Re: Pianoteq 4 foretaste

Talking about the attach... I think that the most pianoteq pianos have a "divided" attach where it is a "th" sound that kind of is separate from the more body sound. Here I think that the Pleyel is more "whole" in the attach, but maybe with even less "bloom"? So, I'm playing around with 2 pianoteq plugs in Mainstage where I have one pleyel and one C3 with hammer hardness to a minimum or lowpass filter so I don't get any attach sound... (almost)... But like many instruments... One day it sounds fantastic but the next day... hm... and the third day it's ok and so on... So, how much is in our head? ;-)

Re: Pianoteq 4 foretaste

safari70 wrote:

It would be great if Modartt could add velocity settings (tweaked and tested) for various DPs to properly get that connection between keys and sound (similar to Nord's tweaks on their Nord Piano)... I know we can edit the velocity curve in PQ3 but after few attempts (using the demo) i just couldn't get a natural connection between keyboard and PQ3...

Check here for your velocity curves: Keyboard velocity curves

Re: Pianoteq 4 foretaste

About the "bloom," or the sound just after the attack. I'm thinking of the term as denoting a perceived rise in amplitude--a brief bubble--just after the attack. Does the note just take a few milliseconds to reach the peak before the first decay? In other words, we hear the hammer strike and all of the partials that are created, but the partials reach their highest  amplitude very slightly later?

Trying to think as a physicist or engineer, which I'm not, this seems impossible. The maximum energy should be at the attack. But if the "bloom" instead appears a very few milliseconds later, how could that be? Just thinking aloud, here:

1. The early reflections from the cabinet and nearby walls take those few milliseconds to reach us, and their combined amplitudes cause the delayed peak?
2. The strings are prevented from vibrating very briefly by the hammers--the hammer contact duration is longer than might be expected?
3. With the pedal not down, the impedance of the dampers takes a few milliseconds to be overcome?
4. The impedance of the bridge very briefly delays the transmission of the energy to the soundboard?
5. The coupling of the strings takes a bit of time, and the energy is most efficiently transferred to the soundboard after this coupling?
6. A combination of these things?
7. None of them or something else?

I wonder about the partial structure of this peak, too. If it is in truth slightly delayed, does that mean that what we hear most prominently is the note when it is already started to decay, even though we perceive it as sounding its loudest? In other words, we hear it as loudest after the attack stages of all the amplitude envelopes of the partials. Harder strikes would be closer to the Fourier structure, since the sound would reach us sooner and the hammer contact would be briefer?

Or am I making assumptions? What else could create the "bloom," assuming that it is an actual phenomenon?

Last edited by Jake Johnson (19-01-2012 08:28)

Re: Pianoteq 4 foretaste

Ecaroh wrote:

@safari70
If it is in preset, adjusting velocity won't help. Perhaps you should try hammer hardness instead for example.

BTW what keyboard did you try to use with PTQ?

I think you're right, it's probably not velocity response but rather the "reaction" of the sound or probably "characteristic" of the sound in reaction to the touch (not sure if i'm making sense here)...

I'm currently using a Roland FP-7F with touch set at Medium and offset +7...
The FP-7F's sound is decent (specially after some tweaks) but what's really great is the connection between touch/keys and sound, i.e. playability... So even if the sound is not super-realistic like sample-based, i really enjoy playing it...

I would like to get something similar with PTQ - it's kinda ok now (definitely better than most sample-based Kontakt instruments i tried so far), but there's still something not quite right in the overall playability... Instead of fiddling with the velocity curve i'll try few sound adjustments as you suggested maybe it'll help...

Thanks.

Re: Pianoteq 4 foretaste

Thanks Doug for the link... there seems to be an FP-7F post in there so i'll give it a try.

doug wrote:
safari70 wrote:

It would be great if Modartt could add velocity settings (tweaked and tested) for various DPs to properly get that connection between keys and sound (similar to Nord's tweaks on their Nord Piano)... I know we can edit the velocity curve in PQ3 but after few attempts (using the demo) i just couldn't get a natural connection between keyboard and PQ3...

Check here for your velocity curves: Keyboard velocity curves

Re: Pianoteq 4 foretaste

Jake Johnson wrote:

About the "bloom," or the sound just after the attack. I'm thinking of the term as denoting a perceived rise in amplitude--a brief bubble--just after the attack. Does the note just take a few milliseconds to reach the peak before the first decay? In other words, we hear the hammer strike and all of the partials that are created, but the partials reach their highest  amplitude very slightly later?

Trying to think as a physicist or engineer, which I'm not, this seems impossible. The maximum energy should be at the attack. But if the "bloom" instead appears a very few milliseconds later, how could that be? Just thinking aloud, here:

1. The early reflections from the cabinet and nearby walls take those few milliseconds to reach us, and their combined amplitudes cause the delayed peak?
2. The strings are prevented from vibrating very briefly by the hammers--the hammer contact duration is longer than might be expected?
3. With the pedal not down, the impedance of the dampers takes a few milliseconds to be overcome?
4. The impedance of the bridge very briefly delays the transmission of the energy to the soundboard?
5. The coupling of the strings takes a bit of time, and the energy is most efficiently transferred to the soundboard after this coupling?
6. A combination of these things?
7. None of them or something else?

I wonder about the partial structure of this peak, too. If it is in truth slightly delayed, does that mean that what we hear most prominently is the note when it is already started to decay, even though we perceive it as sounding its loudest? In other words, we hear it as loudest after the attack stages of all the amplitude envelopes of the partials. Harder strikes would be closer to the Fourier structure, since the sound would reach us sooner and the hammer contact would be briefer?

Or am I making assumptions? What else could create the "bloom," assuming that it is an actual phenomenon?


Hello Jake,

I was one of the people who mentioned the term 'bloom' in this thread; here is how I describe it:

As a piano tuner who regularly works on real acoustic pianos, I am keenly aware of the way the sound (envelope of natural harmonics) physically changes, when the vibrational mode of the string changes from the purely vertical plane (in a grand piano) to one that incorporates the horizontal plane as well.  This real effect is intentionally created by the bridge pins, which literally stagger the path of the string as the string passes over the bridge.

When one views the orientation of a soundboard's spruce grain, you will notice that the grains run deliberately parallel to much of the bridge.  This is part of a real piano's soundboard design, because it is a known fact that soundwaves travel with approximately five times (5X) the efficiency in the direction parallel to the grains, than in the perpendicular direction across the grains.

Now this may seem contrary to common sense, but the loudest noise comes from the attack of hammer upon the string, followed by a natural decay of sound (when the string's vibrational vector is primarily in the vertical direction), but is AGAIN followed by a slight perceived increase in sound energy (at least in the form of hearing various new, different harmonics emerge) after the string's vibration has turned horizontally, and the increased soundboard efficiency occurs due to the sound being distributed along the direction of the soundboard's grain.

As an aside, every piano manufacturer saves its best spruce (with the most densely packed wood grains) for soundboards of its more expensive models in the product line.  This is not by accident!  The best pianos may be anticipated to exhibit 14-20+ parallel wood grains per inch in their soundboards.  In contrast, the cheapest models of a given piano manufacturer exhibit comparatively "wide" grain spacings ... sometimes only ~8 parallel grains per inch.

I hope this comparatively unknown feature of real acoustic grand pianos might shed some light on the subject of "bloom" in a piano's sound characteristics.


Cheers,


Joe

Re: Pianoteq 4 foretaste

Hi Jake,

To add to Joe's very informed comments on real instruments "bloom", I would like to say, after listening closely to Ptq notes versus the V-piano's same (test mp3s of the V-piano are available at the pianoworld forum in their ongoing DPBSD project useful for that purpose) that the same pattern of changing harmonics can be heard on each one, meaning that string and soundboard modelling gives similar results for both, but, to my ear, Ptq's bloom is masqued by the attack that is not as short, loud and percussive as it should be in my opinion, and not related enough to the velocity of the attack. It sounds more static and detached.

I would like to point to this very interesting current (2010) publication on piano modelling (refering to Modartt and Roland by the way) that defines in section IV a proper physics approch to hammer modelling. It is clearly written also that both hammer and string models have an influence on each other (which can lead to numerical instabilities) but that a real-time complete simulation could be implemented on a current machine (too bad the source is not available, I would have loved to experiment with it...)

I am just extracting here a small conclusion but, even if you are not an engineer familiar with differential equations, filter design and the Z-transform and such (which I am not really either) reading the full article can give you an idea of some of the difficulties of instrument modelling.

Here it is: http://home.mit.bme.hu/~bank/publist/taslp10.pdf

In my opinion, Pianoteq's best instrument to date (v3.x) is the Pleyel, because hammers are very prominent and an interesting inharmonic content in the instrument gives more life to the attack. I suppose, in the interest of fidelity to the source instrument (especially the historical ones) that the characteristics of the hammer attack are somehow extracted and added to the string and soundboard model giving the final result, but that the string/soundboard model itself is excited seperately. I'm only guessing here, but that would explain why Pleyel's hammers sound so different from, say, C3's or K1's. This would be (if I'm correct) a very general approach allowing modelling a variety of similar instruments, at the cost for the piano of not having a more realistic hammer/string interaction. In fact, this is already a hybrid approach in that sense.

Of course, modelling the hammers as in the paper as a generic mass with non-linear spring would give a standardized sound to the attack, but, in my opinion, expressivity would be greatly enhanced.

Maybe I'm completely wrong here and the strings in Ptq are excited in the same way as in that paper, but if not, this would surely be my personal demand for the future...

Re: Pianoteq 4 foretaste

For a few years now people have been comparing Pianoteq with Truepianos, and it seems that most everyone is firmly in one camp or the other, very polarized, just like Mac vs PC.

The only thing that TP seems to have that to my mind is better is the attack, otherwise the sound in some registers just doesn't seem right to me, it doesn't breathe. But IMO that type of attack, perhaps being able to dial in the amount, would be outstanding in Pianoteq.

Listen to the Christian Teuscher demo on this page to see what I mean: http://www.truepianos.com/demos.php

Michael

Re: Pianoteq 4 foretaste

Michael H wrote:

The only thing that TP seems to have that to my mind is better is the attack, otherwise the sound in some registers just doesn't seem right to me, it doesn't breathe.

I agree.  Before I settled on PT3, I auditioned some sample libraries and tried the TruePianos demo (I forget what version it was at).  I absolutely hated it, I find it soulless.  I watched the Christian Teuscher TP2.0 tech preview video, and while the player's performance is stunning, I hear the same 'deadness' in the sound (good performances can blind you to the true character of the instrument, I find I need to listen a minute or so before I can start to pick out the character of the sound itself).

PT does 'breathe', the string ringouts and interaction really 'sings', that's what sold me on it instantly.  I also tend to tweak longer decays & more resonance to really let the strings dance, something I loved about the few real acoustics I've played - I think PT is totally unmatched here.  The only missing part is more character.

Re: Pianoteq 4 foretaste

Hi everybody.

I leave you my own point of view, Pianoteq is just another new instrument we have to learn to play with.

Ok, it's name starts with 5 letters everybody knows. But what kind of piano factor do we have here ?
The first need is LOVE to build such a virtual instrument, with a "real" beautiful sound.
As Jacques et Jean Pierre Favino are in love with ebony, cedar and all theses woods used for guitars, as Pleyel, Erard, Steinway, for piano, etc. They were all of them in love with their instruments.
I am sure that Philippe is in love with abstract woods like mathematics, FFT, differentials, interpolation, topology... and piano too. I am in love playing music. But what is fascinating in this Pianoteq, is that you can change the strike point, put dampers on every string, put a 7.84m string length... and you can drive theses parameters on the fly, in real time and/or in midi files, in low or high resolution.
In fact, we just have to learn how to use it, how to compose for it, to search new shapes to make a beautiful sound and music with it, to combine spectrum profile to get other sounds...

A new goal, try to reach kind of new perfection, may be. Probably. Surely.

Re: Pianoteq 4 foretaste

Jake:

I've been sick for a few days, and just read your long post.  I know what you are saying, and I suggest that Joe has written a post that explains things very well.  And I agree with him completely.

Some additional comments directly addressing some of your points:

2)  I don't think that the hammer/sting contact time is significant - the hammer bounces away from the string (and the string bounces away from the hammer).  'This time interval is not likely significant.

3)  I suspect (help me Joe) that the dampers have little significant effect on the string when the hammer strikes the string as when properly adjusted they come up immediately upon striking the key.

4)  There is impedance that theoretically result in a time lag, but I recall having brought this up a while back, and one of the developers suggested it was negligent.  I concur.

5)  Coupling of the strings and soundboard - perhaps Joe's comment about perceived energy change being the result of harmonics that change with time explains this.  Call it psycho-acoustic if you will, but when the pitch of a note changes, I suggest that we tend to perceive the change as a change in volume.

However, consider the analogy of two children on adjacent swings of different length, hanging on to each other while being pushed as a unit.  While each has a natural frequency  that is different from the other, because they are "coupled", they must swing at the same frequency until they let go of each other.  Strings on the other hand are coupled with the bridge/soundboard.

Just some late night thoughts - maybe I'm not thinking quite straight.

Glenn

Last edited by Glenn NK (22-01-2012 11:21)
__________________________
Procrastination Week has been postponed.  Again.

Re: Pianoteq 4 foretaste

Jake Johnson wrote:

About the "bloom," or the sound just after the attack. [...]
What else could create the "bloom," assuming that it is an actual phenomenon?

Regarding blooming, reverberation can also play an important role, particularly when used with some delay. Reverb delay can go as high as 0.1 sec (it's a high value, 0.05 is more typical), and in such cases a blooming appears. It is not rare in sample libraries that the piano was recorded with some reverb, so that even when using the samples dry, there's already a reverb blooming which would not be there in the real world if there was no reverb. Such a built-in reverb is not necessarily perceived as reverb, because the damping of the sound at key release occurs after reverb, whereas the natural path is that reverb occurs after note damping (the two effects do not commute).

Re: Pianoteq 4 foretaste

Glenn, Joe, and Philippe, and all:

Thanks for the responses. Interesting to learn about how the early reflections reach the ear before damping begins. But then the partial structure changes in the ongoing reverberations as the dampened strings reverberations reach the ear--an ever evolving sound, with the reflected sound always slightly brighter\containing louder upper partials?

Glenn and Joe: Interesting conjecture about the change in partial structure being perceived as an amplitude peak. But I'm not sure that I understand the idea. Some partials are attenuated more abruptly, leaving others more exposed, and thus perceived as briefly louder?

I'm still reading in the literature in a haphazard, Google-search sort of way. One  factor that does seem to have an effect on the sound--the hammer contact time--may or not contribute to the bloom (I was just tossing out ideas), but seems to be a consideration in the partial structure and possibly the initial sound, according to some of the literature. I can't pretend to follow the calculations in the second article, but the discussion of the results is interesting:


Page 116-118 of Benade's Fundamentals on Google books: (I don't know if Bernade is still valued as a source--I don't find him referenced in the other literature very often.)

http://books.google.com/books?id=cCW5Ng...mp;f=false


Starting at the bottom of page three in this article by Avanzini and Rocchesso:

http://recherche.ircam.fr/equipes/repmu...es/P46.pdf

I'm not sure that I understand the meaning of their term "spectral centroid." The set of partials that contribute most to the sound at a given moment? I couldn't find an exact definition in the article. Is this a well-known term?

EDIT: Just did a google search for "spectral centroid" and found 20,000 references. Appears to be a well known term...

Last edited by Jake Johnson (25-01-2012 02:36)

Re: Pianoteq 4 foretaste

Philippe Guillaume wrote:

The audio examples from the Hermode Tuning site are not issued from the Pianoteq version 4 in preparation. The only version 4 audio examples currently available are those two linked at the top of this thread.

One might have different opinions about the Hermode Tuning system, but I do hope that Pianoteq 4 will support Logic's global tuning settings. It is so convenient to set the tuning once for all instruments in a project.

Last edited by tutman (28-01-2012 12:21)

Re: Pianoteq 4 foretaste

tutman wrote:

I do hope that Pianoteq 4 will support Logic's global tuning settings. It is so convenient to set the tuning once for all instruments in a project.

Unfortunately the Logic tuning settings seems to be only available to Logic internal instruments, and not to third-party AU plugins such as pianoteq

Re: Pianoteq 4 foretaste

Jake Johnson wrote:

Glenn and Joe: Interesting conjecture about the change in partial structure being perceived as an amplitude peak. But I'm not sure that I understand the idea. Some partials are attenuated more abruptly, leaving others more exposed, and thus perceived as briefly louder?

Hello Jake,

The simple answer to your question is, 'Yes."  However, I am never one to be content with furnishing short and simple answers.

Think of the vibrating string as containing a large number of partials in the harmonic series, especially in the first few seconds after the hammer has just impacted the string.  Initially, the string is vibrating in a direction perpendicular to the soundboard, with most or all of its harmonics being excited, because the hammer's impact caused it to vibrate in this direction.  So far so good.

The nature of the bridge is to transmit sound to the soundboard, simply because the bridge is glued to the soundboard.  The other aspect of the bridge is that it has so-called bridge pins (two per string), designed to jog the string through a little detour, if you will.  This slight change in longitudinal direction, as the string crosses over the bridge, intentionally forces the string to begin also vibrating in a direction (mode) which is parallel to the soundboard. 

If you were to visualize the combination of the string's vibrations, surely some of it will be in the vertical direction and some of it in the parallel direction to the soundboard in a form imagined as a "plus sign" in arithmetic!  Actually this is not quite the case:  Much of the time, the string's vibrational pattern (viewed from the end of the string) is a combination of vertical and horizontal component at the same time!  High-speed movies of vibrating piano strings have captured the string's "path of vibration", and found the sum of the string's various vibrating "directions" (vertical, horizontal and diagonal) string to trace out a circle or a slightly elliptical nature, rather than simply vertical OR parallel.

Since (mostly) spruce soundboards are designed to carry sound along their grains, it is natural to imagine that different partial harmonic frequencies will be resonant ... when that component of the string's cross sectional vibration is parallel to the soundboard, rather than perpendicular to it.

EDIT:  Now here's the important part -- despite the overall vibrational energy of the string to be decaying over time, the change in vibrational mode to include parallel to the soundboard can actually "focus" different resonances (at the expense of others that have been constrained) whose volumes can actually seem to emerge from the overall sound.

Hopefully this helps more than it confuses.


Cheers,

Joe

Last edited by jcfelice88keys (29-01-2012 08:07)

Re: Pianoteq 4 foretaste

Ecaroh wrote:

@safari70
Latency. If your sound system has latency, it won't feel natural no matter how much you adjust velocity curve. After adapting my playing with my Nord Piano, I feel latency everywhere...

Every system and everything has latency, even an acoustic ~1ms/34cm so if your ears are 60cm away from an instrument you got 2ms latency

Last edited by Rytmenpinne (28-01-2012 22:20)
http://sharpattack.bandcamp.com/ my very own one man band project

Re: Pianoteq 4 foretaste

I just tried a small "thought experiment" (call me nuts if you wish):

Imagine something is generating a series of short tones at constant pitch and at constant sound pressure level (SPL).  They are repetitive and go on for five or ten seconds.

Suddenly without warning, the sound pulses change in pitch by one semi-tone, but the SPL remains the same - same actual loudness.

Do you think the sudden change would stand out in your mind?

Is it possible that it could be perceived as a change in volume?

I leave this for the readers; in particular Jake and Joe.

Glenn

__________________________
Procrastination Week has been postponed.  Again.

Re: Pianoteq 4 foretaste

Glenn NK wrote:

I just tried a small "thought experiment" (call me nuts if you wish):


Suddenly without warning, the sound pulses change in pitch by one semi-tone, but the SPL remains the same - same actual loudness.

Do you think the sudden change would stand out in your mind?

I leave this for the readers; in particular Jake and Joe.

Glenn

Hello Glenn,

Thank you for posing your thought experiment.  In particular, I would like to address the question of whether a sudden change in pitch by one semitone would stand out in one's mind.  My answer is an overwhelming 'Yes', even if the total sound pressure level (SPL) remains the same.

Here is my reasoning for this reply: 
If you have a series of tones that suddenly repeat one semitone higher, one has just perceived the theme written by John Williams of the movie JAWS!.

F-F#.................F-F#................F-F#...F-F#...F-F#-F-F#-F-F# ... etc.

I would contend that, even if the change in a semitone (let's be reasonable and assume the change in tone is perceived in the pitch range of the Grand Staff, say, in the range of C2 to C6, I believe that everyone in this forum would perceive the tone change as standing out in one's mind.

* * * * * * *

Now, when I tune the unison pitches of three middle C4 strings in an ordinary acoustic piano, by slightly varying the unison tuning, I can often "dial in" a predominant octave overtone, or a predominant twelfth overtone, and if I am fairly far out of unison tuning (but not enough to perceive much of a warble in fundamental tones) it is possible to hear a simultaneous pair of predominant overtones, namely a twelfth paired with a seventeenth.  In terms of notes on the keyboard, I can dial in C4 with relatively predominant overtones of : C5,  or C4 with G5, or C4 combined with G5 and E6.  Of course, I strive for C4 with only a predominant C5 overtone.

* * * * * * *

Proceeding to address your question whether some overtone partials become actually louder, I must answer subjectively (as I do not own an SPL meter) as 'Yes, sometimes'.  As a case in point, when I normally tune the fundamental pitches of notes in the range where two copper-wound strings are found -- usually G2 to E3, many comparatively mediocre pianos (uprights and grands) exhibit painfully loud seventh harmonics -- even when one of the two strings has been silenced in order to tune the nonblocked string.  I believe what is happening in this case, is that the acoustic energy gets "focused" at the seventh harmonic.  An analogy is to focus the sun's rays (on a warm summer day) on a dry leaf with a magnifying lens.  The same amount of light falls upon the leaf, but the sun's rays are focused into a small area -- and the dry leaf begins to smoke and eventually burn, despite the total amount of photons entering and leaving the glass lens has not changed -- only redirected.

An example might be C3 (an octave below middle C) and its seventh harmonic sounding as Bb5.  Sometimes, in order to supress these annoyingly nasty overtones, I must resort to retuning these individual notes -- and then tune the rest of the piano one or two hertz above or below a diapasion of A440, so as to keep the customer from complaining about "those horrible overtones" arising from these copper-wound two-stringed notes!

Cheers,

Joe

Last edited by jcfelice88keys (29-01-2012 08:18)

Re: Pianoteq 4 foretaste

Rytmenpinne wrote:
Ecaroh wrote:

@safari70
Latency. If your sound system has latency, it won't feel natural no matter how much you adjust velocity curve. After adapting my playing with my Nord Piano, I feel latency everywhere...

Every system and everything has latency, even an acoustic ~1ms/34cm so if your ears are 60cm away from an instrument you got 2ms latency

Warning - this has nothing to do with my previous post - except we're doing some more thought experiments.

Case One:

Sound travels at about 340 m/s (1115 fps).  Suppose you are the pianist in an orchestra sitting at the extreme left of the stage.  The brass (trombones) are at the the extreme right about 10 metres (33 feet) away (both as viewed from the audience).

33 feet divided by 1115 = 0.0295 which is just under 30 milliseconds.  There is a lot of latency between these two players, yet they manage to play in time.  How do they manage?

Case Two:

Play a bass note on a grand piano very softly.  From the time you touch the key, until the string and sound board are vibrating is several milliseconds.  What's interesting is that the latency in the high treble is less (the hammer doesn't have to travel as far).   I have no trouble playing a grand piano - well not true, but the latency is the least of my worries.

Case Three (to demonstrate how small a millisecond is):

If you have a watch with a stop watch feature built in (mine does), try starting from zero, and then attempt to stop precisely five seconds later.  Can you repeatedly stop the timer within 1/100th second (10 ms)?  I can't get it right very often, and I've used the feature many times to determine the beat rate of a recording.

How problematic are latencies anyway?

Glenn

EDIT:  changed 1/10 second to 1/100th second.

Last edited by Glenn NK (29-01-2012 20:10)
__________________________
Procrastination Week has been postponed.  Again.

Re: Pianoteq 4 foretaste

Case One:

33 feet divided by 1115 = 0.0295 which is just under 30 milliseconds.  There is a lot of latency between these two players, yet they manage to play in time.  How do they manage?

Glenn


The answer is that you watch the conductor's baton, having faith that the speed of light (7-1/2 circumferences of the earth per one second) supersedes the speed of sound.  The two individual players may sound out of timing with respect to one another, but the audience perceives the timing as synchronized.

Cheers,

Joe

Last edited by jcfelice88keys (29-01-2012 05:48)

Re: Pianoteq 4 foretaste

Glenn NK wrote:

Case Three (to demonstrate how small a millisecond is):

If you have a watch with a stop watch feature built in (mine does), try starting from zero, and then attempt to stop precisely five seconds later.  Can you repeatedly stop the timer within 1/10th second (10 ms)?  I can't get it right very often, and I've used the feature many times to determine the beat rate of a recording.

Glenn


Actually 1/10th of a second is 100 millisecconds ==>  100/1000 = 1/10

I have read elsewhere that two short pulses sound as one, until their spacing exceeds about 25 milliseconds.  Surely, by 30 milliseconds, a person can hear two distinct sounds.  You can demonstrate this for yourself by experimenting with a hardware or software reverb unit -- analog or digital -- if you set the initial reflections to last longer than 30 ms, you will hear two distinct sound "sources" in your reverb.

EDIT:  If a virtual instrument is set to allow 512 samples, then 512 samples being run through a clock of 44,100 Hz means one should expect a latency of 9.3 milliseconds unless there is an additional latency delay from your computer's CPU, hard drive or general electronics. End EDIT.

Cheers,

Joe

Last edited by jcfelice88keys (29-01-2012 08:31)

Re: Pianoteq 4 foretaste

Well, it's true that every sound system has latency. That's why, for example, it's important to put your speakers close enough and to same distance from you. All this seems probably very obvious to us, but it's funny that music dealers seem to forget it: they may try to sell the most expensive pianos with very bad monitor positioning.

I think all of us have our own acceptable tolerance for latency. It's simply matter of what we are used to deal and play with. I said earlier that my Nord seems to be very fast. After I've adapted to it, almost all the other digital pianos seem to have more latency. Before Nord I used Rolands for maybe 15 years. I remember that when Roland started to use "supernatural" technology their latency increased. Perhaps because of the modeling?

So we should not underestimate the effect of (software or soundcard based) latency at playing feeling especially for new guys coming from hardware world. To myself, if I wanna have similar feeling with PTQ which I have with my Nord, I have to use very low audio buffer size. Fortunately my mac based sound system can handle that most of the time. On the other hand, I have no problem with some latency, I've used softwares so many years...

Last edited by Ecaroh (29-01-2012 09:41)

Re: Pianoteq 4 foretaste

On the stage the bass player and his amp is often/allways right beside the drummer to get the best timing between them.
If not they maybe allso needed a conductor :-)

Re: Pianoteq 4 foretaste

Thanks Joe for correction.  That was sloppy of me.

The correction makes my point a bit stronger - perhaps?

Maybe my example should have used one of the old "swing" bands - Benny Goodman say.  When he played the clarinet, I wonder who was swinging the baton?  And those bands filled up quite a large stage - and they seemed to be working together.

This reminds me of the arguments going on in photography forums where "noise" is argued incessantly - yet it rarely shows up in a print - but is quite obvious with pixel peeping.

Maybe it comes down to what is significant, not what is measurable.

Glenn

PS - I'm going to edit the other post.

Last edited by Glenn NK (29-01-2012 20:08)
__________________________
Procrastination Week has been postponed.  Again.

Re: Pianoteq 4 foretaste

jcfelice88keys wrote:

I have read elsewhere that two short pulses sound as one, until their spacing exceeds about 25 milliseconds.

I think you are out by a factor of 10 or so.

The approximate number has been known since the early days of 'talking pictures'. When the first electronic amplifiers arrived, they had to be used with horn speakers, which were the only type efficient enough to fill an auditorium from 10W or 20W. When a film of a tap dancer was shown, the audience heard an echo on every step. It was quickly realized what was happening - the loudspeakers were multi-way systems, and had been installed with their mouths aligned. When the individual horns were realigned so that their drive units were within a metre or so of each other in the vertical plane, the effect disappeared. The effect was rediscovered by drummers when electronic percussion kits arrived - some drummers could not adapt to them. The reason: the speakers might be 3 or 4 metres away from them.--

Re: Pianoteq 4 foretaste

Four metres divided by 340 m/s = 0.0118 second = almost 12 milliseconds.

I'm coming to the conclusion that there is no one "right" answer as to how much latency is acceptable.

Acceptable latency varies from person to person (don't most things?) and whatever works for the individual is what is acceptable or "right".

IOW, if it bothers you, it's too much.

Glenn

__________________________
Procrastination Week has been postponed.  Again.

Re: Pianoteq 4 foretaste

hyper.real wrote:
jcfelice88keys wrote:

I have read elsewhere that two short pulses sound as one, until their spacing exceeds about 25 milliseconds.

I think you are out by a factor of 10 or so.


Hello Hyper.real,

Upon performing a Google search for JND (just noticeable difference), there is a Stanford University research paper that quotes between 30 and 40 milliseconds as the time between two +20dB pulses is required before they become discerned as two individual entities.  Below this duration, they remain perceived as one.  This threshold of time is reported as the "rise time" of so-called ossicles in the ear become activated -- I stand by my original assertion, but thanks for your input.

Cheers,

Joe

Last edited by jcfelice88keys (30-01-2012 04:03)

Re: Pianoteq 4 foretaste

julien wrote:

Unfortunately the Logic tuning settings seems to be only available to Logic internal instruments, and not to third-party AU plugins such as pianoteq

OK, thanks for settling that. Too bad, though. But it made me fill out a Logic feature request at Apple's site. Maybe others are interested in this too.

Re: Pianoteq 4 foretaste

I'm curious...

What will be of the actual pianomodels after update to Pianoteq 4 ???

Will the old models work in a improoved way, like if pianoteq 4 have new algorithms for soundboard etc...  or the old piano models will not get extra quality ?

Re: Pianoteq 4 foretaste

Good Day. Long-time-never-post here.

First, Jake "About posting recordings of those desired notes..." gave me my best laugh of this morning.

And next, listening to Gilles' two performance-posts on Page 2 gave me the idea the desired "attack" (over Pianoteq's) many are on about consists of a little pyramid of emphasis to begin each note. And every note, monotonous to me - I like PTQ's Pleyel lots better, once I get over its slight boxiness. However, Even a dummy (me) can see this pyramid'd cut through the mix of sound onstage from a band, and that would be exactly what the piano player will want.

So maybe they'll be interested in what the Manual describes at Item 4.5, Direct Sound Duration. Haven't tried it, only read. It doesn't seem to beef up the level, only control attack's duration, I presume including shortening PTQ's available pyramid, whereby it may offer a bit more, er, attack, for free.

ADDED: Make that "perceived attack". And if it indeed helps with cut-thru, in view of the PTQ-engineers' elsewhere having spoken of their interface's "exposing" a limited quota of their modeling, yell at them about exposing MORE of it, re this issue.

MORE: Mulling the issues here over (and making no claim to ever have tried these, dead unadventurous), I see that 4.5's brief presentation mentions Hammer Hardness in connection. It may thus chance that this alone would beef up the perceived attack, and all the needed control is "exposed" already. But can't say by own self-hands-on. Have a look.

Heh. My first piano I DID buy because of its bright tone. After having forked over the cash and got the thing, I shortly found that B natural delivered a BONK instead of a note. Severely mismatched and dead middle of where you play. Got a book on tuning, fast, and soon found out the bright tone was due to glue in the hammers, plus that NO amount of pricking the felts fixed the BONK standout. I never did trace the cause, either, despite sawing out bits of soundboard, dropping a felt curtain between hammers and strings, learning to tune (focus B natural). Nada Nada Nada. Last I saw of the box it was upending into landfill. The action went decades later.

Still got the wrest pins, and to end this tale of woe where it began, yes I sympathise with bright tone fanciers, even if I tend to think "glue".

Last edited by custral (07-02-2012 08:31)

Re: Pianoteq 4 foretaste

Something just occurred to me while listening to some recorded piano: On a real piano, is there a slight delay in the sympathetic resonance? (Does it take a very few milliseconds for the vibrations to travel down the bridge and excite the other strings? A matter of bridge impedance?)

Last edited by Jake Johnson (13-02-2012 08:16)

Re: Pianoteq 4 foretaste

The attack has become a feature of this posting. Could we have another mp3 sample of the attack in Pianoteq 4?

Re: Pianoteq 4 foretaste

Jake Johnson wrote:

Something just occurred to me while listening to some recorded piano: On a real piano, is there a slight delay in the sympathetic resonance? (Does it take a very few milliseconds for the vibrations to travel down the bridge and excite the other strings? A matter of bridge impedance?)

Hello Jake,

The quick answers to your three posed questions are ... Yes.  Partially Yes.  Partially Yes.

I would completely agree there is a slight delay in the onset of sympathetic resonance in the strings.  However, that actual sympathetic resonance (in the form of undamped strings being excited into sympathetic vibration) is not 100% conducted through the bridge.  Rather, it is mostly caused by vibrations in the air! 

Try this experiment, yourself, if you have access to any grand piano, or an upright piano with the front of the case removed such that the strings are exposed:

1) Depress the sustain pedal, in order to free all of the strings to be able to vibrate (if excited), and hold it down with your foot during this entire experiment;

2) Place your mouth somewhat close to the strings, say within about 12" / 30cm or so;

3) Yell any word you wish to say, such as "Hey!!!" -- or any expletive of your choosing -- into the undamped strings.

The sound and pitch of your voice will be sympathetically captured by the strings!  If you want proof, simply release the sustain pedal, and the strings will cease vibrating sympathetically in response to your voice.

Even the longest strings' upper harmonics will vibrate in sympathy with your voice.

Now the question is:  Did your voice excite the strings via vibrations in the air (hint:  Yes), or were the strings excited mostly by the vibration and impedance qualities of the bridge (hint:  probably not). 

In my opinion, the sympathetic vibrations of strings are limited in time only by the speed of sound in air.


Hopefully this helps you answer you curiosity.


Cheers,

Joe

Re: Pianoteq 4 foretaste

jcfelice88keys wrote:
Jake Johnson wrote:

Something just occurred to me while listening to some recorded piano: On a real piano, is there a slight delay in the sympathetic resonance? (Does it take a very few milliseconds for the vibrations to travel down the bridge and excite the other strings? A matter of bridge impedance?)

Hello Jake,

The quick answers to your three posed questions are ... Yes.  Partially Yes.  Partially Yes.

I would completely agree there is a slight delay in the onset of sympathetic resonance in the strings.  However, that actual sympathetic resonance (in the form of undamped strings being excited into sympathetic vibration) is not 100% conducted through the bridge.  Rather, it is mostly caused by vibrations in the air! 

Try this experiment, yourself, if you have access to any grand piano, or an upright piano with the front of the case removed such that the strings are exposed:

1) Depress the sustain pedal, in order to free all of the strings to be able to vibrate (if excited), and hold it down with your foot during this entire experiment;

2) Place your mouth somewhat close to the strings, say within about 12" / 30cm or so;

3) Yell any word you wish to say, such as "Hey!!!" -- or any expletive of your choosing -- into the undamped strings.

The sound and pitch of your voice will be sympathetically captured by the strings!  If you want proof, simply release the sustain pedal, and the strings will cease vibrating sympathetically in response to your voice.

Even the longest strings' upper harmonics will vibrate in sympathy with your voice.

Now the question is:  Did your voice excite the strings via vibrations in the air (hint:  Yes), or were the strings excited mostly by the vibration and impedance qualities of the bridge (hint:  probably not). 

In my opinion, the sympathetic vibrations of strings are limited in time only by the speed of sound in air.


Hopefully this helps you answer you curiosity.


Cheers,

Joe

Yes, I forgot the "sing into the piano" experiment. I wonder if the transmission along the bridge does much. Seems as though it might--if the vibrations in the air can overcome the pressure of the damper, surely the vibrations along the bridge could, too?

Re: Pianoteq 4 foretaste

In reply to Jake and Joe:

This really gets interesting; I never thought about the explanation by Joe of SR until now, but I have no doubt that one's voice will do this; I recall it happening on my Yamaha G2, so I can confirm this to be true.

But if the bridge/soundboard is vibrating (note I said IS not ARE, as they are virtually one and the same for physics purposes - generally glued together quite well ), then would not a unison of three for example be exciting the bridge/soundboard, and hence the other some two hundred strings?

As an aside, this raises the interesting case of the upright piano that is up against a wall (where they are typically placed).  At what distance should they be placed so that the sound waves coming off the soundboard, striking the wall and returning, do not cancel/amplify the newer waves of sound from the soundboard?   This was discussed  in a past forum by Modartt et al in regard to the lid of a grand piano, in regard to all the reflections that occur within the case before the sound escapes to our ears.  Complex - exceedingly complex.

On Jake's question about the delay of SR, I would suspect that theoretically there is a delay, but we must be talking about very small time intervals.

Recalling the comments by Joe about repetitive sounds that sound like one - I think the sound delay in this situation is extremely small or small enough to not be detectable by the human ear/brain.  But this is just my conclusion from thinking about it, not any result resulting from calculations or measurements.

On the matter of impedance, every physical body has mass and stiffness which results in impedance (resistance to being moved, deformed, etc), - so there has to be a time delay.  How large is the delay?  I don't know, but again I suspect it's really quite small.  Is it significant?  I'm not sure, but many things occur/exist in this world that are not significant.  The trick is determining what is significant and what is not.

Glenn

__________________________
Procrastination Week has been postponed.  Again.

Re: Pianoteq 4 foretaste

jcfelice88keys wrote:

Now the question is:  Did your voice excite the strings via vibrations in the air (hint:  Yes), or were the strings excited mostly by the vibration and impedance qualities of the bridge (hint:  probably not). 

In my opinion, the sympathetic vibrations of strings are limited in time only by the speed of sound in air.

Joe, have you shouted at any harps recently? :-)

Sympathetic vibration requires energy input into the system - in this case from shouting. The coupling between energy in the air and the piano strings is poor compared to the coupling with other parts of the system. The initial dominant sound is from acoustics of the cabinet as a resonant chamber excited by specific frequencies in the shouting. (Get your wife to shout into the piano :-). This is primarily what distinguishes a piano from a harp in the first instance. Once the shout is inside the cabinet, more of it impacts the soundboard than the strings. The coupling of the soundboard to the the strings is very good compared to the coupling of the cabinet to the string, and the energy in the soundboard transfers to the strings more easily than from the air to the strings. This begs the question why the ear picks up the resultant vibration of the strings in preference to the resultant vibration of the soundboard. 

The sympathetic vibrations are limited in time by the mechanical resistance (also called damping) in the entire system - not by the speed of sound in air. Speed of sound is different depending on the material, but a characteristic of the propagation of energy is not a cause. Time limits on the amount of power in the system are dependent on the rate of energy loss from the system. One loss is by the mechanical connection of parts with other parts (where impedances matter). The other loss is due to internal dissipation within the material from internal mechanical resistance (also called damping). A piano's sympathetic vibrations are curtailed by dropping the dampers. Some characteristics of felt can be demonstrated by shouting at it :-)

To pick up the dangling question, the wood of the soundboard has higher internal resistance(s) than the strings. This alters the physical character of resonances in its pattern of vibration - they are not as 'loud' and their active frequencies are 'smeared' over a larger range. Psychoacoustically, the perception of resonances depends on both of these physical factors, but not only these. Another important factor is that the multiple resonances in the soundboard are different from those of a string, and are not typically heard (though not by musical instrument makers) as being pitched in the way that a string under tension is.

So is this any more enlightening? Or, for that matter, relevant to the topic?

From the point of view of the physics, that is of simulating the acoustic response of a piano, there are two aspects: the transient behaviour (as the energy input initially travels around the mechanicals), and the steady state behaviour (as the flows net off).

From the subjective side, comments in this thread suggest to me that (some) users feel the attack portion of the sound can be improved. I also note general comments about the timbre of the sound from Pianoteq in different reviews and forums. These tend to be of two types, but they could be different ways of making the same observation. One type of criticism I read is that the sound has a metallic edge. Another is that the sound lacks the woodiness of actual pianos. I think there is an element of is-the-glass-half-full-or-half-empty variety to these. Nevertheless, such criticisms persist - even after accounting for the fact that a demo version of Pianoteq is available which allows users to mitigate in various ways characteristics of the sound which they don't like.

My general impression of Pianoteq 3, FWIW then, is that the product does well modelling strings, but has potential for improvement in respect of modelling the other components of a piano. This is perhaps not unexpected given the relatively complexities of the physics of strings and of planks of wood.

Nevertheless, I like what I hear of Pianoteq 4 work-in-progress. I may find when it emerges that I've been barking up the wrong tree here :-)

Re: Pianoteq 4 foretaste

I agree that the coupling between the strings and airborne sound is very weak; the coupling between the soundboard and airborne sound is strong (otherwise the piano would have to be amplified - and please don't call the sound board an amplifier - it's a transducer).

The most striking demonstration of this (that can be checked at a guitar store), is to strum a solid body guitar with the power turned off.  Very faint sound.  Then strum an acoustic guitar with the same energy input.  Good ones will fill an auditorium with sound without amplification.

I've discussed this with an old friend that plays and builds classical guitars.  He of course knows all about this.

Glenn

__________________________
Procrastination Week has been postponed.  Again.

Re: Pianoteq 4 foretaste

hyper.real wrote:

I like what I hear of Pianoteq 4 work-in-progress

Like what you hear ?

What do you hear ?    Just tidbits at the beginning of this thread ?

Last edited by dondascher (15-02-2012 21:50)

Re: Pianoteq 4 foretaste

Modartt will give us another bit of sounds for V4, or should I kdnap a beta tester ?

:-)

Re: Pianoteq 4 foretaste

Déclaration d"âme mûre"

Permettez moi:
d'après ce que j'ai lu ici et là, la plupart d'entre vous semblent être des professionnels de la musique, des techniciens du sons,etc... Enfin des gens apparemment très qualifiés concernant la musique, et particulièrement le piano.
La plupart du temps, vous parlez de vitesse du son, de caisse de résonnance, de l'importance de la dureté du marteau sur la longueur des cordes, de la résonnance par rapport à ceci, par rapport à cela, etc...
Je dois avouer que je ne comprends pas un traitre mot de tout ce charabia technique. Pour ma part, parler du piano c'est parler musique, donc parler d'Amour.

Depuis toujours; mon rêve a toujours été de posséder un piano. Pas un vulgaire piano, mais un piano qui deviendrait "mon" piano... et il serait unique, et majestueux, et sa sonorité serait si particulière parce qu'il mesurerait trois metres de long, et je pourrais l'adapter à ma (si particulière) personnalité, et ce piano ne ressemblerait à aucun autre, et puis surtout: il serait à moi, seulement à moi...
Mais pour x raisons, surtout financières, ce rêve n'a jamais pû être exaucé, et j'avais fini par faire une croix dessus. C'est la vie, c'est comme ça...
Puis j'ai découvert qu'il existait des simulations de pianos, virtuels certes, mais là était peut être une réponse « acceptable» à mes attentes... Alors j'en ai tésté des douzaines, des vingtaines...

En aucun cas j’oserais affirmer que Pianoteq a réussi à creer la simulation de piano parfaite, je ne suis certainement pas suffisamment qualifié pour cela étant donné que je ne connais pas «l’incidence de telle ou telle partie» sur «telle ou telle autre partie» de l’instrument (personnellement: savoir qu’il serait mieux que le vecteur X soit plus rapide et plus approprié, ou que l’incidence Y  résonnerait bien mieux comme ci ou comme ça ... mais je constate que: outre les possibilités quasi infinies de «farfouiller» dans le logiciel pour réellement créer de nouvelles sonorités, il m’est possible de jouer du piano et surtout, point important pour moi, de régler celui ci avec une précision d’horloger suisse et à ma propre convenance...

Pour finir, je suis, personnellement très satisfait de ce logiciel qu’est Pianoteq 3 . Et il y a fort à parier que la prochaine version m’enchantera encore davantage...

J’estime que c’est bien plus qu’une simple simulation de piano qui; bien que prenant très peu de place dans mon ordinateur, occupe une place tant importante dans mon coeur... mais ce que j’en dis... dans le fond; je n’y connais rien en musique...

PS: râler parce que l’on aura à payer une somme ridicule pour une mise à jour ...
me parait franchement déplacé.

                                 *************************

Allow me:
From what I read here and there, most of you seem to be professional musicians, sound technicians, etc. ... Finally people seem very qualified for the music, especially piano.
Most times, you talk about speed of sound, sounding, the importance of the hardness of the hammer on the string length, the resonance from this, compared to that, etc. ...
I must admit I do not understand a word of all that technical mumbo jumbo.
For me, speak about the piano, is speaking about music, so: speaking of Love.
Historically, my dream has always been to own a piano. Not an ordinary piano, a piano that would become "my" piano ... and it would be unique and majestic, and his sound is so special because it would measure three meters long, and I could adapt it to my (very special) personality, and this piano would be unlike any other, and above all : it would be for me, only me ...
But for some reason, especially financial, that dream has never been heard, and I ended up doing a cross over. That's life, that's how ...
Then I discovered that there were simulations of piano, virtual certainly, but there may be a response was "acceptable" to my expectations ... So I tested dozens, scores ...

In any case I would dare say that Pianoteq has managed to create the perfect simulation of piano, I am certainly not qualified enough for that because I do not know "the impact of a particular party" on "this or that other side "of the instrument (personally: that it would be better than the vector X is faster and more appropriate, or that the incidence Y resonate much better like this or like that ... but I find that : besides the almost limitless possibilities of "rummage" in the software to actually create new sounds, I can play the piano and especially important for me to settle this one with a precision of Swiss watchmaking and my own convenience ...

Finally, I am personally very pleased with what this software Pianoteq 3. And it's a safe bet that the next version will enchant me even more ...

I think it is more than just a simulation of piano, although taking up very little space in my computer, occupies a place as important in my heart ... but what I say ... in the background, I do not know anything about music ...

PS: bitching because you have to pay a ridiculous sum for an update ...
seems to me truly outrageous.

Re: Pianoteq 4 foretaste

Would be cool to have another demo to make the wait more bareable.

Re: Pianoteq 4 foretaste

Rohade wrote:

Would be cool to have another demo to make the wait more bareable.

I second that.