Adjusting sound

[kiwigeo]

As an experiment try plucking the E string at different points along it's length and see how the fundamental and harmonics look on the spl vs Frequency plot.

[Dave M]

I had a mis-sounding string the other day which took quite a while to find.

It needed a magnifying glass to spot a tiny, shallow bevel on the body side of the saddle, causing a sitar effect. Quick file to get a clean beak point fixed it. I was doing a nut set up at the time and was convinced the sound was coming from the nut. They can be damned hard to pin down sometimes.

[Mark McLean]

While TallDad is now off chasing a buzz, this thread has kicked me into gear to do some Frequency Response Curve (FRC) analysis of the falcate braced nylon guitar that I built about a year ago and have been a bit unhappy with the sound (top 3 strings lack clarity and sustain – especially the G string).
I had downloaded VA to my laptop a while ago but didn’t have a microphone to hook up to it, so I never got around to using it. On Craig’s suggestion I also have now loaded the FFT app onto my iphone – and found that using it is a piece of cake. In 10 minutes I was getting all sorts of interesting data, and I think I have diagnosed some issues with my Engelmann/sassafras nylon falcate. But I haven't done this sort of stuff before - so please tell me if I have got it all wrong.......

Here is the FRC plot. I think a relevant issue is that it has a fairly live back (sassafras not being very stiff). This is a L-00 sized guitar – i.e. fairly small bodied.

My reading of this is that the peak at 108 is the main coupled air response, and then there is a double/overlapping peak at 192 and 230. I think the 192 is the main top resonance – which is a bit higher than ideal, and rather close to the frequency of the open G string (196). That might have something to do with the poor sound around that frequency? I also hypothesized that the 230 peak was the resonant frequency of the live back. This seems to be confirmed when I repeat the FRC with the back dampened by a pillow – the 230 peak disappears.

And if I dampen the top and just tap the back – lo, a strong peak at 230.

So – could my sound problems be related to the main top resonance being too close to G, and also fairly close proximity of the top and back resonances? It seemed like a good idea to bring the top resonance down a bit. I could also consider killing the back resonance by adding some bracing – or, alternatively, always playing it with a pillow on my lap.

I consulted the Good Book looking for divine guidance. In the Gospel according to Trevor, Chapter 2, Figure 2.4-2 I received the wisdom regarding the effect of side mass. I had installed side mass loading struts when I built it, so now it seemed like a good time to make use of them. I picked up some lead fishing sinkers from the local tackle shop and drilled them out to take a wingnut. Two weights at 215 grams each should be enough to make a bit of a difference.

weight.jpg (86.71 KiB) Viewed 11585 times

Here is the repeat FRC once 430 grams of side mass is screwed in place. As prophesized by Saint Trevor, the main top resonance has dropped from 192 to 180. The other peaks are unchanged, but the top and back are now a bit further apart. I tell you, Brothers and Sisters, this is some miraculous sh*t right here.

So – does it sound better? That is an entirely subjective matter – and I also changed the strings while doing this work, introducing another confounding variable. However, yes, praise the Lord - I think it is better.
And this is now one heavy guitar.

[johnparchem]

Interesting work Mark. It is fun to see the graphs play out and the effect of the mass. In your case I like the idea of stiffening the back. You might flatten the lower bout back brace a bit and glue a brace on top to make the back a bit stiffer. The top will rise as well (but you can tune it to 190 with some mass) You should be able to separate the top and back peaks. I like the overall sound of the 180 Hz classical guitar I made but they sort of wash out on the high notes on the E-string. I find the 190 the best balance for my taste in sound.

[Mark McLean]

Thanks for the comments John. I am totally new to this stuff so I am learning a lot. It has been really fun to play around with. I know most people are collecting their FRC data using Visual Analyser. I found the FFT app to be great to use on my phone. It requires almost no set-up, and you can add a title, capture the plot as a screenshot, and then email it to your computer. I think it cost me $AU16.

[Jim watts]

Thanks for posting that Mark, i's fun to see this stuff applied. I haven't used the weights in a guitar yet as the overall weight of the instrument kind of goes against my grain, but between this and the book I may have to give it a try.

[Mark McLean]

Jim, it is still a whole lot lighter than a stratocaster. And no-one seems to worry about the weight of them.

[Jim watts]

True enough!

[TallDad71]

I learnt so much from this thread, thanks.

I solved my problem in the end. Unfortunately it had nothing to do with looks at frequency curves and everything to do with removing the volume control for the acoustic amplification that was stuck underneath the sound hole.

Lesson Learnt : Have a little more faith in the work you do and a little less faith in factory built components from China!

Thanks again.

[johnparchem]

I learnt so much from this thread, thanks.

I solved my problem in the end. Unfortunately it had nothing to do with looks at frequency curves and everything to do with removing the volume control for the acoustic amplification that was stuck underneath the sound hole.

Lesson Learnt : Have a little more faith in the work you do and a little less faith in factory built components from China!

Thanks again.

Whoops sorry I never knew you were talking about an amplified sound.

[lamanoditrento]

I used Visual Analyzer at first but switched to the FFT app by Andrew Smith a couple years. It is about $10. As far as I know, it works only on Apple iOS, iPhone and iPad. VA is kind of quirky to set up and use. FFT by Andrew Smith requires minimal set up, very intuitive to use, saves files for future interactive recall and overlays, saves screen shot images to camera roll. Drag a guideline in from the side of the screen for quantifying peaks. X & Y axis can be zoomed in/out and shifted left/right/up/down with smart phone gestures like pinch, spread, touch and drag. Because it is on my phone, it is always handy and ready to go. Accuracy seems every bit as good as VA. I use FFT on an iPhone 7+ now which is a big improvement over the smaller iPhone 5. Any of the iPads would be good too, but I like having it in my pocket.

Hi Craig, what settings are you using on FFT?

[edit: quoted the wrong post]

[Mark McLean]

Trent
Craig will have more experience - my recent posts were my first experience with it and I chose settings pretty much by trial and error. However, to get you going here are some screen shots of the settings I used. I set smoothing at 1/12 octave, and graph decay at "peak hold".

[

If Craig says something different - do what he says, and so will I next time.
Mark

[lamanoditrento]

Trent
Craig will have more experience - my recent posts were my first experience with it and I chose settings pretty much by trial and error. However, to get you going here are some screen shots of the settings I used. I set smoothing at 1/12 octave, and graph decay at "peak hold".

Cheers Mark!

[Craig Bumgarner]

FWIIW, I use the default settings on the settings page. On the main display, I use No Smoothing on the smoothing options in the lower left corner and Peak Hold in the lower right. I have not utilized the many custom options because 1) FFT seems to work fine with the defaults and 2) I'm lazy. Don't let that stop you from trying though, there are probably some improvements hiding in the options that I have not even considered.