Welcome to our forum. A Message To Our New and Prospective Members . Check out our Forum Rules. Lets keep this forum an enjoyable place to visit.
Check out our 2023 Group Christmas Project HERE
Log In
Register
Home
Topic RSS
(0 votes) 
Regulars
Offline
A while back there was a @Fiddlerman poll about whether people would be interested in a violin that played an octave lower than normal.
Last night I noticed in The Cambridge Companion to the Violin, p.255, a comment that from 1969 Jean-Luc Ponty employed a "violectra", which was an electric instrument that sounded an octave below the violin.
However, Wiki says "He sometimes uses a six-string electric violin called the Violectra, with both the low C and low F strings (not to be confused with the "violectra" he played from the late 1960s to the mid-1980s that had the usual four strings but tuned an octave lower)"
Also Wiki: The original Violectra "was the trade name of an electric violin produced by Barcus-Berry with the pitch equivalent of an acoustic tenor violin, sometimes called baritone violin. It is tuned an octave below normal violin,"
I think I answered 'no' to the poll, but I can't remember why. Perhaps I wanted to try a viola first?
=====================================
Lol, I've just noticed to the right of this post: -
Would you be interested in a violin which tunes and plays one octave lower? Octave violin.
- Yes (56%, 425 Votes)
- Maybe (23%, 176 Votes)
- No (21%, 164 Votes)
Total Voters: 765
Andrew
Offline
The attached image is a spectral trace of the open G2 - at the extreme left, there is a low amplitude component, just below the 100Hz marker - that's the G2 fundamental. In fact, the G3 isn't overly present in the sound either when you notice the contribution of the 3rd, through the 6th harmonics. ***CAVEAT*** these tests were performed during tests using a cable-style super-lightweight tailpiece ( via @Irv ) and the dampening of these harmonics is likely to be significantly less (making them more prominent) than would be observed on a normal tailpiece. The VERY low level of the fundamental G2 at 98Hz, however, is surely due to the body response curve - it's acting like a huge bass-cut, compared to the "lift" across the normal frequency range of the instrument.
Finally, just for fun - if you havent seen it (don't recall if I posted it to the forum or not) - here's the Octaves in play...
I seriously recommend not copying my mistakes. D'oh - 
Please make your own, different mistakes, and help us all learn :-)
Regulars
Offline
There is mystery involved in a violin's low G. We are told that the output of the fundamental is nearly zero and (like a piano's low notes) our brains reconstruct (i.e. imagine) the fundamental from the overtones. Yet if you pluck the G string and let it settle, you can only be getting a fundamental sinewave from it, and it is audible.
What is going on?
I guess I'm assuming that the fundamental, rather than missing, is just quieter than the overtones, and spectrum analysis needs interpreting, just like any other graph or statistical representation of numbers.
Another possibility is that the distorted rectangle caused by bowing doesn't produce a fundamental, but that would be true of the other three strings also, or the D string, at the very least.
Andrew
Offline
Gordon Shumway said
There is mystery involved in a violin's low G. We are told that the output of the fundamental is nearly zero and (like a piano's low notes) our brains reconstruct (i.e. imagine) the fundamental from the overtones. Yet if you pluck the G string and let it settle, you can only be getting a fundamental sinewave from it, and it is audible.What is going on?
I guess I'm assuming that the fundamental, rather than missing, is just quieter than the overtones, and spectrum analysis needs interpreting, just like any other graph or statistical representation of numbers.
Another possibility is that the distorted rectangle caused by bowing doesn't produce a fundamental, but that would be true of the other three strings also, or the D string, at the very least.
Indeed @Gordon Shumway - I'm not well up on this (the "psychology or physiology" if it is even that, of hearing) - but yes - in some ways this "fill-in" of the sub-tones appears to be similar is the famous guitar power chord 5ths (best obtained when heavily over driven and distorted LOL) - oh - its like the beat frequency brings in an octave lower - hang on - example easier - say you play A440 and a 5th up on E660 (nice easy numbers !!!) - sure - played together the beat freq has to be 220Hz - i.e. A3 - I get that.... and, that surely is "real" - there are two sound sources in action, beating against each other. But, what I also don't wholly get is exactly what you are questioning...
--- and - oh - this is a trap I regularly fall into on spectral analyses - the spectrum peaks in the image in my earlier post are NOT JUST G2, G3, G4, G5 etc - the individual peaks are of course G2 (in Hz) apart - so we have in the spectrum earlier -
G2 ( 98Hz ), G3 ( 196Hz ), D4 ( 294Hz), G4 (392Hz), B4 - rather "flat" but "just!"(490Hz), D5 (588Hz), F5 (rather flat) (686Hz), G5 (784Hz) and so it goes on, with lots of interesting, but increasingly smaller overtones the further we go... [ my "rather flat" comments relate to the 12-TET values we commonly "accept" as being the frequencies of notes, and as used (largely) on keyed instruments. ]
So, nothing new there to either of us, that's just how it is....
I'm having to question your assertion that
"Yet if you pluck the G string and let it settle, you can only be getting a fundamental sinewave from it"
... and trust me on this - I'm not saying you are incorrect - I just have to question this, largely because I had never really thought about it... I am guessing that regardless of how gently we pluck the string, there has to be some initial impulse taking the string from rest (and causing an initial displacement) and imparting energy into the string. The point at which the string is plucked, and indeed just HOW it is set in motion may well have something to do with this and what we hear... I do not know, this is all conjecture on my part. I am going to have to go do an experiment this afternoon, once I have finished painting the ceiling, oh, and cleaning the now white-paint-blotched keyboard on my laptop... D'Oh....
I seriously recommend not copying my mistakes. D'oh -
Please make your own, different mistakes, and help us all learn :-)
Offline
Well, I'm glad you raised thoughts about this Alf @Gordon Shumway - interesting, and thanks! Isn't it great to share ideas and thoughts !
It appears that the plucked string has a significantly (very significantly - remember - the vertical scale is dB, not linear) higher contribution of the fundamental - however - no - it is not sinusoidal by any means. It still shows overtone and harmonic contributions leading to a complex waveform, although, quite different compared to that of the bowed note.
I have other related images from the very quick experiment, but pointless sharing them here.
I did try exciting the string by tapping on the instrument, with the other 3 strings deadened - but - the largest contribution I was able to capture was the ringing of the body.
It MIGHT just be possible to excite the G alone by driving it (the violin body, I mean) with an audio source (like a hacked-up old loudspeaker with the cone removed and the moving coil assembly in physical contact with the body). Perhaps a pure sinusoid (impulse burst, not continuous) at 1/2 the G3 frequency in an attempt to excite the G3 ???? The driving G2 contribution could then be subtracted from the recorded response, and what would be left would be whatever was emitted from the G3 and the body.... Now THAT may well be the fundamental sinusoid only, of close to it... but - no, I'm not going there... LOL
Yeah, I think it all has to do with the action of the pluck, as the string is initially deformed and stretched, then suddenly released, and the impulse must surely travel both towards the bridge and the nut, arriving at different times, and being reflected back (and continually being "echoed" as the energy wave arrives at the fixed closed nodes of the bridge and nut, to gradually die away), leading to the more complex vibration mode of the string.... Someone somewhere MUST have a proper research paper on this - but I'm happy with my currently slightly improved understanding !!
**EDIT** for some reason, my composite-results image (just a .jpg) would not load as an attachment - it may be too large - so I have just this minute placed it on DropBox - here is the global link - https://www.dropbox.com/s/o1ia.....s.jpg?dl=0
I seriously recommend not copying my mistakes. D'oh -
Please make your own, different mistakes, and help us all learn :-)
Regulars
Offline
@Billyg how does your pickup or Mic contribute to your readings? if the range of the Mic falls off on the low end or has a bump in its top end that could be skewing the data. not saying thats happening but as conversation its a possible contributor to your graph. ideally youd want one that has a flat response curve at least from 20-20k assuming a thats human hearing range I would think.
thoughts?
Offline
I seriously recommend not copying my mistakes. D'oh -
Please make your own, different mistakes, and help us all learn :-)
Regulars
Offline
" you make a good point (naturally)..."
haha...more like captain obvious or the old blind squirell and acorn saying...but thanks! just making sure. I figured you had point covered.
Offline
In use in my music room as I mentioned earlier - yes - there are small "lifts" - but really not overly significant - I guess they come from the localized echo/reverb and the various hard surface reflections in the room (I like the sound like that)- and that's the way the tests on the plucked string were done. But really - on the log dB scale - these artefacts are close to unnoticeable - and in fact can be swamped even by the display software in Audacity ( not the actual MEASUREMENT side - that's as good as it gets - and if you NEED the real data values, it's all there for export as text / csv format).
However mentioning the "display software", it is a point worth noting for folks who may "read too much" into "relatively small variations" between traces - what happens is a "rounding error" occurs as you resize the display in Audacity vertically (to expand it) - and - occasionally two adjacent and "close-valued-but visually different in height by a few pixels" will suddenly swap-over - with the low peak now taller, and the taller peak now lower! It's just math rounding, and not unexpected since we're dealing with values in a 32-bit floating point number range, and re-plotting that on a logarithmic vertical scale - not surprising at all.... but something folks should be aware of before rushing to judgement when looking at such graphs!
I seriously recommend not copying my mistakes. D'oh -
Please make your own, different mistakes, and help us all learn :-)
1 Guest(s)
Mouse
tonyluo
ELCBK
