boxy sound, missing openness & clarity, muddy mids?

Thanks for your tip! Unfortunately that preset does not solve it for me.

It may also be a modeling thing, something that you can't overcome with the current version. If I browse through the legacy piano models I notice that this flat/muffled sound has improved from model to model, and *in comparison* the sound is really lively and clear today.

So, I am definitely a big fan of Pianoteq as well, since the real feeling rules for me. But I think we are not there, there is room for improvement. While version 4 is honestly very good, maybe  5 or 6 will real perfect. :-)

It could sound obvious but check your velocity curve also; when playing hard you should easily reach velocity values from 100 upward. Otherwise edit the graphic or, if you're in trouble, start with velocity presets (clicking in 'velocity' tab over the graph), or do the very easy calibration.
Then I'd move hammer hardness sliders on slightly higher values.

We are much of a mind. So I'll post my latest FXP (it has mike-treatment as its focus, otherwise is nearly default), after finishing up here.

Exceptions to the mike focus are : no Reverb (no mud from that source); Blooming feature is used (OK, it's a sort of mud, a sparkle sort in the treble which adds sheen, plus is a coloration I've heard in all real pianos  I've ever come across); somewhat reduced Dynamics (default is 50, and it's reduced to fight the heavy hands of the pianists who as a breed produce the world of MIDI files); Limiting's applied (same reason), when playing oneself matching touch to hearing is natural, so Limiting  isn't needed . Too, the scala scale I'm using is easy changed (it's there to provide a pleasant variety across the scales, to escape the Equal Temp straightjacket). Likewise everything else is easy turned off, or varied to taste.

The miking itself is kept as simple as possible, and as close as possible to the hammer level - and incidentally pretty much on the path to a player's ears.

I played around myself a little and came up with this modified D4 Blues preset

http://www.forum-pianoteq.com/uploads.php?id=1739

It's a mellow piano but does not feel as boxy as the normal D4. Well, my best result so far, though I am not 100% content yet.

I noticed that my patch somehow sounds like these cheap artificial stereo makers. Phase shifted somehow. When comparing directly to the original D4 Blues this strikes me. However after a few minutes I like it.

Just the same shortcoming I am seeking to overcome now within the framework of PT and also with some external tools (I have Reaper with it numerous AddOn Effects and Tools).

This is a longer process process,  I'll share my more definitive results later. Just some directions:

My fixa idea was for long that the major shortcoming of digital pianos, especially of sampled ones, is the fixed narrowed down sound perspective with them. (It is the logical limitation of the sampling process, where the samples have to flow through the close microphones which are depicting one particular phase image of the string which you can't get rid of fully with afterprocessing.)

It was a very amazing effort by Pianoteq - beside of not needing this critical micing  - that it lets make use of the flexible micing options. So I have bought an Audio Interface (Focusrite Scarlett 6i6), which allows the flexible redirection of multiple channels (Microphone Outputs) to more than a pair of stereo boxes. I can now combine a a good quality monitor pair (Genelec 8020b)  with the internal sound system of the KAWAI CA65 (+later also with my 12" SUB or som other better computer sound system).

(I have the possibility to drive on 6 output channels completely separately, with the help of my old EMU0404USB which I need for convert the digital S/PDIF output of the 6i6 to balanced audio out.)

This all costs some time to learn, (along with the many possibilities and peculiarities of the Scarlett and it's MixControl SW) but the first results are encouraging. Since the dull attack problems as I perceive it is with the lower mid one ore two low-mid octave range, I experimented with a very close mic to capture set to capture that string area, near to the hammer hitting point and mix that mic throughout the other channels. The first results I found encouraging, but as said, I need more weekend time for further experimenting sessions.

Another direction would of course be to improve the key attack in this key range separately. I think the adding some bold disturbance effects in this key range could help, but I don't  have any experience with the PRO version so far - but this is a future.

One interesting question however remains, whether some limitations or internal parameter settings of the present v4.5 should be improved at the first place or the Instrument Models are not yet optimized with the last engine update. (They were brought out before the last engine update - these seem to be supported by Bright Open Grand FXP which is a good try, along with the D4_Blues_2_, e.g. ).

I am convinced, PT has the potential to overcome this major handicap, where sampled instruments might have their most advantage yet IMHO: lacking of attack brilliance in (low-)mid key range.

That last example sounds excellent to this pair of ears! Definitely sounds "open" instead of "boxy".

Many thanks for that helpful explanation, Philippe.

So "level compensation" is just a shorthand way of equalizing the relative volume of the various microphones.  I guess what I'm hearing as more "open" is just my personal preference for the natural amplitude/volume differences resulting from microphone proximity.

Aside from my preference for the expressiveness and playability of Pianoteq (over sample-based digital pianos), it's so much more fun to be able to play with all of these parameters!  Many thanks to Modartt for such a wonderful instrument.

That's very kind and generous of you: The demo was just a quick grab-bag to show off the fxp.  The first minute is my variation of an old Vince Guaraldi theme, and the bit afterwards from my latest song-in-progress.

Hope you enjoy the preset file!

Just a little follow up on this: I think that post EQ (outside Pianoteq) can help a great deal also to make the sound less boxy. Pianoteq tends to emphasize the mids (which might be the natural piano sound), if you damp frequencies in the range ~ 200 - 1000 Hz this helps. The internal Pianoteq EQ works a little differently I think, you cannot really reach the same results with it.

I have greatly improved the tone of PianoTeq by opening it inside my DAW(I use MixCraft 6) and using a different EQ and reverb than PianoTeq's.  I use ValhallaRoom for the reverb and Pultronic Tube EQ. The difference is astounding!! Also as already mentioned getting the velocity setting right is crucial.

I'm actually using one of the factory presets on the Pultronic that is for piano, but the whole point of my post is to that you need to use a better EQ in general. I've tried several different EQ's and most all of them sound better than what is in PianoTeq. Same goes for the reverb.

Hi , you can try the demo here....

http://www.g-sonique.com/demos.html

I am still trying to narrow down what bugs me about Pianoteq's sound. And it is definitley the microphone setup. If I move microphones around, it sounds like changing a comb filter, or sometimes even like a formant filter. Definitley there are partials in the spectrum which get boosted or damped depending on the microphone setup.

I mean, I know this effect from reality, like when a jetplane passes it sounds like the tonality is changing (I don't mean the Doppler effect here, I mean this kind of phasing sound).

So even that this seems to be real physics, you never notice it so strongly with a real piano.

But I am not a microphone guy. Is the sound changing that much depending on the microphones position?

I've tried playing with the mics and found to just leave them in the default position works fine, but closing the lid and using a different EQ and reverb have made the best improvements to my ears and easy to do. Have you tried this yet?

I was able to listen to some sampled piano demos, and I think what I was criticizing about Pianoteq's sound is just the reality of a multiple microphoned piano. All sampled pianos have that same sound characteristic that I thought was bad modeling. But it seems it isn't, it rather is good microphone modeling.

The fwy noise must be dreadful, to drive you to Metal. However if it's your chosen poison, no harm. An alternative is to listen to the fwy for features you might say mock, creatively. Perhaps grab unusual rhythms, then capitalise these into song. Revenge served cool, hot, or as you like it; at any rate in the spirit of rising above it just because you've got the mind.

Or plain make a project of living in the country. Raise wheat or corn. Freshwater lobsters. Frogs. Bees. Hm, noise, perhaps not.

Hah!

While playing around with parameters I found another source of the something in Pianoteqs sound that bothers me, maybe THE source of it: the strike point!

If I touch that and move it around slowly, I constantly think "nah.... nah.... nah.... not yet.... nah.... not either... nah... ". I am trying to find the right spot. But then after a minute or so, I tried the whole range and still there is no setting that I like. Everything sounds like a notch filter or comb filter or something the like, that sounds odd to me.

This effect gets especially noticeable when you only use one microphone, mono sound.

So, maybe Philippe can say something about this feature? Is it some kind of overtone modelling, or is it really just a filter? Either way, I think here may be potential for improvement.

Well I know all that (except I wasn't sure if its maybe just a filter). But it is for sure not doing "exactly" what happens in a real piano, because it is only a model ;-) And that's what I tried to say, varying this parameter shows where the weakness of Pianoteq sound lies, and where there is room for improvement.