How does a violin reproduce overtones? - Theorizing a model

[Andreas Preuss]

12 hours ago, Don Noon said:

First iteration of the latest Frankenfiddle test (detailed previously) is done.  As predicted, results are not hugely conclusive or good...

Don, thanks for your experimental contribution here. Just for completeness I would like to know 

bridge height

(I suppose you used standard bridge feet width of 42mm) 

bridge weight

string angle

string brand(s) and thickness

distance between f holes at upper lobes. 
 

Thanks.

[Marty Kasprzyk]

2 hours ago, Carl Stross said:

From Amazon : 

https://www.amazon.com/Soundpost-Violin-John-Ph-d-Mclennan/dp/1543403107

"This book dispels the mystery that the position of the sound post in determining the sound quality of the violin is critical."

 

This tells me all I need to know about his ability to discriminate sound quality...............

 

Apparently the Amazon book reviewer didn't actually read the book.

It is well known that tiny soundpost position adjustments can change the sound heard by the player.  Some players are so sensitive they can hear a sound change when the soundpost isn't even moved.

[Carl Stross]

1 minute ago, Marty Kasprzyk said:

Some players are so sensitive they can hear a sound change when the soundpost isn't even moved.

That would rather be the norm, I think. Good point !

Nonetheless, I have to suppose the author agreed with the review. Reminds me of the countless "explanations" from Woodhouse, all of them pretty useless. 

[Don Noon]

13 hours ago, Marty Kasprzyk said:

I recommend  John McLennan's 2017 book "The Soundpost in the Violin"

I'll pass.  I did find a free article of his "The violin (as I understand it)", where he states:  "The most important sound producing function of the violin is the monopole action."

I suppose at age 90+, you might not be able to hear anything over 600 Hz, and therefore it might be true.

Andreas, I'll post the details you want when I get the Frankenfiddle back together again.  Regraduation is done, new bass bar soon.

[Marty Kasprzyk]

9 minutes ago, Don Noon said:

I'll pass.  I did find a free article of his "The violin (as I understand it)", where he states:  "The most important sound producing function of the violin is the monopole action."

I suppose at age 90+, you might not be able to hear anything over 600 Hz, and therefore it might be true.

Andreas, I'll post the details you want when I get the Frankenfiddle back together again.  Regraduation is done, new bass bar soon.

Another criticism of a book from someone who hasn't read it.  

The book is just a literature review of all sound post studies done up until 2017.  McLennan's hearing ability is irrelevant.

[jezzupe]

1 hour ago, Don Noon said:

I'll pass.  I did find a free article of his "The violin (as I understand it)", where he states:  "The most important sound producing function of the violin is the monopole action."

I suppose at age 90+, you might not be able to hear anything over 600 Hz, and therefore it might be true.

Andreas, I'll post the details you want when I get the Frankenfiddle back together again.  Regraduation is done, new bass bar soon.

By "sound producing function" does he mean amplitude? or "tone" ? the way the sentence is stated could be assumed he meant one or the other, if we were talking about "sound" meaning volume, I might be inclined to agree with him, if we mean "tone" , then no I would dismiss that too, I assume you are thinking in terms of "sound" as "tone producing function" ? I'm not sure, it's not a very good sentence, did the article extrapolate on what he meant by "sound producing function"?

[Don Noon]

3 hours ago, Marty Kasprzyk said:

Another criticism of a book from someone who hasn't read it.  

The book is just a literature review of all sound post studies done up until 2017.  McLennan's hearing ability is irrelevant.

The book might be OK, I don't know.  I can find out what I need to know about the soundpost by my own tests and moving it around, and don't need the book.  My criticism was of a quote elsewhere indicating his understanding of what's important, which differs from what I'm quite sure is more important.  The hearing thing was a joke, and mine is heading in that direction.

Jezzupe, google the guy's name and poke around.  You can find a lot of free articles he has written, if you're interested.  Or you can google the title of the article.

[Andreas Preuss]

On 8/16/2021 at 10:00 AM, Marty Kasprzyk said:

You're right. The soundest prevents its side of the top plate from vibrating much thereby reducing out of phase far field sound cancellations. The sound post increases the sound output.

If you like a more detailed explanation I recommend  John McLennan's 2017 book "The Soundest in the Violin"

I've never met him but he's an inspiration to me now that I'm getting older.  He got his PhD at the age of 84 (!) with his thesis "Violin Acoustics from Baroque style to Romantic style instruments".  He along with Oliver Rogers. Norman Pickering and Carleen Hutchins have shown that actively doing violin research increases your life span. 

 

 

 

 

 

 

Thanks for the book recommendation. I am still chewing on your last recommendation ‘Why we hear what we hear.’ so I am still busy with digesting. (Still so much interesting stuff to read! And it is written in a way that a simple mind like mine can understand most of it.)

Concerning a whole book talking about only the sound post and written by someone who presumably didn’t do too many sound post adjustments himself (if any) I rather stick to what I got in my finger memory to move and adjust sound posts. I would be afraid too that I might fall asleep after reading 10 pages of that book… 

Anyway, so far I can say string angle matters much more than sound post adjustments. At least it makes sense to change first the string angle and then readjust the sound post. I don’t know anyone who would do it the reverse. (To me this logic says all)
 

I think for each body structure there is one optimal string angle. The more sensitive the body the more important this becomes. On a well balanced body structure the right string angle will bring out all sound characteristics a player wants to have, on an imbalanced or badly proportioned body this will have limits. Sound post adjustments are in this game the finish touch. (Still important enough)

[Don Noon]

25 minutes ago, Andreas Preuss said:

Anyway, so far I can say string angle matters much more than sound post adjustments. At least it makes sense to change first the string angle and then readjust the sound post. I don’t know anyone who would do it the reverse. (To me this logic says all)

Exactly what is involved when you change the string angle?  Resetting the neck for overstand and projection?  Diddling with the saddle height?  Changing the bridge height?

I can't imagine much effect with the first two... testing saddle height didn't do much, and I'm not about to test changing the neck set.  Absolutely a different bridge height would matter, and you'd have to mess with the neck angle to match... but string angle is more of a byproduct of the important variable, and not a cause of the change.

[Davide Sora]

21 minutes ago, Andreas Preuss said:

I think for each body structure there is one optimal string angle. The more sensitive the body the more important this becomes. On a well balanced body structure the right string angle will bring out all sound characteristics a player wants to have, on an imbalanced or badly proportioned body this will have limits. Sound post adjustments are in this game the finish touch. (Still important enough)

I think you mean a sensitive body needs a more open angle (less pressure) and vice versa, am I right?

But what do you think are the minimum and maximum acceptable limits for an effective angle? There doesn't seem to be much room for variation.

[Andreas Preuss]

8 hours ago, Don Noon said:

Exactly what is involved when you change the string angle?  Resetting the neck for overstand and projection?  Diddling with the saddle height?  Changing the bridge height?

Technically this is a little bit of a headache because it can't be done very quickly. 

The easiest way to change the angle is changeing the lower saddle height, but there are limits because a very high saddle can flip over. For experimental purpose I invented a saddle which is not glued on and its height can be changed quickly by adding slivers of wood underneath. Though audible changes are rather minimal it can give indications whether it makes sense to go further by changeing the neck angle

On the neck I do it usually with the raise pitch method to go higher. Going lower is only possible when you already have a sliver of wood in between neck heel and top plate. While figureing out the best setting, I don't glue the sliver so that it can be easily removed. Neck heel overstand comes more or less from experience values, and I don't think there is any better method. Roughly speaking, I see a need to go higher on high tension strings. More or less this seems what restorers have been done to counterbalance tighter strings on famed fiddles to soften the down pressure from the bridge. (Or, what used to be maybe 4-5mm neck overstand in the 19th century became slowly 6-6.5, and nowadays with some really strong strings it might be advisable to go even higher)

[Andreas Preuss]

8 hours ago, Davide Sora said:

I think you mean a sensitive body needs a more open angle (less pressure) and vice versa, am I right?

But what do you think are the minimum and maximum acceptable limits for an effective angle? There doesn't seem to be much room for variation.

Mostly I think in terms of the top plate and how it is framed. On a soft frame and thin and flexible top this must be very low and on a stiff top with a stiff frame this must be very high. 

I think there is more variation than what we think. It is mostly a matter of how religiously you follow standard measurements for the neck overstand. Besides, I had moments in my experiments when lowering the bridge only 1mm made some difference, and therefore I think string angle is pretty sensitive, especially if you build the structure thinly.

My new concept violin project was mostly trying to figure out how thin and light you can go, and if you do so what is necessary to make it work. On this path I ended always up to adjust the neck angle and the bridge height to get the best results for a given situation. (If I hadn't done so, I had probably tossed the project since long) While this didn't bring so far the final result I was looking for, it taught me on which concept I have to work on to get there. It is a slow mental crystallization process and I think I can see now the end of the tunnel to get the result I want.)

[Don Noon]

On 8/16/2021 at 4:49 AM, Andreas Preuss said:

Don, thanks for your experimental contribution here. Just for completeness I would like to know 

bridge height

(I suppose you used standard bridge feet width of 42mm) 

bridge weight

string angle

string brand(s) and thickness

distance between f holes at upper lobes. 

32mm bridge height, 42mm width. Bridge weight 1.66 g (lighter than I would like, but that was the only one in my junkpile that sortof fit the top and was high enough).  String angle varies between 155 and 157 degrees. Strings are extremely old Evahs, except the A string is probably only a year or two old (replacing one that unravelled)... but that wouldn't affect impact spectra.  Eye spacing 42mm.

The regrad is done, taking off 5.4g from the plate and .25g out of the bass bar (new bar).  Almost all of the mass was taken out of the upper bout.  M5 of the bare plate went from 332 to 316 Hz.  I also opened up the F holes, as they were just roughly done on a scroll saw.

I won't bother with spectral plots, but the notable changes:

A0 gained 7 Hz (due to opened F holes) and gained some amplitude, ~2dB

B1- dropped 4 Hz and also gained a dB or two

B1+ stayed exactly the same frequency and amplitude

Some things moved around in the transition hill, but not much; bridge hill also moved around a bit and overall slightly weaker.  (Edit: soundpost might be slightly closer to the bridge than before, giving this effect)

Total effect:  some extra meat on the G and D strings, A about the same, slightly weaker E.  But it still kinda sounds the same.

Conclusion:  inconclusive and confusing (as predicted) with regard to controlling the bridge hill part of the spectrum.  What IS fairly repeatable is that thinning a top plate enhances the lowest frequencies.  Whoopee.  Time to get back to real work.

[Salve Håkedal]

59 minutes ago, Don Noon said:

...

Whoopee.  Time to get back to real work.

Yeah. But we love following your experiments!

[Andreas Preuss]

8 hours ago, Don Noon said:

32mm bridge height, 42mm width. Bridge weight 1.66 g (lighter than I would like, but that was the only one in my junkpile that sortof fit the top and was high enough).  String angle varies between 155 and 157 degrees. Strings are extremely old Evahs, except the A string is probably only a year or two old (replacing one that unravelled)... but that wouldn't affect impact spectra.  Eye spacing 42mm.

The regrad is done, taking off 5.4g from the plate and .25g out of the bass bar (new bar).  Almost all of the mass was taken out of the upper bout.  M5 of the bare plate went from 332 to 316 Hz.  I also opened up the F holes, as they were just roughly done on a scroll saw.

I won't bother with spectral plots, but the notable changes:

A0 gained 7 Hz (due to opened F holes) and gained some amplitude, ~2dB

B1- dropped 4 Hz and also gained a dB or two

B1+ stayed exactly the same frequency and amplitude

Some things moved around in the transition hill, but not much; bridge hill also moved around a bit and overall slightly weaker.

Total effect:  some extra meat on the G and D strings, A about the same, slightly weaker E.  But it still kinda sounds the same.

Conclusion:  inconclusive and confusing (as predicted) with regard to controlling the bridge hill part of the spectrum.  What IS fairly repeatable is that thinning a top plate enhances the lowest frequencies.  Whoopee.  Time to get back to real work.

The conclusion is that we have to look somewhere else. because overtone spectrum is not only generated from the material we use there is something in the construction. 
 

Besides, I would lower on your experiment the string angle to 158-160 degrees and measure the graph again. Experimenting with overtones is tricky because it DOES involve setup parameters.

ADDITION

If I think about it, there might be a mistake in the experiment. Do you agree that thick tops need more bridge pressure than thin tops when the arching is a constant value? 

This means, if you want to make top in an experiment thinner then the string angle should be lower than the starting thickness demands. In other terms I see in your experiment from the beginning too much pressure from the bridge to get a real conclusion.

This might be as well the mistake I made on my experiment scraping from the outside getting no conclusive result. 

But in the end I still think there are better trigger points for overtones.

Edited August 18, 2021 by Andreas Preuss
addition

[Don Noon]

1 hour ago, Andreas Preuss said:

I would lower on your experiment the string angle to 158-160 degrees and measure the graph again. 

158 - 160 degrees would be an increase in string angle... meaning a new, taller bridge or decreased neck overstand.  I can't lower the saddle, since there is none currently.  If you want to pay for shipping, I'll send it to you so you can reset the neck or make a new bridge.  IMO, the sound capability is mostly fixed in the wood and arching, and don't expect it to change much (although I'm sure it could be made worse with radical deviations from normal elsewhere). I'm all experimented out for now, and have stuff to build for clients.

I will keep an eye on this fiddle to see if it settles in and changes over time.  With a new top, even torrefied, there is usually still the issue of tonal change after it is first strung up.

[Don Noon]

oops... double post

[Andreas Preuss]

47 minutes ago, Don Noon said:

158 - 160 degrees would be an increase in string angle... meaning a new, taller bridge or decreased neck

?

160 degrees is closer to 180 degrees (=flat) than 155, or not?

Regardless I mean flatter. 

You can also loosen the back around the upper bout and pull it out, though this needs some forced reglueing only recommended on experimental violins.

[Andreas Preuss]

1 hour ago, Don Noon said:

If you want to pay for shipping, I'll send it to you so you can reset the neck or make a new bridge. 

How much is the shipping? I am serious. PM me.

[Don Noon]

35 minutes ago, Andreas Preuss said:

?

160 degrees is closer to 180 degrees (=flat) than 155, or not?

Regardless I mean flatter. 

You can also loosen the back around the upper bout and pull it out, though this needs some forced reglueing only recommended on experimental violins.

Brain fart.  Yes. you're right.  I could reset the neck (ugh) or add a saddle... which would be a vastly different effect than lowering the bridge to acheive the same string angle.

[Andreas Preuss]

On 8/18/2021 at 11:52 PM, Don Noon said:

Brain fart.  Yes. you're right.  I could reset the neck (ugh) or add a saddle... which would be a vastly different effect than lowering the bridge to acheive the same string angle.

Just following up. Did you make a lower saddle to reduce the string angle?

[Don Noon]

52 minutes ago, Andreas Preuss said:

Just following up. Did you make a lower saddle to reduce the string angle?

No... that would not be possible, since it had no saddle to start with.  I did ADD a saddle, which INCREASED the string angle slightly, with no apparent effect.  I also added a coat of varnish, which had the usual effect of slightly less dB, but more refined tone.  It's a good fiddle now... but still really ugly.