played - a typical violin response would not be too dis-similar to this - 

As you can see - although a very complex curve -  it does exhibit a type of "band-pass" characteristic.    What it shows is that frequencies in the range of around 200Hz to around 10KHz are variously affected, some being "lifted" (in volume) more than others.   Frequencies outwith that range are highly attenuated, effectively the fiddle itself is not "responding" to them, so they will be present to be heard, but at a significantly lower level.

Fine.  Not rocket science.....  

But - yeah - although obvious now - it took me a few minutes to explain the following spectrum - taken from playing the octave string G2 on my 4/4 fiddle.

OK - before I post that - for those who don't get the physics involved - let's go back a step.   

When a string is bowed, what happens is that the hair grips on a string, then slips back, and the "grip and slip" cycle repeats as the bow is drawn across the string.  The grip-and-slip gives rise to the characteristic saw-tooth waveform produced by a bowed instrument - it's kind of like /|    /|    /|  where the "/" is the grip, with the string being forced out of its rest-position, and the "|" is the point at which the hold of the hair (the grip) breaks, and the string slips .    We don't need to go into the math behind this - but - any repetitive waveform can be constructed from a sum of the fundamental frequency and a (some) number of its harmonics...  here's an animated example of how a saw-tooth wave is constructed building from the fundamental (first harmonic) up to the 50th harmonic - the final waveform getting ever closer to that "perfect" sawtooth....

That is clearly just a mathematical synthesis - in "real life" we don't get perfection like that - the largest difference being the "slip" - which here, is apparently instantaneous (mathematically correct, physically impossible)

OK - so - that was a digression and will let you follow when I talk about the fundamental and its harmonics - so - here's the spectrum analyses of my open G2 octave string - it is not pretty ( nah, it's gorgeous really, tells a lot ) -

Observe that at the far left there are two peaks - the leftmost one is PROBABLY fan noise/vibration (the mic was on the table beside the laptop).

What REALLY got me head-scratching was that the second-left peak, close to the 100Hz marker IS THE FUNDAMENTAL - it is G2 (98Hz)  

Now - from the construction of the sawtooth wave, we really expect the fundamental to be the highest amplitude, with its harmonics dropping off 

But that's not what I see here.   I see the G2 (the fundamental - also known as the first harmonic - of 98Hz, just to left of the 100Hz marker) - but it is REALLY low in amplitude.......

I see G3 (second harmonic in the "saw-tooth waveform additive sine wave component" somewhere to the left of the 200Hz marker - much higher than the first harmonic, but still relatively low)

Things start to "go right" from there on - we have a HUGE G4, then a falling off (the major peaks - ignore the dross at the foot of the spectrum for now) up to the 20th harmonic and beyond ( 20 * 98 = 1960Hz - the highest peak just to the left of the 2000Hz marker)

HOWEVER - once I go to play notes in the normal range a violin is designed for - let's go for the open E ( that's an E4 on octave strings, which would be the E on the D string of a normally strung fiddle ) - we get this truly beautiful spectrum - ( E4 = 330Hz approx - so the first 3 peaks are approx 330, 660, 990... etc )

There is still that low-level "fan noise" down below 100 Hz - but the rest of the spectrum looks fine to me....

SOOOOOO.....

Other observations - playing down on the G and D on the octaves, I can easily feel the vibrations through chin and shoulder.  Obviously the energy is going somewhere - but where it is NOT going is into the fiddle itself to be radiated out as sound energy.

Spectrum measurements taken on the A3 and E4 (on normal strings, A3 is on the G3 string and E4 on the D4 string)  strings are fine- again - these are now of course within the normal range the violin was designed for....

QUESTION:  What happens if I install the octaves on my cheap old EV ?   The EV has no resonant body, just a piezo strip under the bridge.    Before I bother, I should check (a) the piezo transducer response curve (should be available somewhere) to check it will still work down at 98 Hz - I rather expect it will - and (b) a greater unknown - there's an electronics module built-in to that EV - it is JUST POSSIBLE it has a bandpass filter (like the first image in this post) built in to it to tailor overall response - I won't know until I try.....

Of course, that will mean amping it up - oh, that'll annoy the neighbors half-a-mile-away

It's been an interesting day....