Topic: Matching Pianoteq to the sound of a real acoustic piano

While I have been a fan of tinkering with Pianoteq in many ways and trying many of the presets as well as many of the files that people of uploaded, I must say that the closest electronic piano, either sampled or modeled, to my own acoustic piano is by using my piano's MIDI sensor rail to play "The Grandeur" from Native Instruments (my piano is an 1885 Steinway upright model F). Nothing from Pianoteq seems to do quite as well for me, unfortunately, based on my current set-up which of course involves my speakers (Emotiva Airmotiv 6's), speaker positions (I did have them on stands to each side of the piano facing the ceiling to be more omnidirectional, but today found that they sound better on each side of the piano facing each other across the underside of the keyboard), and even the large resonant wooden box made by Steinway in front of me that adds to the sound playing the electronic instruments.

One gadget that has made this quite clear is my purchase for travel of a CME X-key keyboard. This has allowed me to sit with my piano played acoustically and the other keyboard played electronically, altering back and forth, and even simultaneously. (This was something that was a bit difficult to do on an instant basis using solely the Steinway's keyboard, as it involved pulling a lever to engage the stop rail and pushing buttons to silence or un-silence the pickups from the QRS system.)

Despite the relative accuracy of the Native Instruments "Grandeur" piano, as well as the acoustic sound of my upright Steinway, I love going back to Pianoteq, because my goal is to figure out a generalized representation for the pianos to sound best from the player's position at the keyboard on my set up. This would allow me to have more fun with everything from the Steinway's D and B as well as the other modeled instruments.

This task can be best broken up into two categories: the first is of microphone placement and settings, and the second is of modifications to the "dry sound" with delay, reverb, etc.  I will discuss what I know briefly in reverse order:

Microphone placement is my most recent bugaboo. Although I had not initially had success just using the earphone representation in Piantoteq, because playing this back through the speakers on both sides of my piano give a fairly sterile and closed sound. As a matter fact, if I put on my own earphones and listened to the sound through them, it is usually from a non-earphone representation that sounds the most open (in other words, open enough that sometimes I have to take off the earphones to see if I'm really hearing through the earphones or through the speakers outside them). At the same time, the characteristics of the sound of the piano vary tremendously depending upon where the microphones are placed. Placing microphones where the pianist sits does not seem to give the most realistic sound. Similarly, placing them to the sides of the piano, where my current speakers are, does not give the most realistic sound either. Instead, somewhere over the soundboard is likely best, but then playing it back that way does not give the sensation that the sound is coming from the piano directly in front of me, even though the timber of the sound is better.

For some reason, many of the setups that people use mix two different microphones both into the left and the right channel – I do not know why. Also, not talked about anywhere in the forums or in the instructions is the use of the delay setting for each microphone. Certainly, a microphone placed further away will have its own delay characteristics, and there are separate settings to turn off the distance delay per microphone, as well as an adjustable delay for each microphone, as well as a setting for delay in the sections that modify the sound overall – some delays on some microphones seem to add to the realistic impression that I am sitting in front of the piano, where other delays hurt that impression.

And if that isn't enough, the different types of microphones each have their own characteristics. I am not a sound engineer, and I am not familiar with these other than by trial and error. Many of the different representations using Pianoteq are quite realistic in the treble or midrange, but poor in the bass, and vice versa. I never seem to know if that is from microphone placement, microphone characteristics, or perhaps in the representation of the instrument itself – very confusing. Once again, the sampled piano "the Grandeur" sounds more realistic from top to bottom on my system.

In this long email, now getting to the second aspect that I mentioned a few paragraphs ago, the difference between 'wet' sound and 'dry' sound, more things help or hurt realism: in theory, I would think that sitting in front of a piano and trying to duplicate it that I should want a fairly dry sound. The truth is that even my own piano has a wetter sound than most of the dry sounds from the Pianoteq instruments. While the reverberation of the room that you are in does not need to be added to by Pianoteq (in other words, since I am just trying to simulate my own living room, that type of reverberation comes both from the acoustic piano as well as from the electronic piano played from the speakers; but if I wanted to pretend that I am in a cathedral, that would be another story). While room effects may not need to be added, it is the piano effects themselves that are a little bit more complicated: my own acoustic piano has its own inherent reverberation from the structure of the piano. And electronically modeled piano played out of speakers on either side of my piano both needs to have a little bit of reverb to simulate the casing of my own piano, but even that is not perfect as some of the sound out of the speakers in fact placed through my own piano even though the speakers are on the sides and pointed behind it. Once again more confusion!

So this is where things are this Saturday as I keep working to understand the system, and try to get the Pianoteq pianos to sound as similar as possible to my acoustic upright.

David

- David