Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum The right wire for late 19th century grands

That is obvious from what I said. At that point, the string is displaced the most, and therefore it is at its longest length. It corresponds exactly to the maximum transverse displacement.

What I would like to know is whether what I call longitudinal mode corresponds to what you call longitudinal mode, and how it differs from transverse mode. That is a simple enough question.



I am the one who is asking questions, which should be an indication that I am interested. Sorry if my questions bother you, but you brought the subject up, and you should be able to discuss it.

L modes is responsible for forwarding the attack noise and tone precursor.
If a wire does not very well generate Lmodes, it may be the reason of the "hole" that is perceived immediately after hammer impact, and that I thought was due to some internal damping.

Lessening the impact noise can be a good thing, probably not in all regions.

I will listen to the samples with and without L modes to see if this helps me.

The importance of the attack transients (?) is under evaluated by tuners and re builders, in my humble unauthorized opinion;)

Yes I believe this may be related, and not only because of the acoustical tones produced by L modes and their mix with usual modes.

WHy can't we find analyzed spectra samples, and comparaisons , even if not at a high precision laboratory level, any one owning a decent sound card and a mike can make samples, and analyse the shape of the tone (with some error margin indeed but are not tenedancies shows enough so to make temporary conclusions ?)

When I see a spectra I see the most prominent partials and their level

When I analyse the "envelope" (dont know if this is the good word. I see how the tone behave.

Even Audicity, which is free, allow top do so
.

BDB;
I am glad you are interested in wire type and L-mode behavior. You can study Ellis's book, articles published in The Journal of the Acoustic Society of America, Patents issued to Conklin, and to Ellis, and others. I don't think a forum is the place to post copyrighted material or to self-publish a text. I look forward to your comments on the body of writing that presently exists.

I would settle for a brief definition of longitudinal mode. I gave one, which would coincide pretty much with the other modes of vibration. Another definition, which may explain why Conklin says it is about 10 times the frequency of transverse vibration, would just be related to the speed of sound in the material. In any case, if it is 10 times the frequency of the transverse vibration, that pretty much eliminates it from being a factor in all but the bass strings, as it would be inaudibly high elsewhere.

if you want to hear it BDB jut rub a string, assuming you can get up of your chair you will be able to hear it, assuming yours ears are yet functional

Tried it. The results are inconclusive, and I have a number of problems with the method. After all, a violin is played by rubbing the strings, and yet most of the sound is due to the lateral motion of the string.

high pitched, similar as some pîtches produced by hammer matin g problems, but thicker, it should be easy to investigate if those are the same tones

Now when rubbing with a rosin impregnated finger we hear one pitch probably the most fundamental L mode.

Seem more than evident to me that thoses tones mix with the other form of vibration.

Audicity is basically recording software. It gives you a good time-series plot of the sound envelope but it does not tell us anything about frequency. We can only get volume (power) relative to time.

To get frequency we need a spectrum analyzer. And separating out energy generated by longitudinal modes relative to transverse modes gets tricky. The best way is to incorporate accelerometers that can isolate vibrations relative to the direction of motion but good ones are expensive. A decent tri-axial accelerometer can easily run better than $2,000. And then there are the electronics to drive it an process the signal(s).

I have a pair of fairly good single axis accelerometers that one day Real Soon Now I’m going to try to set up to measure some of this but I don’t know how successful I’ll be until I try it.

If I do make it work I’ll try to publish the results somewhere.

Until somebody does this we have to rely on human ears which are not all that bad a measuring and analysis device when coupled with some experience and thought. I await Ed’s articles with some interest.

ddf

BDB;
Read Ellis book and you will see how some of the L-modes and T-modes interact. You will have to wait for my Fully Tempered Duplex Scale article for me to disclose other mechanism's.

Hi thank you for chiming, you may relate to older versions of audacity, you can ploit the frequencies and see the peaks in audacity, based on a selection. DOyou mean a freqency analysis tool must be real time initially ?

But even on a PC there are more sound analysis oriented tools as WavePad http://www.nch.com.au/wavepad/fft.html
.
I had one very professional one, but not the hardware adequate to use it. but a simple ZOom H2 is yet enough to take strings samples and see what one hear

I have been working as a tuner at the IRCAM lab in Paris when they begin to use accelerometers, but it was on an old Petrof and the tone thickening was not very apparent on old strings.

Often I have seen the experiences did show interesting things but nothing conclusive came from unisons, and they where not analyzed as such.

When I will have time and contacts, I will try to setup something so we can try to see how the mix Tmode Lmode is evolving during unison tuning.

High speed camera should be excellent also to see how the phase works between strings, the only videos I have seen did not include sound samples, nor time frame, only polarisation (wapin) was seen. no idea of the way the unison was "build"

BTW I asked that on another thread, where it seem that lower tension could raise the sound volume when usiong the Excel spreadsheet based on the "calculating Technician" .
To me mor etension mean more volume, is not it ?
or do you know if volume is measured as a first partial value (in the Travis book) ?

Thanks

I.O.

My copy of Audacity was fairly old. I’ve just downloaded the most recent version and I see what you mean. That feature plots the spectrum of a wave envelope without reference to a specific time. To be useful in analyzing the sound produced by a piano it needs to be able to isolate the spectrum to a specific instant of time.

Since most of the information that enables us to identify and quantify a piano tone comes in that first burst of energy at impact and immediately following impact it is necessary to be able to isolate the energy spectrum and watch what happens during this time.




Yes, there are good PC tools available but they are not cheap. (Actually, there are several software fft programs available as shareware. They are pretty fussy and I’ve never been able to get them to do anything really useful but I’m not a computer guru. Someone else could probably do better.) Neither are the peripherals such as measurement mics (those contained in Zoom products are not good enough), mic pre-amps, etc.

The mic needs to be flat from at least 20 Hz to 20,000 Hz and not many are. The cost of a suitable measurement mic, at least, has come down considerably over the past few years. Twenty years ago nothing was available for less than $1,500 to $2,000. Today mics adequate for our purpose can be had for $200 to $300. They are not quite as good as those from the dedicated measurement instrument manufacturers but they are good enough.

ddf

Yes Del, but frankly to do the basic checking I wanted to do (just see how the most important partials behave during unison tuning) the software (free version) I gave the link is perfect

Good display and some editing functions, not much analysis tool indeed but that is yet positive.

The clearing and quiet behavior of the second partial is visible at the same time it is heard. (real time then)

Another point , less easy to see but may be related, is that the fundamental is not as stable as expected, and get better with hammer mating (I file a little the crown in the recorded samples)

It is fun to see how the top spectra get quiet, with a smooth extinction, as soon a sthe unison is tuned clear.

I will try similar recordings with a different way for unison, and see if I can get a more stable fundamental while the rest of the spectra is also good.

I agree that certainly to work in the regions of Lmodes and to analyse their interations with the sensors you tell of, expensive software and material is needed.

There is a very good software availeable for free only for one month , the name escapes me, but the phase is manageable, lot of add on availeable, I will post the name asap.

I've probably tried most of them by now. The one I currently use when I'm traveling is SpectraPLUS. With options. That, along with an Audix measurement mic and a good preamp/phantom power supply makes a fairly good portable system.

My two accelerometers are fairly small -- the cables take up most of the space -- but the constant-current power supply/preamp is USB powered and, for some reason doesn't communicate with Windows 8. It looks like I'll have to replace it. More money gone.

ddf

I haven't used Audacity in a while, but I believe one can limit the spectral plots to specific points in time. I think what you'd have to do is to take data over the whole time of interest, save it to a file, open it in Audacity, select the start and stop time, and then look at the spectral plot. Setting the start and stop time may discard the rest of the recording, but if you have a copy saved on your hard drive that shouldn't be a problem.

I think you're right. I just downloaded it and looked at it briefly but I don't know why what you're suggesting wouldn't work. A little cumbersome if you're interested in more than one time period but it is free. And, of course, it's missing my favorite waterfall plot.

ddf

Yes Spectra plus, seem to be one of the most adaptable ones , that was the name I was looking for

I hope you will be able to setup your equipment soon

You have a waterfall and a free version for the soft I gave the link, http://www.nch.com.au/wavepad/fft.html

sure realtime is more useful. you use the Quad capture Roland ?

I've been using my basic travel setup -- including SpectraPLUS on my laptop -- for years. It's not the best fft system available but it is relatively low cost (a little more than $1,000 with options) and, obviously, portable. I can carry everything I need in a small, canvas case in my computer bag through airport security with only a few raised eyebrows. (Through the X-ray machines the microphone looks like the barrel of a gun....) They sometimes don't like 15' of microphone cable either so I pack that in checked luggage if I'm taking any.

I've been using other equipment in my office/workshop at home but it is less portable and more complicated and I'm not sure the results are all that much better.

It was only recently that I lost the use of my accelerometers. But I don't usually travel with them. I purchased the accelerometers to work with the bridges and soundboard units on my string test frame. They work well for this but mounting them requires modifying the bridges and/or soundboards to mount them.

I’ve been considering adding a new one that would travel along with me but, again, these things are pricey. The two I have now ran about $350 each and the one I’m considering next is way more than that.

ddf

Oleg and all:

For analysis, do not forget Spear, the program created by Michael Klingbeil. Its graphics are not striking, but it lets you see partials and their amplitudes shift in time, displaying them as lines running across the screen. Since it allows you to then synthesize the note into sine waves and broken into moveable nodes, you can, within the program, silence a partial or several partials, change their relative amplitudes, or change their pitch at specific moments in time to hear the result: http://www.klingbeil.com/spear/

Note that for the best accuracy, one can calibrate the lowest amplitude it will register. As in all similar programs, the better the recording, and thus the less noise\hiss\hum, the better the result.

Hi ! I certainly wish to setup analysis with accelerometers, but I understand the difficulty to position them well, unless a 2 dimension goodie is used.

I seem to recall that they where glued to the bridge, and where fairely small, at the IRCAM.

Is not it possible to use simple mikes to detect phase shifts, I suppose at the soundboard/strings level only.

I want study of unisons, what happens when the tuner decide to change the stabilisation delay between fundamental and the rest of the spectra; do we use any of the L modes for that.

Difficult to have a setup that will play the notes "musically" and with similar energy each time but I believe it is possible just with mass and height, and may be some smooth material to simulate the amortizing done by the pianist.

I also wish to analyse those "self crunching" unisons I hear regularely, they are may be good for all percussive music but leave no much options to make the tone sing, as it saturates too much before getting quiet.

Greetings