steve200 wrote:

Philippe,

If you don't mind, I thought it best I mention that the problem I am highlighting is most significant at ff.  At that velocity, Steinway D Bb1 and K2 C2 are dominated by the overtone.  It is almost hard to hear the primary. 

best,
steve

You are right, this happens too on a real piano, those longitudinal modes get often relatively stronger in the ff-fff range.  Note that we will reduce them slightly but won't kill them, they do give character to the piano sound when playing fff: you do want that something happens when playing fff, not just getting louder.

I would vote to leave them as is. For ff-fff notes in that register on a real piano I expect drama to happen. The effect will surely be different in low volume on computer speakers or headphone, but that should not be the standard.

On a more principled note, if you go for realism, go for realism, not for glossy and air-brushed. Or should we expect a new slider, going from ‘realistic’ to ‘dreamy’.

Maybe, but the detractors say pianoteq is too "synthetic".

I could argue it either way, i.e.
a) for removal of all the nasty artifacts of wooden pianos wherever possible.
The things that piano designers, builders and technicians would remove if only thy could.

b) Include the "warts and all" of pianos built in traditional ways from traditional materials.
The things that pianoteq detractors usually refer to as "character", incidental mechanical noises are already in.

Compromise is to make them switchable, control to attenuate them down to zero.
OK, so that is in Pro, but I feel it should be in Standard.

dagife & steve200, I did a quick & dirty "fix" for you guys; changed the string length and strike point slightly, just for that Bb note, to tame the twang. (To do it "properly" by editing the harmonic profile would be a lot more tricky.) See what you think of the resulting preset, which you can download here:

https://drive.google.com/open?id=1BZdj-...tS8aXNCtbA

This will work in Standard but not in Stage, which does not allow note-by-note edited presets.

It's not that easy. It's hard enough to take it out. It's even harder to put it back in. Unless it was explicitly shown as a reshaped harmonic profile, which it wouldn't be if it was part of the basic instrument model.

In my case I am using a real piano keyboard with a QRS MIDI strip attached.  I should be able to teach it that it's really hard to get to fff, but even on my real piano when the stop bar is left out, there's just only so loud it will play even if I really whack it.

I feel a "Spinal Tapp" moment coming on...:

"When you're really giving it your all, and you just want that 'extra push'...Where do you go?  127....That's why our amplifiers go to 127!"

Just kidding (sort-of!).

Philippe,

If you don't mind, I thought it best I mention that the problem I am highlighting is most significant at ff.  At that velocity, Steinway D Bb1 and K2 C2 are dominated by the overtone.  It is almost hard to hear the primary. 

best,
steve

You are right, this happens too on a real piano, those longitudinal modes get often relatively stronger in the ff-fff range.  Note that we will reduce them slightly but won't kill them, they do give character to the piano sound when playing fff: you do want that something happens when playing fff, not just getting louder.

As a side note, I recommend that you calibrate your keyboard in such a way that you really need to put a lot of energy to reach the ff-fff range, as on a real acoustic piano. On many digital keyboards, it is much too easy to reach the velocity 127, which corresponds to a huge smashing on an acoustic piano.

If you don't mind, I thought it best I mention that the problem I am highlighting is most significant at ff.  At that velocity, Steinway D Bb1 and K2 C2 are dominated by the overtone.  It is almost hard to hear the primary. 

best,
steve

A while back it was pointed out to me that the twanginess that cycles on and off during a long-held note is the audible manifestation of a vibrating string that might have started oscillating with an up-down motion, then precessing to a side-side motion, and then back again.  As one who's plucked a more than a few taut rubber bands in my childhood and seen this happen (and heard it), it makes sense to me a a potential cause for some of this sound 'artifact' (and I hear it on some notes on my real piano).  Philippe, does this hold in Pianoteq?

This can happen too, but it is not what was pointed here. Here the things you are hearing are the longitudinal modes, which can be very loud in some cases (see details in my first post). Controlling those modes for the piano manufacturer is a difficult task, particularly regarding the cupper wound strings. For example, their pitch from note to note can be quite irregular: although it goes usually increasing, it can happen that it decreases from one note the next one a semitone higher.

Thanks very much for your consideration of this issue.  I understand there are many constraints you are trying to satisfy!   Upcoming changes appreciated!

steve

If you allow me to make a comment: those twangy notes are not imposed by the physical model that we use, they are there just because the pianos that served as references had them, and sometimes even much more pronounced (If you listen for example to some Horowitz recordings, you will notice some enormous buzzy/twangy notes, not all notes though). When it is too pronounced (the meaning of that being of course a pure matter of taste), we attenuate them. The most difficult is to find the sweet spot that makes as much as possible everyone happy, which is what we try to achieve during the beta tests. Some people like twangy notes, some people want it more calm...

I say that because this problem is even more pronounced in Pianoteq's K2 piano, in note C2.  Have a listen.

My understanding is that K2 is a pure model, not based on any real piano.  There is no corresponding real piano that has that bad note.

I can't begin to speak for Philippe.  He has vast knowledge about this, and I have little.    But, my take on Philippe's response is that Pianoteq has a model they believe captures a piano.  The model makes notes with those pronounced overtones.  They consider that to be a "pure" and "correct" model, and so that is how it sounds, by default.

But in the end, as a consumer, this is simple:  these notes sound bad, and I don't think it is either fair or good marketing to expect me to upgrade from $150 Stage to $500 Pro to correct it.

I'll add that in blues piano, the root bass note, often Bb1 or C2, is prominently played, so these excess overtones are particularly noticeable in that context.

steve

Now S&S either provided/allowed a less than average instrument - to keep a software version safely below the sound of their instruments*, but quite close, or the average D does sound like this. I have played Ds in the past, but cannot hold in my memory that level of detail.  If the average D does have this characteristic ex factory, then there must be a way of prepping them for professional concert use. If pianists on the forum are troubled by it, and I assume they are not all first tier concert pianists, then 1st tier concert pianists would also be troubled by it and S&S technicians would fix it.

I personally hear the note in question to be slightly harsher than those nearby, but with similar characteristics and am quite happy to play an instrument that approaches the sound of a D but which may not have been expertly prepped to suit a tier 1 concert pianist. (If that is indeed the case) I can tweak the sound somewhat in any case and do what all pianists do apart from those touring the world's concert venues: adjust my touch for notes that don't quite mesh with their neighbours.

* I worked in the food industry: the quality of all manfacturer's products were kept slightly above the 'private label' versions they also produced for supermarket own label products.

I would suggest that Pianoteq attenuate these by default in Stage and Standard, and let the Pro guys increase it if they want.  In a real piano, a tech would try to eliminate this.

Steve

I agree. Indeed, Pianoteq is revered for its realistic piano modeling but not to sound patronizing but I don't think any piano VST provider would want to advertise being so realistic that a note or two will have a displeasing ring to it just like how real pianos could have. Don't get me wrong though, I am very much enjoying Pianoteq and I look forward to buying other instrument packs but my friendly criticism still stands.

Obviously I can only speak for myself but I'd argue that the majority of Pianoteq users would prefer to not have this piercing sound present in the Steinway instruments. Just like how I'd imagine most wouldn't want to hear it in a real instrument either. Here's hoping that Pianoteq developers can "fix" this sort-of problem without needing Pro.

Greg