Thanks.  Great ideas.  I have done much of this, but not done much with reverb (though I love the convolution reverb file that sounds like a piano's wood case reverb.  Here on my cellphone, I forgot the name - and I wish we could add two reverbs: one for the case, and one for the room).

Do try out the free MuVerb. The interface looks stripped down, but it's a serious vst effect from the people who make Mulab.

I would imagine that the last thing you want to hear about right now is still more software to add to the effects chain, but there are some vst reverbs that allow you to use two wave files and control the way that they intermingle. I have not tried them, but this one looks interesting, since it offers so much control over files and how they intermingle:  https://www.liquidsonics.com/software/reverberate-2/

I'm not sure that two wave files are needed, but it's hard to avoid being curious about this new, to me, kind of reverb. (I'm going to post a general question about it as a new thread. Perhaps we can learn more.)

1. Experiment by trying different presets, of course. You may find a given preset that works better than others. A preset intended to sound like a relatively distant classical piano, after all, just won't sound close to you as you sit at the keyboard.

2. If you have the pro version, find a preset that comes closest to what you want, and then move the mics around, imagining that each mic is one of your monitors\speakers. Do try swapping out those mics.

3. Again, if you have the pro version, on the Mics page, try right-clicking on the image of the mic above the grid and turn OFF the Proximity Effect Compensation. But be aware that each mic will of course then will up more bass freqs.

4. Experiment with the reverb. Note that you can change the pre-delay and the degree to which late or early reflections are emphasized. You can do a lot with these two settings. (And try out the two Piano Room presets, which are listed near the botton of the presets list for the Reverb.)

5. Try using a wave file for a convolved reverb instead of the default reverb. There is much that can be done here. Many wave files for rooms and some for piano soundboards can be found. Be prepared to have some fun, here, as well as to get both good and bad results. 

6. Something I've only recently found: In the VST version, turn off the internal PT reverb and try a third-party reverb that allows you to set separate pre-delays on each channel. The free MuVerb may be of interest. This may (MAY) help to create a sound that you like. Setting separate predelays, if done with some care and by ear, can imply to the ear that the one wall is slightly closer to the piano or at least the player, than another. If the predelays are set too differently, the effect is too noticeable, and if the predelays are set too close, you won't hear any change in the sound at all. But there's a sweet range, to my ears, in which the freqs overlap as they arrive at different times. This can occur with a small predelay on one channel and with the other predelay set to about double the time or less. Again, this is only valid if the initial predelay is set so that the predelay is not noticeable as a pause. (One of the effects of these settings, to me, is that there is a slight "thickening" of each note, since the freqs are arriving at the ears at very slightly different times, but still overlap and mingle. This may be what we are accustomed to when sitting at a piano in a real room, where the piano is not centered. and often we are not sitting dead center at the piano. I, at least, find that these separate delays creates a variation in the impression of the piano. I don't know if I can accurately say that it is better or more realistic than the default reverb. Thre's always subjectivity at play.)

Does anyone understand why this is?

https://www.dropbox.com/s/guuxf68fnrb4h...4.fxp?dl=0

+ I use an app (BOOM 3D) that expand the audio in 3D, so de FX is better on the sense of et the impression you are sitting on the front of the piano

The only thing I dislike about the Bluthner presets is the excessive key-off noise. For single notes it's okay but for chords (e.g., Chopin Prelude Op.28 No.4 left hand) it's way too much. Not an issue in Pianoteq Std or Pro, because you can just turn it down; I'm not sure about the Stage version but I think it's adjustable there too under Action. In any case I recommend you try the Bluthner presets in demo mode.

However, if you find the D4 sounding great when listening to it on a recording away from the keyboard, then keyboard and speaker placement will be the main issue. For example, if you have your keyboard and speakers set up against a wall, it's going to be very difficult to get a good sound. Ideally you need some space between the speakers and the walls of the room, and even more space between your keyboard (i.e., your ears) and both the speakers and the walls. If that is not feasible, then well placed acoustic insulation on the walls could help (acoustically absorbent material that reduces reflections).

Put your monitor speakers below you, and point them upwards at the ceiling. 

I have done that with one to either side of my keyboard (an 1885 Steinway F upright piano with a QRS Sensor Rail) and have now added an additional one behind the piano, splitting the right channel (I would have added both right and left behind the piano, but one of my two smaller speakers is not working currently). 

With that setup, and a little playing with volumes, microphone placement, and  setting the stereo width at 0.35, I can hardly tell real vs. pianoteq+mute rail.  Ok, so I can, but Pianoteq+mute rail is REALLY good with this set-up.

David

Sarasota, FL

I like the IDEA of this, at least in theory.
With what I have and where I have it - I can't do it.
(and in any case my priority is more toward improving my playing than improving the illusion that I have a physical piano).

I think it gives the room the opportunity to respond as it would respond to a (physical) grand piano, which project MOSTLY vertically (and somewhat horizontally via the lid).

Most of the sound we get when playing comes to us VERY indirectly, this probably imitates that indirectness quite well.

I have done that with one to either side of my keyboard (an 1885 Steinway F upright piano with a QRS Sensor Rail) and have now added an additional one behind the piano, splitting the right channel (I would have added both right and left behind the piano, but one of my two smaller speakers is not working currently). 

With that setup, and a little playing with volumes, microphone placement, and  setting the stereo width at 0.35, I can hardly tell real vs. pianoteq+mute rail.  Ok, so I can, but Pianoteq+mute rail is REALLY good with this set-up.

David

Sarasota, FL

Lanny

Kalessin, regarding your comments about wooden transducers etc.: That makes sense in theory. But doesn't Pianoteq model all the piano sounding-board and other physical aspects of the piano with such realism that that side of things is more than adequately covered?

Yes, and no. It's two different things.

Pianoteq models a piano more or less completely. That is you can think of it as e.g. a virtual Steinway Model D sitting in a perfect studio with absolutely neutral acoustics and no noise floor, and an arbritrary number of 1-5 microphones that can be freely positioned. On this side of course the resonator behaviour of the piano is modelled quite well, so well in fact that you can get all the problems 'real' sound technicians face when they position the microphones (e.g., sound wave minima and maxima at different positions).

This does not, however, change the fact that the only device creating sound waves in your room is a pair of speakers, which are (from a physics point of view) much closer to point sources than the actually rather huge active area of a piano. A concert grand has a resonator surface of 30-40 (or even more) square feet; your two speakers have about 2 between them.

So what Pianoteq actually does is simulate a recording situation. Recording a piano with carefully chosen microphone positions and playback over speakers can result in a somewhat realistic 'illusion', but I fear it can never be quite complete, much like a console game will never be real life, regardless how detailed the graphics are and whether you even use a 3D monitor. You will always need a certain amount of acoustic 'suspension of disbelief'.

The problem with room acoustics is that they are never absent, at least if you don't actively insulate the walls. Sound waves propagate in all directions, and some will be reflected from the walls and travel back to you from there. That is, your ears will always hear a mixture of 'direct' speaker output and 'indirect' speaker output. The near-field speaker approximation is more or less the (actually rather unrealistic) hope that if you sit in front of the speakers 'just so', you will mainly hear the direct speaker output and don't have to worry about room acoustics.

And yet, this is most often still untrue. For example, longer wave lengths tend to be reflected more efficiently. This means that a wall behind the speakers will often seemingly amplify the lower frequency response. Hard surfaces in the room can mean that high frequencies are bouncing around, meaning you might also have an unwanted peak there.

Apart from that, no-one does ever 'not move' their head. We always move our head ever so slightly, it is how directional hearing works to a large degree: we register small phase differences when moving our head (which is not really even a conscious thing), and our brain calculates probable sound source positions from that. A complete illusion would require using headphones and tracking all head movements, permanently calculating the correct sound input. This is theoretically and even technically possible of course, but very, very difficult to get right.

In your case, I still believe that the problem is either too much or too little room acoustics. As I mentioned above, virtually no room is acoustically neutral. This also means that most humans find acoustically insulated rooms highly unnatural and often even uncomfortable. This is why music recordings on occasion implement deliberate frequency distortions: they are pleasant to the listener because they appear more natural. Usually this means that lower frequencies (below 500Hz) and moderately high frequencies (above 2kHz) are a bit amplified; many hifi speakers and headphones do this as well.

So, in your case the near-field approximation might also actually hold 'too well': you might just hear sound that is 'too dry'. Which means that it is missing all those subtle distortions that make it appear more natural. You will never be able to make it appear like a 3m grand piano behemoth is sitting in front of you, but increasing reverb, maybe adding a delay filter and possibly even pushing low and/or high frequencies in the equalizer might 'help'. I'm sorry that I cannot give you a more concrete answer than that.

I think so? Anyone know for sure?

Yes, you can. Demo version only has two limitations, 1) 20 minute time limit, 2) numerous silent black keys.

rjpianist and njaremko, thanks for the FXPs. Do you know if it's possible to try them in the Demo version to hear the effect of your adjustments?

I think so? Anyone know for sure?