WX5 Reed Bend vs. Displacement

The windlist has often debated the reed/mouthpiece performance of the WX5 compared to older Yamaha wind controllers like the WX7 and WX11. Of particular controversy has been whether or not the WX5 reed/mouthpiece has a wide "flat spot", a reigon of little or no bend change versus bite pressure (displacement).

Many of us old-timers (though not all) who play the WX7 and WX11 feel the WX5 has a significant "flat spot", and that this flat spot is very wide, and hence makes subtle vibrato difficult compared to the WX7 or WX11. One theory for this wide "flat spot" is that it's easier for new players to adjust the bender offset (lip zero) to no-pitchbend.

Looking at how the WX5 reed interfaces to the mouthpiece, it's pretty easy to see how the reed quickly flattens, and with little pressure required. It is also easy to see that for extreme bite pressure, the final curve in the beak of the mouthpiece is markedly sharper (higher slope) than in the lower part of the mouthpiece lay. Several players, including me, have indicated that by observing lay of the mouthpiece, the existance of a flat spot is clear. But other players, some with years of professional experience with several Yamaha controller models, swear that in a configuration of max lip gain and biting on the tip of the mouthpiece, it's possible to play with no flat spot.

So, to eliminate the subjective, I've started a series of technical experiments to quantify reed bend vs. displacement, and graph the results. My data (below) does prove the existance of a "flat spot". The current experiments that have been done include:

12Nov2000 Experiment

  • WX5 reed displacement vs. midi bend data

25Nov2000 Experiments

  • WX5 reed displacement vs. midi bend data (repeat of original)
  • WX5 bend value with adjusted offset (from original nominal)
  • WX5 bend value with adjusted gain (from original nominal)
  • WX5 reed sense lever only

27Nov2000 Experiments

  • WX5 reed displacement vs. midi bend data (repeat of 25Nov data)
  • WX7 reed displacement vs. midi bend data

Future Experiments

  • WX11 reed displacement vs. midi bend data

And now here's the data I have so far. I believe you will agree that there is a substantial "flat spot" on the WX5.

12Nov2000 Experiment

This was the first experiment, with my system set to collect bend data with my WX5 set to a nominal playing position. This was just a "first look" at how the reed bender performs.

graph of 12Nov experiment

25Nov2000 Experiment

This was the second experiment, with my system set to collect bend data with my WX5 set to a nominal playing position (repeating data from above), an increased lip gain, and an increased offset, and the reed lever only. Of specific interest here is that the "flat spot" exists even without the reed in place, disproving my original theory that the lay of the reed on the mouthpiece was the cause of the flat spot. Since it shows up in the lever-only data, it's presumed to be programmed into the WX5's look-up tables.

graph of 25Nov experiments

27Nov2000 Experiment

This was the third experiment, with my system set to collect reed bend data with my WX7 set to a nominal playing position and comparing that to the nominal reed bend data from the 25Nov2000 experiment. Of note here is the WX5 has a nominal "flat spot" that is 0.007" wide, and the WX7 has a nominal "flat spot" that is 0.012" wide. This is curious, because I find the WX5's reed performance to be vastly different from the WX7, with my personal preference being for the WX7. This is most likely due to the slope differences and wider operating span of the WX7 mouthpiece vs that of the WX5. The WX7 has a gentler slope, so the transition into and out of the flat spot is less noticeable than it is on the WX5. See the graph below. This certainly accounts for my personal preference for the WX7 reed performance. Others prefer the WX5's setup (Hi Bob Norton!)

WX7 flat spot ~0.012" wide    WX7 working range ~0.120"
WX5 flat spot ~0.007" wide    WX5 working range ~0.045"
graph of 27Nov experiments

Here are two photos of the setup for collecting the data. test setup

test closeup The WX5 is on it's back in a simple wood jig I made to hold it still and support the mouthpiece. A dial indicator is supported on an Enco indicator base that's clamped to my bench. Near the green Enco base, there is a screw/spring-hinge mechanism that allows me to essentially raise and lower the dial indicator. Using 5-min epoxy (easy to break free when testing is done), I glued a wood dowel to a fixed portion of the dial indicator parallel to the indicator's plunger. The end of the wood dowel is gently rounded where it pushes on the WX5 reed. As I turn the screw on the base, the dowel is slowly lowered onto the reed. In parallel to this, the indicator is pushing against a fixed metal reference sitting on the same primary surface that's holding the WX5 and indicator base. So, as the wood dowel pushes on the WX5 reed, the indicator measures displacement. The WX5 is connected to my VL70m for power and midi connection. The acutal midi data is being recorded at the other side of the room where I have my studio computer running MidiOx and the midi monitor screen on my VL1m. I have a 25' midi cable connecting my bench to the studio computer. My 7-year-old lab assistant, Brendan, called out midi values as I changed the reed displacement in 0.001" increments.

One final comment: my sincere apologies for the non-metric units of measurement. My dial indicator is graduated in 0.001" increments, and I'm too lazy to make the conversion.