Wishlist for Pianoteq V2.x (Page 3 of 4)

This is not so much a request as a question about the desirability of a feature. I'm writing this after reading about how azrael4 is using two instances of PianoTeq as separate velocity layers, much as though they were velocity layers in a sampling program like Kontakt or Giga. (See my note at the bottom of this post--I'm aware that this notion is problematical, considering the goal of modelling.)

Would it be good to have the ability to assign fxp's to midi velocity layers in Pianoteq and stack them? Even without the ability to stack them internally, the ability to assign an fxp to a velocity layer could be good.  We could then open several instances of PianoTeq, each assigned to a different velocity layer. We can of course already do this using Cantabile or a sequencer such as Nuendo, as azrael4 does. But would it be valuable to be able to do this directly in PianoTeq?

The load on the processor might prevent us from drifting into the world of creating 5-6 layers. Imagine how this line of thought might otherwise play out: people creating something like one of the newish 40 gig pianos with 20 layers.

All of which leads to a larger theoretical question. Would creating many, many layers be back-pedaling or creating the piano version of Frankenstein's monster? Or would it instead just be combining the possibilities of synthesis with what has been done in sampling. "And" instead of "or." Maybe. Yet it seems to go against the basic principle of Pianoteq--that the model renders the sound of a piano without having to use old-fashioned sampling workarounds like layers. And the load on the processor would be terrible. And a few of us would spend many waking hours experimenting with all of the possibilities. (Splits, multiple velocity layers, different presets loaded on each and edited into fxps, control over the sound sources for them all.)

Sorry if this notion is opening a can of headaches. If so, please blame azrael4 for leading me to bring it up.

1)
I would like to have some models without any room information. I can hear the room where the piano was originaly recorded in, before it is analized and (how ever) transfered into a model. (But maybe I´m wrong and hear something else?) It only turned  into a problem when I changed the overtone-sliders. Then the
"room" changed in a way that did no good onto the sound.

I don´t know if it´s possible to render out the room-info or to record it in a room completely soundproof.

2)
Would it be possible to have something like a "block diagram" or "structure" of the model to see where and how parameters are connected?

After all, pianoteq is innovative and inspiring!

I'd like to request a control that I think Beto-Music requested when the forum was first started, when he asked for control over "how quick the damper cut the sound of strings". (Beto-Music, you related this to the Attack.)

I think I'm thinking along the same lines. (I think.) The problem, for me, is that the dampers cut the sound too much and too fast. I notice this on fast runs, when lifting the pedal, and even more in the overall difference between the sound when the pedal is up and pedal is down. Often, I create an fxp with the pedal down so that I can hear the full decay of each note, and get exactly the sound I want, but when I then play a few notes with the pedal up, they sound muffled or sometimes synthy. I'm just posting these observations as subjective impressions. Conversely, if I create an fxp in which the pedal up notes decay at the rate I want,  the notes sustain too long with the pedal down.

I'm not sure what's happening in terms of the physics. I understand that the damper of course both restricts the vibration of the strings and absorbs some of the force from a strike, while also being bounced off the strings a little when enough force is applied. Is there more going on? Does the damper reduce the length of the string, so different transwaves are created, adding entirely new freqs?

In any case, I guess I'm asking for the ability to control both the amount of damping and the rate at which the damped freqs decay. I understand that in a good, well-maintained piano, the dampers kill the sound fast. (Maybe I'm asking for these controls as "age and condition of the piano" controls?)  (Are we asking for the same thing, Beto-Music?)

I'd still like to be able to switch off the help tags. They keep covering up information I need--I'm thinking especially about adjusting the volume and the help tag pops up right in front of the bar graph.

Simple request :

EQ (or high pass filter) for the "hammer noise"

Thanks
Guillaume Jodoin

I wish we could hold some few parameters, like velocity curves, while changing from one piano model to another.

   Sometimes we take a time adjusting the velocity curves, since each Brand of digital piano or controller have a particular feel of touch, and need to do that again while changing to another piano model on Pianoteq.

    I also desire a "firm base ability to velocity curves too. Like we careful adjust the velocity curve to ou controller (digital piano) and that be set as fix. And when we would dowload a FXP created by someone, the velocity curve of it would be summed up with the firm basevelocity curve.
      The firm base velocity curve would be just a compensator to get a linear sensibility from the controller to Pianoteq.

     In my digital piano I need to do a deep change in velocity curves to get bether dynamics from MF, F, FF to FFF.  So if I upload somebody elses FXP which had a own velocity curves, I will need to readjust it. And maybe the FXP vel Curves was created to make it like some vintage piano touch, and the creator of the FXP made the velocity curves for both, compensate his controller to Pianoteq, and to make it similar to a vintage piano feeling touch.

     With the "firm base ability" everyone would first "fix" his controller with Pianoteq, making a linear responde adjusting the velocity curves graphic, and all the posterior velocity curves adjusts would be summed up to the "firm base"
velocity curves adjust.
      So perhps it' bether call it as preliminar adjust or compensator adjust between midi-controller velocity response and Pianoteq velocity response.

The more I experiment, the more I'd like to have volume controls for:

The Remanent Sound and Direct Sound
The Global resonance
The Sympathetic resonance
The Full Sustain Resonance

The goal here would not be to raise any one volume a lot, but instead to be able to adjust each slightly in relation to each other.

A separate EQ for the Full Sustain Resonance would be good, too. Would give us a lot of control over shaping the sound.

These controls wouldn't necessarily go against the model: velocity and the other parameters would still control the relative volume\effect of each sound contributor\modifier.

According to the model, increasing one sound source, say the Direct Sound, would still automatically increase the volume of the sound reflected off the soundboard, etc. However, giving us control over the volume of each sound source, and ideally a control over the extent to which each increase in volume affected other volumes later in the sequence (a scaling tool like the velocity scaler, with each parameter represented by a line?) would let us shift the listening position to a degree and make slight adjustments to get different balances among the sound sources.

I second the "morphing" request uttered here recently. Sometimes I'd like a bit of the "twang" of Bechstein, but also the overall soft base of the C2. Don't know if this is feasible, but if so, it would be of great use I think.

Sorry Jake for my incomplete explanation… The "direct sound duration" slider affects the volume of direct sound with respect to remanent sound because it controls the direct sound duration. So why does the direct sound duration affects this relative volume? This is because of its definition. In each unison,  there are slow decaying modes and fast decaying modes. Let's suppose that there are only two such modes: a slow one and a fast one. Imagine that at the beginning of the note, the fast decaying mode is louder than the slow decaying mode. Then, at a certain time, both modes will be equally loud. This moment is by definition the direct sound duration: before this moment, the fast decaying mode is louder - this part is called the direct sound, and after, the slow decaying mode is louder - this part is called the remanent sound. Now what happens if you make the direct sound duration longer? That means that it takes a longer time for both modes to be equally loud, that is, the fast decaying mode was louder at the beginning than before. Thus the rule: increasing the direct sound duration is similar to increasing the direct sound volume with respect to remanent sound volume. Note that the slow and fast decay rates are not affected by this change. Hope this helps, and thank you again for sharing your passion about piano sound.

Hello,
I would suggest a Keyboard Level Scaling (a curve that will modulate the level by a specific amount, based on key pressed - a la DX) of the following parameters (with priority on my wish list !):
- Hammer Noise
I'd like to be able to have a curve like this
\_
   \

- HH Forte
- HH Mezzo
- HH Piano
I'd like to be able to have a curve like this
\_
   \
I'd notice to my taste that I could use less HH and HN in the top range (C5 and above) and more in the low range (C1 and below)

- Character
I'd like to be able to have a curve like this
\_

Maybe somebody had already requested it, sorry I haven't read all the thread.

Michael

Are you sure you have room in your submarine along with that tremendous cathedral pipe organ ?

With full MIDI controller implementation, anything is possible.  I think I'd make mine out of the H. R. Giger school of design.

My main wish is for keyboard scalability of all parameters with a graphic interface using, perhaps, vector curves that can be visually adjusted.  ...And maybe the ability to reach inside and strum the strings with the back of the nails for that harp effect.....

I second the request for (elaborate) keyboard scaling. Like Michael said here, it would be great if one could decide how much hammer hardness and hammer noise there is in different parts of the keyboard (for example, as also stated by him, soften things a bit in the upper range).

A simple one-dimensional "scale" parameter, while it would already be great, would not quite be enough though. The suggestion of the "DX 7 style" keyboard scaling would be more like it.

Or something even better: There is a relatively unknown FM synth plugin named "Octopus". There, they have implemented keyboard scaling in a rather sophisticated way.

http://www.linplug.com/Instruments/Octopus/octopus.htm

You can download a demo and try it. This thing has a very flexible keyboard scaling (in the graphical area at the left, 2nb button "KBD"), comparable to the velocity curve in PTQ. I'd love to see that in PTQ for various parameters.

Incorporating these ideas into a sleek GUI might be a challenge. Maybe one could have a "hidden layer" where such controls could be found. The Octopus interface, however, gives a good example of how much can actually be shown at a glance. Though it admittedly looks a bit "techy".

I believe scaling options would greatly help with fine-tuning sounds and adaption to various situations (in live playing, one would use different settings than in recording etc.).

A small and three not so small requests for some future version:

1. The ability to press just a number to call a preset\fxp. In the standalone version, you can call a preset by pressing Alt-1, Alt-2, etc. This request may seem trivial, but being able to just press a number, without the Alt key, would be nice: a slightly faster way to choose a preset. This small change would also let us use a usb or wireless numeric keypad set on top of the midi keyboard, and not have to worry about swiveling around to choose a preset. (There doesn't seem to be any ability to call a preset with the Alt-# key-press in the VST version--is the assumption that the host will let the user create shortcuts?)

2. Sliders, possibly adjacent to the current Spectrum profile sliders, that let us reduce or increase the decay rate (or sustain time?) of each partial. We could then set both the volume and decay rate of each partial in each note, and thus have enormous control over the evolving sound. Ideally, these decays would be modulated by velocity, and the sound sources and influences that come later in the sequence of producing the note, such as the full sustain resonance and pedal position, would register any changes made and adapt accordingly.

3. Decouple the body resonance and harp resonance, so we could choose to raise or lower one without affecting the other. This decoupling would be particularly good to have if the EQ range extended lower, to let us raise\lower the volume of still lower frequencies, or if there was a separate EQ for the body resonance that let us edit the amplitude of these lower frequencies. There might also be an option check-box to let us keep them coupled, so we could create settings that had a given ratio of harp to body resonance, recouple them, and then move the already existing Global Resonance slider so the volume of both was raised while retaining the new ratio.

4. Sometimes, on some presets, with the pedal down, for a reason I think I understand, the sound of the Global resonance seems to rise slightly--not in pitch, but instead to literally pan up as it decays. This is not what actually happens, of course. The real problem, I think, is that in the midrange and upper treble, particularly, the sound of the Global resonance seems to originate from a location that is a little too high: It doesn't seem to come from slightly below me, inside the piano, but instead from slightly above my head, and in fact, from a place that is above the sound source for the main note, if not the lid. Thus, as the main note decays and the Global resonance is more exposed, the sound seems to rise. This situation may occur because the original intention of the model was to achieve a more distant sound, and thus the listener hears the global resonance as it is reflected off the lid. Or am I the only one who hears this? Time to start drinking just water while I play?

Hi jake, good suggestion about the default midi controllers. We will keep that in mind for a future release

It occurred to me that these sliders might in a sense go against the model. But I still wanted them... Cameleon does look like fun, but I was more interested in the changes in piano timbre that we might be able to find.

Do the sliders go against the model? We can already control the amplitude of the first 8 partials, and create a piano that's several yards long. So the model is flexible--each variable contributes to the harmonic content and the length of the sustain\decay or adds its own sound. The sliders would just be another set of variables, and variables that are "realistic" since strings at different ages, and at different listener distances from the piano, are going to produce standing waves that take varying times to decay. I think. But yes, this is over my head.   

Or do you mean that the sliders would create a mathematical mess and overload the cpu, requiring too much processing note by note? In any case, no one else has mentioned these sliders, and that's probably not a good sign.

I guess you are talking symptoms of CPU overload. At least, that's what I can relate to. I have never encountered any sonic overload / distortion in PTQ.

The solution here would not be a limiter or something but lowering the sample rate (lower right part of the panel). And, if that still happens, go easy on low notes and/or sustain pedal. Or buy a faster computer.

OK -- managed to get a nice RnR overdrive by cranking it up. Hidden feature

Why would you not just keep the output level at bay? Works for me...

Ah! Good stuff.
You did not mention this in the release notes!

The reason why I haven't been able to solve this by "just keeping the volume down" is twofold:

1. I usually play Pianteq in a rock band situation, and when you are fighting with electric guitars, you often need to turn yourself up louder!

2. I have a volume swell footpedal on my MIDI master keyboard, and it is difficult to set it accurately and leave it in one place... particularly when the crackle point for volume overload was about 50%!
In the end, I had to turn off the volume pedal for pianoteq as it was just resulting in crackles all the time.

Maybe now I will be able to turn in back on again in 2.3!

trying Pianoteq 2.3 today with a high volume setting,
the behaviour does indeed seem improved from 2.2 -
less out-right crackles and glitches (although it still tends to distort a bit more than I might like)

Slightly off-topic, but take a look at this website to see what some guy did to his Burger&Jacoby piano: http://www.music-express.ch/akustik/klavier.htm - the photos explain it even if you can't read the text.

EvilDragon, all your remarks are welcome as we are precisely working on the future version.

Concerning your request “Add the property which controls how much the sustain pedal will damp when released and pressed again”, you can control it directly with your foot if you have a progressive pedal. You can also control it through the “key release duration” in the options menu (bottom right), as this controls the efficiency of the dampers – thus influences both the release of individual notes or of all of them when stopping the sustain.

Add-ons are in preparation - as you know we regularly produce new ones, and if ever somebody can provide samples from a Cymbalum, we could construct a model and add it to the KIViR collection!

I would like to let everyone here to know that I had treid to contact a famous brazilian museum, The Imperial Museun, which have a very rare and old pianoforte, claimed to had belong to a monarc from Imperial times in Brazil history, who proclaimed Brazil's independence.
   The idea was about create a Pianoteq Add-on of this historic pianoforte.

   The pianoforte it's a english Broadwood, dated from early XIX century.
   They also have a ver rare golden policromated spinet builted in Lisbon Lis 1788 by Mathias Bosten.

    It's the piano in the left corner os the photograph in the page: 
http://www.museuimperial.gov.br/tour_musica.htm

    Here a picture of aother Broadwood from about the same period, just let a idea of the shape:
    http://www.squarepianos.com/broadwood%20grand.jpg

   Despite all the work Moddart had to create a Add-On be free, and the technic not envasive or harmfull, seens the musem don't want to collaborate.

You mean you use some form of resynthesis to actually model the sound virtually?

- Since I don't think it's possible in the standalone version (if it is, it's not very well documented), I would like to be able to assign to each instrument/preset a midi program number of my choice. Right now it goes like this: C2 chamber = 1, C2 concert = 2, etc, with all the add-on presets having the highest numbers.
With this feature I could be able to organize a better instrument selection on my Kawai digital piano (for example I could have a button only for harpsichords, two only for historical pianos, one for my favourite ones, etc).
Being able to assign Pianoteq settings to midi program changes would be useful too (example: program number 56 = set equal temperament, etc).

- Upright sound modeling would be a nice thing to see. I would particularly like a physically modeled version of one of those old (historical) and very large upright pianos that no manufacturer makes anymore and a smaller and cheapish one for a more unsophisticated (but good) sound.

- No piano is equally in tune (or out of tune) in all keys. One or two "random detuning" sliders could add some realism or fun factor by setting the amount of detuning desired and how many keys in average should be affected. Also, defects other than tuning could be affected.