Re: Adding a Transducer to a piano's Soundboard
Well, I see by the first forum post in this thread that I've been shying away from this project for four years. I guess that I've been afraid to do something that would permanently harm my piano, as old and beat-up as it is. But my inspiration has been the Yamaha TransAcoustics, which are super-cool to play, and the Steingraeber hybrid electro-acoustic pianos (though I've never seen one in real life).
Anyway, with some home remodeling going on, I had occasion to move my piano out in to the middle of the room, thus making its soundboard accessible, as well as needing to remove the two monitor speakers and subwoofer (my auxiliary speakers just got placed on top so I can still use them). This gave me the excuse that I needed to start figuring out the best ways to attach transducers to the soundboard, to see if I can have a modeled piano sound like a real piano, and thus do away with the large monitor speakers that normally flank the sides of the piano (not present in the photos today).
I have several Dayton Audio emitters ( https://www.daytonaudio.com/topic/excitersbuyerguide ) that I purchased, so my first trial was very simple: I had one weatherproof transducer that has a screw-mount (HDN-8), so I just screwed it directly into one of the ribs on the bass side of the soundboard on the back of the piano, driving it with Dayton's SA-98E amplifier. The results are encouraging, but this is no Yamaha TransAcoustic so far: the sounds are piano-like, but a bit brash, especially in the treble, where the sustain rings forever, possibly from the heavy transducer attached to the rib. And there's an echoic quality that speakers don't have. I had to go to the Grotrian Player to find a soft enough piano to not give acrid mid and treble tones, but the examples that I recorded were with the Steinway Model B (which is acoustically the closest to what my piano sounds like itself).
Here are photos in a Google Album (captions have been added to the "info" field):
https://photos.app.goo.gl/tJWaVQwidwcm9i3AA
And here are two videos, one showing the comparison of the acoustic versus the transducer-driven mids and bass, and the other showing the treble (the iPhone videos do not demonstrate the acoustic differences as well as I would have liked):
Bass and mid: https://photos.app.goo.gl/qZUXbp3EEpTp9rxh6
Treble: https://photos.app.goo.gl/FDmsvvri6nNv1sJ57
These are some other findings and thoughts:
i) The sound comes from behind the piano. This may seem obvious, since I'm driving the soundboard from the back, but I didn't realize that having the acoustic action enabled (stop bar disengaged) produces a sound that is much more 'forward' than the current transducer position. I somehow thought that most piano sound emits centrally from the soundboard, no matter how it is driven.
ii) The sound isn't as broad when driven by a single transducer as compared to being driven by a hammer-struck string. For some reason, even though the whole soundboard is resonating, the sound appears more localized. Another thing that should have been obvious, but somehow I didn't expect in real life.
iii) The initial percussion of the hammer hitting the string isn't the same. While we are driving the soundboard with a transducer, just like in an acoustic piano, we are not really driving the strings, except by resonance off the soundboard. So it's like a piano in reverse. It's not as correct-sounding an emulation as I would have hoped.
I am looking for experience and guidance from anyone who can help. Here are some topics of concern:
1) What's the best way to attach the transducers? My first attempt used a screw-peg, screwed into one of the soundboard's ribs. Yamaha attaches the transducer to the piano's frame, and then has the moving coil drive the soundboard through a separate lightweight point actuator. I don't know how Steingraeber deals with this issue, but their photos suggest that the tansducers are mounted to the soundboard. I have in my photos an experiment with hot melt glue with a transducer that I mounted to a test board several years ago and then removed today with rubbing alcohol - with little damage to the wood after removal.
2) Should the transducers be mounted to a rib or to the soundboard itself? I don't know how piano soundboards work, but there must be some sound transmitted through the string to bridge and string to agraffe connections, as well as sound driven through vibrations in the air from the string to the soundboard.
3) Where should the soundboard be driven? Dayton Audio recommends a 2/5-3/5 relationship on linear panel dimensions. I don't know how this corresponds to traditional piano string and soundboard setup.
4) What transducers would you recommend? I have several from Dayton, plus some beefy wall-mount transducers from Hidden Audio Systems (Model 401, made to drive drywall sheets making up a wall of a home, for example: https://hiddenaudiosystems.com/ ).
5) What about the amplifier? Will I hear much different sound from a more powerful amp than the small Dayton Audio amp?
6) Related to the amplifier question, I seem to get different clarity/less harsh tones by varying the relationship of Pianoteq Volume to Amplifier setting, more than I would predict would make much difference (since I'm not driving this to levels of audio distortion).
Anyway, that's a lot for one forum post. My next experiments will likely be gluing several different sized and shaped Dayton Audio emitters to different parts of the soundboard, and then driving each one independently.
I am interested in any ideas that you have for this project.
Addendum - I made one other change tonight: SInce I got annoyed that the Volume was not freezing appropriately between presets when I clicked the check-mark in the Freeze section for Volume, I allowed Pianoteq to play at 100% volume, and reduced the amplifier to about 30%. Surprisingly, this reduced some of the bloom (despite the overall output sound volume being the same), but it didn't reduce the bloom enough. I guess that this suggests that a better quality amplifier may help, though not cure the problems.
Thanks,
David