PDA

View Full Version : I've just corrected the intonation of every ukulele I own by using a matchstick!



anthonyg
01-29-2016, 07:58 PM
Well, I've JUST ABOUT corrected the intonation of most ukuleles I own by placing a matchstick on the fretboard against the nut.

Someone else gave me the idea, probably from here so I thought I should try it at last.

I have plenty of ukuleles that were built with the saddle placed at the theoretical scale length or slightly short of the mark leaving not enough room for compensation. This is probably one of the reasons I'm crabby about the lack of theoretical knowledge of intonation in the industry.

I had read about and seen nut compensation so what I've done is to move the nut contact point towards the 12th fret, which flattens the intonation which compensates for the intonation going sharp due to a lack of saddle compensation.

Sure its not perfect, yet I'm a little gob smacked at how well it worked. There would be a limit to how far you could go when compensating at the nut. The matchsticks are only 1.9mm thick (measured with verniers) yet its been quite effective for me so far.

Now of course if the nut to centre 12th fret distance wasn't accurate in the first place then your adding significant errors at this end of the instrument. For the record I had measured all the instruments up and most were accurate at the nut end but some weren't. If by chance your intonation is going flat then you could compensate for this by adding a spacer at the nut end.

Definitely try this one for yourself.

LOTS easier and cheaper to fiddle with the nut than move a bridge.

Anthony

Michael N.
01-29-2016, 10:44 PM
I'm not surprised it's improved matters but 1.9 mm is a lot of compensation at the nut. You can buy thin sheet plastic or bone veneer and glue that to the existing nut so that it overhangs the fretboard a little, recut the grooves. My limited understanding of shortening the theoretical position of the nut is that it has more effect on the lower frets, saddle compensation being more effective on the upper frets.
You might be able to recut the existing bridge groove and move the saddle back to obtain better overall compensation, assuming that compensation can't be had by reshaping the saddle itself. That's a little more involved but it was a frequent fix on some cheaper guitars that intonated badly. Of course there is nothing to stop you doing the nut compensation as well.

anthonyg
01-29-2016, 11:22 PM
Well yes I am surprised that moving the nut that far DIDN'T cause more problems than it fixed. Some of the instruments probably did have an error in the nut in the first place so I was partially correcting for that error.

Still, it seems that building a ukulele without any saddle compensation requires a bit of compensation at the nut.

Run your own experiments. So far I've got good results all the way up the neck from the experiment. Not perfect of course, but quite an improvement.

Anthony

Michael N.
01-30-2016, 12:42 AM
Well I don't know how some locate the bridge/saddle slot so innaccurately in the first place. It's not as though saddle compensation is some sort of highly guarded secret. I'd post a picture of my method but it's so simple that a curt explanation should suffice. It just consists of a thin length of wood with a thin block of wood glued at the end. That forms a 'hook' for the nut end of the fretboard. At the other end is another piece of wood that is the same size as the proposed saddle, that slots into the saddle location at the bridge. It's simply a matter of placing these cross pieces of wood in their 'correct' positions i.e. the scale length + a bit (usually around 2 mm's). It's an exercise in simple measuring and marking! A bit of glue helps as well. Clamps not necessary.
The really fine tuning (if you must) can be done on the saddle edge itself and or nut compensation. Of course the frets themselves need to be positioned fairly accurately too but that's another matter.

Timbuck
01-30-2016, 01:36 AM
I just use this simple gauge ..I hook one end into the saddle slot.. Then line up the dots on the 12th fret .. the dot with a line scotched on it is the 12th fret and the other dot is the compensation.. near enough to 2mm .
http://i219.photobucket.com/albums/cc143/shiregreenbod/PICT0021_zps8nulj3h6.jpg (http://s219.photobucket.com/user/shiregreenbod/media/PICT0021_zps8nulj3h6.jpg.html)
http://i219.photobucket.com/albums/cc143/shiregreenbod/PICT0023_zpsxs8qlk6b.jpg (http://s219.photobucket.com/user/shiregreenbod/media/PICT0023_zpsxs8qlk6b.jpg.html)

Michael N.
01-30-2016, 03:07 AM
That's very similar to the one that I made except that I reference it from the nut end of the fretboard. Providing you make these things accurate (not difficult) it's virtually impossible to get the placement wrong, unless you allow the bridge to slide around whilst gluing.

BTW here is a test that you may (or may not) like to take and gives an indication of how good you are at recognising differences in pitch. Try it with headphones or decent speakers. With practice you can get better. I usually come out at 1 HZ but with a little practice I can get it down to 0.75 Hz. Good but hardly amongst the best. I think my results translate to near 3 cents. Do for me! If you can get down to 2 cents or less you'll be amongst the elite.
Getting to that level doesn't worry me. In fact it could be a grace. A bit like not wanting perfect pitch.

http://tonometric.com/adaptivepitch/

Rakelele
01-30-2016, 03:47 AM
I have tried the matchstick trick on my steel string Tenor Guitar as well, and it did improve intonation - or at least, didn't make things worse (dampened the sound a little, though). However, if you calculate the distance of the frets relative to each other with a tool like this one on stewmac (https://www.stewmac.com/FretCalculator), there is no way that the nut can be misplaced by almost 2mm, which is what the matchstick trick suggests. Mathematically, everything seems to be pretty accurately built... :confused:

anthonyg
01-30-2016, 11:16 AM
That's very similar to the one that I made except that I reference it from the nut end of the fretboard. Providing you make these things accurate (not difficult) it's virtually impossible to get the placement wrong, unless you allow the bridge to slide around whilst gluing.

BTW here is a test that you may (or may not) like to take and gives an indication of how good you are at recognising differences in pitch. Try it with headphones or decent speakers. With practice you can get better. I usually come out at 1 HZ but with a little practice I can get it down to 0.75 Hz. Good but hardly amongst the best. I think my results translate to near 3 cents. Do for me! If you can get down to 2 cents or less you'll be amongst the elite.
Getting to that level doesn't worry me. In fact it could be a grace. A bit like not wanting perfect pitch.

http://tonometric.com/adaptivepitch/

I agree that in isolation, a 2 or 3 cents difference is very difficult to hear. However when your playing chords the difference starts popping out. Look, I'm not really complaining about a 2 cent error. What really annoys me and is really EASY to hear is when you play a chord such as Amin (2,0,0,0). You have 2 A's and if the fretted A goes 4 or 5 cents sharp or more with even light pressure then it stands out like a sore thumb.

When you guys are setting up your ukuleles try testing the fretted A against the open A. The fretted G against the open G and the Fretted C against the open C. (open position of course)

Anthony

anthonyg
01-30-2016, 11:27 AM
I have tried the matchstick trick on my steel string Tenor Guitar as well, and it did improve intonation - or at least, didn't make things worse (dampened the sound a little, though). However, if you calculate the distance of the frets relative to each other with a tool like this one on stewmac (https://www.stewmac.com/FretCalculator), there is no way that the nut can be misplaced by almost 2mm, which is what the matchstick trick suggests. Mathematically, everything seems to be pretty accurately built... :confused:

Yes I'm sure that a matchstick is a little soft for steel strings. Yes, your deliberately introducing an error at the nut to cancel an error at the saddle.

Mind you if it works it works. One thing I am concerned about and could happen is a concept I will call "crossed curves" I've pinched this from photographic theory. With colour photography you can graph the colour response from the 3 colours used in film. Red, Green and Blue. Plotted on a graph they form curves which should all be parallel. If these plotted responses ever cross each other you end up with say a green tinge in the shadows and a red tinge in the highlights. Pretty ugly really.

SO, intonation adjustments at the nut or saddle that are too radical due to compensating for errors in construction, COULD lead to intonation being flat at some frets and sharp at others. Something to watch for.

So far my matchstick technique hasn't lead to any such problems.

Anthony

seattle
01-30-2016, 11:41 AM
The posting of this thread was good timing for me. I just got a tourist grade instrument called a charango. It's a 10 string, 5 course instrument with a soprano sized body and tuned similar to a uke.

The G string wouldn't ring out when not fretted. I put a 10mm sized piece of a wood match stick just in front of the zero fret and now the string rings out.

I think the real problem is the left side of the first fret is too high (not even). When I change strings I'll see if I can tap that down and then I won't need the matchstick trick.

For now, it's gets me playing. :)