Topic by peitzmann

  • Avatar
  • peitzmann
  • Member
  • Offline
  • From: Utrecht, The Netherlands
  • Registered: 23-08-2007
  • Posts: 10

Topic: design question - piano size and soundboard

Hi,

I should first say that I do find the achievements by Pianoteq truely amazing! I have a design question (mainly to the developpers, I guess, but may be somebody else is in the know...).

There are settings for the soundboard and piano size, which I would think should be coupled in a real piano design. They are not here, and may be that leaves more freedom, but I'm not sure whether the "normal" design behaviour is actually implemented. What I mean is the following:

A larger size piano will have both longer strings and larger soundboard. Longer strings means larger inharmonicity which is tuned by the "piano size" setting. It will also change the sustain properties of the string. This can apparently be controlled by the soundboard impedance, which may be fine to model the same thing. But the "soundboard" parameters have no size parameter. With size the resonances of the soundboard change in a complex pattern - e.g. if the contact point (the bridge) for a certain string happens to be at the location of a node of the appropriate resonance closest in frequency there would be very little energy transfer to the soundboard from that string. (These are actually potential imperfections of a real piano, so it may not be wanted to simulate these, but I'm curious how this is done.) The dominant effect from changing soundboard size however would be decreasing the frequency of the lowest resonance. Below that resonance, which is usually higher than the fundamentals of the lowest strings in a grand piano, almost no sound is radiated from the soundboard. This is often called the "missing fundamentals". I would have thought that this is one important detail of piano design. Is this implemented in Pianoteq?

Also for different sizes one might want to vary the location of the bridge relative to the edge of the soundboard - is that been done?

Best regards, Thomas

Reply by Philippe Guillaume

  • Avatar
  • Philippe Guillaume
  • Admin
  • Offline
  • Registered: 25-04-2007
  • Posts: 5,243

Re: design question - piano size and soundboard

Hi Thomas,
thank you for the kind words
Regarding the settings that you mention, you are right, they are coupled on a real piano, but the coupling is quite 'loose' and depends very much on the manufacturer choices, so letting impedance and size uncoupled gives the maximum of freedom to the user. It is like having first decided how long your piano is (the longer it is, the smaller the inharmonicity), then you still can decide how thick your soundboard will be.

Regarding nodes, note that the bridge is very rigid, so a given string does not 'communicate' only with the soundboard point which is just below, but with a broad part of it, reducing the risk of poor energy transfer to the soundboard.

Regarding the critical frequency below which sound radiation is inefficient (it is btw the frequency for which the soundboard transverse waves speed is less that the sound speed in the air), increasing the rigidity of the soundboard increases the speed of the transverse waves, thus reduces the critical frequency. So you are right, this couples impedance (depending on size, thickness, ribs, wood properties, etc…) and critical frequency. But here again we made the choice to let freedom to the user. The ‘philosophy’ behind being that in the virtual world, we are not obliged to impose the real life constraints, we can give us more freedom

Philippe Guillaume's Website

Reply by peitzmann

  • Avatar
  • peitzmann
  • Member
  • Offline
  • From: Utrecht, The Netherlands
  • Registered: 23-08-2007
  • Posts: 10

Re: design question - piano size and soundboard

Hi Guillaume,

you're right of course (the longer the string is, the smaller the inharmonicity) I got that the wrong way 'round in my post.

In general your reply tells me (which I could have guessed) that there are a lot more hidden parameters than are accessible to the user. Could some more of those be made accessible? (Like string material, e.g. solid vs. wound strings) What about an editor program for the more advanced features of piano design?

Regards, Thomas

Reply by Philippe Guillaume

  • Avatar
  • Philippe Guillaume
  • Admin
  • Offline
  • Registered: 25-04-2007
  • Posts: 5,243

Re: design question - piano size and soundboard

I confirm that there are a lot of hidden parameters (many thousands, although I never counted them), and we had to choose how many would be available to the user. We wanted something sufficiently rich but still easily understandable. One thing however that we would like to develop in a future Pianoteq Pro version would be to offer control of the parameters for each note. For example, you could adjust the unison width for each note, making only a few ones out of tune...

Philippe Guillaume's Website

Reply by hyper.real

  • Avatar
  • hyper.real
  • Member
  • Offline
  • Registered: 30-05-2007
  • Posts: 110

Re: design question - piano size and soundboard

there are a lot of hidden parameters (many thousands...)

<gulp>

But just to talk about the soundboard some more -  is it being also modelled as an independent radiative element with its own (low level) acoustic output? If so, that would possibly allow you to offer users control over the amplitude and timbre of its contribution to the sound. I'm not suggesting crazy soundboard materials, but I note that some forum posts have mentioned "woodiness" (or perhaps lack of it) in the sound. Perhaps there are some useful parameters here which could be bundled into a range of preset "materials", that users could select?

The other question I have which intrigues me in this vein is about the piano body as an acoustic cavity resonator. Pianoteq has the option of lid up or down, but I am wondering if this is treated as a reverberation characteristic after generating the "basic" (anechoic?) sound, or does it also form a part of the basic sound generation? (I leave out the follow-up question(s) about whether or not this is coupled to the soundboard size and impedance, etc).

Reply by doug

  • Avatar
  • doug
  • Member
  • Offline
  • Registered: 08-05-2007
  • Posts: 245

Re: design question - piano size and soundboard

hyper.real wrote:

The other question I have which intrigues me in this vein is about the piano body as an acoustic cavity resonator. Pianoteq has the option of lid up or down, but I am wondering if this is treated as a reverberation characteristic after generating the "basic" (anechoic?) sound, or does it also form a part of the basic sound generation? (I leave out the follow-up question(s) about whether or not this is coupled to the soundboard size and impedance, etc).

I wonder, too, about where in the chain the cavity resonance is modelled. The lid position setting is in the reverb section of the interface, which makes me suspect that if I turn off reverb I lose any internal resonance/reverberation as well as the room reverb. Is this how it works? I ask this because I prefer to run with no reverb or with an outboard reverb applied to all my instruments.


EDIT: used my ears and checked with the program. Lid position matters even when reverb is turned off. Good.

Reply by hyper.real

  • Avatar
  • hyper.real
  • Member
  • Offline
  • Registered: 30-05-2007
  • Posts: 110

Re: design question - piano size and soundboard

It's nice to know I'm not alone in puzzling about these somewhat mind-numbing questions. Perhaps one day when the model is published in a mathematical journal we might be wiser. Or we might not ;-)

I would expect the lid position to affect the sound in two ways. First, the tonal balance, with the lid down tending to reduce higher frequencies output. Second, as affecting the parameters of the cavity resonance. The former might be susceptible to modelling as a straightforward difference of EQ in the frequency domain, but the latter would I think would require modelling the change of Q as a temporal characteristic.  But I have the feeling that Pianoteq already models change in pitch due to overstriking a string, although whether the algorithms are identical or comparable I would not know.

The reverb topic has been occupying me also. I have found so far that I prefer the result if I apply a Pianoteq EQ to my speakers that gives a "flat" response at my listening position (seated at the keyboard about 1.5m from the speakers at ear level). Then I find the sound is more even and does not have any "lumpiness" up and down the keyboard. But (subjectively) "flat" in an actual room depends on the relative amounts of direct and reflected sound at the listening position. With Pianoteq having optional reverb, it is not clear to me yet how to set things up for an optimum impression of a real piano where one does not exist - mostly I just want to play the thing :-)

Doug, I would be interested to know what reverb parameters you prefer in your outboard, and why (acoustic context). My prior experience with software reverbs has been ugly- I dislike the distortion artefacts - and until I can understand how this operates when playing Pianoteq I am not in the mood to buy an expensive reverb like Altiverb. My preference is mostly for the studio-player preset modification, tho' I sometimes I feel this is too dry.

I think that I would be happy (in a (the?) future Professional version) to trade the option of different room acoustic treatments in favour of a more detailed modelling of the piano's mechanical/acoustical resonances, if it were a choice between the two. A room reverb can always be hung onto the output as I see it, but convolving in a piano cavity resonance in the same manner may not at all amount to the same thing; although I could be entirely mistaken about this.

Reply by doug

  • Avatar
  • doug
  • Member
  • Offline
  • Registered: 08-05-2007
  • Posts: 245

Re: design question - piano size and soundboard

hyper.real wrote:

Doug, I would be interested to know what reverb parameters you prefer in your outboard, and why (acoustic context). My prior experience with software reverbs has been ugly- I dislike the distortion artefacts - and until I can understand how this operates when playing Pianoteq I am not in the mood to buy an expensive reverb like Altiverb. My preference is mostly for the studio-player preset modification, tho' I sometimes I feel this is too dry.

At the moment, I don't have any outboard reverb but I'll probably use whatever comes with Logic Express, which is my next software buy. The main reason for a separate reverb is to run all my instruments through it so they sound like they're in the same room.

Meantime, I turn ptq's reverb off since I find that having it on--even in the player position--makes things sound mushier than I prefer. There's so much natural resonance in ptq that adding reverb (in addition to my room's natural reverb) is too much of a good thing.

Reply by damstraversaz

  • Avatar
  • damstraversaz
  • Member
  • Offline
  • From: lyon
  • Registered: 16-08-2007
  • Posts: 8

Re: design question - piano size and soundboard

doug wrote:

Meantime, I turn ptq's reverb off since I find that having it on--even in the player position--makes things sound mushier than I prefer. There's so much natural resonance in ptq that adding reverb (in addition to my room's natural reverb) is too much of a good thing.

I'm actually trying pianoteq and have the same impression. whereas this reverb sound very nice, I prefer to play without . The feeling of playing a real piano is better in my opinion. with the reverb It seems to me that I'm playing far away from the piano.

Reply by Philippe Guillaume

  • Avatar
  • Philippe Guillaume
  • Admin
  • Offline
  • Registered: 25-04-2007
  • Posts: 5,243

Re: design question - piano size and soundboard

But just to talk about the soundboard some more -  is it being also modelled as an independent radiative element with its own (low level) acoustic output?

yes, and you can control its amplitude through the "global resonance" slider (which actually combines soundboard, cabinet and harp resonances).

Pianoteq has the option of lid up or down, but I am wondering if this is treated as a reverberation characteristic after generating the "basic" (anechoic?) sound, or does it also form a part of the basic sound generation?

As Doug noticed it further, the lid model is independent from the room reverberation, but it is of course coupled to the previous resonances models.

I would expect the lid position to affect the sound in two ways. First, the tonal balance, with the lid down tending to reduce higher frequencies output. Second, as affecting the parameters of the cavity resonance.

This is exactly what we tried to do.

whereas this reverb sound very nice, I prefer to play without . The feeling of playing a real piano is better in my opinion. with the reverb It seems to me that I'm playing far away from the piano.

I agree with you: I often turn off the reverb when I play and want to have the feeling that the piano is just there under my fingers. And I use reverb when I want to listen to a midi file and have the feeling that I am just sitting in the room listening to somebody else playing.

Philippe Guillaume's Website