Sheldon Brown's Wheelbuilding page links to the article: Check Spoke Tension by Ear. Obviously a tensiometer would be ideal, but Sheldon suggests that checking spoke tension by ear is a reasonable alternative. The article was written in 1987 and updated in 1997; are there any developments from the last 10 to 20 years that would invalidate the technique or require adaptation? For example, the article mentions that dished wheels require a different pitch on the left and right sides, but how great should this difference in pitch be? In general, how effective is tensioning by ear?
That depends on how good your ear is. If you can tune a stringed instrument effectively, then tensioning a wheel by ear is very effective. Identical spokes that have the same pitch when plucked should have the same tension within the margin of error for any tensiometer reading. The catch is you need a tensiometer to get the relative tone for the proper kg/f tension on the spokes, unless you have perfect pitch and can remember up to four different tones for a single wheelset.
Keep in mind that for a wheel with any amount of dish, the tension, spoke length, and therefore the pitch, will almost certainly be different from one side of the wheel to the other. Therefore, once you have your tone for a given kg/f reading it will only be accurate for one side of one wheel (there are exceptions, such as a front road wheel, or a flip flop rear wheel). Flange heights and spacing vary from front to rear hub and side to side as well, also with some exceptions. Also worth noting that different gauges and shapes of spokes of the same length and tension will have a different ring to them.
In short, don't assume that the pitch of properly tensioned spokes on one side of one wheel apply to the rest of the spokes on the wheelset, and certainly don't use the pitch of a spoke on a wheel as a reference for a completely different type of wheel.
Anecdotally, I was discussing this very issue with a friend of mine the other day. He said that he believed tensioning by pitch gave a valid reading, but that he couldn't do it because he was mostly tone-deaf. So to reiterate, if you don't have a pretty good ear musically, then this isn't a viable method of checking tension.
To summarize, with a good ear you can use pitch to check tension on spokes, but it's more of a timesaver than a replacement for a tensiometer as you need a baseline tone at proper tension to check the rest of the spokes against.
The method I use (still need to get a tension meter to verify) is as follows:
The formula for the connection between tension (KgF) to frequency is
f = 1/(2L)*sqrt(T/u) where L is the length of the spoke, T is the tension (in Newtons) and u is the linear density of the spoke. (http://en.wikipedia.org/wiki/Vibrating_string)
you can get the linear density by dividing the spoke mass by its length.
you can get the tension T by multiplying the KgF in 9.98 to get it in Newtons.
It is important to use all the length measures in meters and all the masses in Kg.
for example for 140 KgF and 260 mm spoke that weighs 6.5g you get the frequency to be (1/2*0.260)*sqrt(140*9.98/0.025)~450Hz which is approximately A4.
this results also correlates with John S. Allen's article http://www.bikexprt.com/bicycle/tension.htm.
If anyone gets to check this with a tension meter to confirm it would be helpful.
How about this?
It is an iPhone based spoke tension gauge that uses sound to determine tension.
All the wheels I built by ear were within the tolerance when I finally got a Park Tension Meter. However, I can build much more consistent nowadays with the meter. Somehow the range of tension from one spoke to another is so much better that I would never like to build without it ever again. I think the wheels will last much longer with consistency from one spoke to the next. So don't leave any loose ends, just buy the meter.