Thank you for pointing out this. And thanks for all your information/knowledge in this forum. It’s interesting, valuable and useful. Thanks!

The long answer?And questions.
Maybe there is not much reality in the recordings. Many albums rock, jazz etc is compressed, equalized and so on….so they can’t sound realistic. So, we can distinguish differens.
   And then we have the speakers(and headphones). In my opinion, we don’t have the perfect speakers yet (have not found the best for myself). And if we have in the future, can they be made, so they can sound the same in different home and rooms? How can manufacturer eliminate an acustic problem precisely in my room and others at the same time? Maybe sensors in the speakers, that can correct reflections and different acustic problems, like a car’s parking assistent? I am not surprised, if it is the next step in speakers design in the future.
   And we hear different. And people are used to hear instruments in a certain way. And personal preference have influence on how we make a decision, good or not so good sound, real sound or not. 
   I think Pianoteq sounds as a real piano, yes, but we are still struggling (have to struggle) with our musical tools, equipment, and the space/room where we are listening, to get the real sound out (got better acoustics when moved some furniture).
   But, thanks Joe, and everyone in this forum for sharing your knowledge and ideas.
   Well, that’s what I think about it.

When I hear myself playing Pianoteq, with speakers or with headphones, should the headphones\speakers be as flat as possible?
I mean, is this considered to be the best way to use Pianoteq, sonically speaking?
Is that how Pianoteq designed to be used, by the developers?

If so, should I assume that monitor speakers are the best amplification for Pianoteq?

My answer is definitive "no" to all the questions.

Well, we're getting back into the 'realism vs. perceived realism' debate now! I wasn't implying that I was unhappy with the 'perceived realism' of Pianoteq - far from it. With good equipment, you can indeed get pretty darn close to the experience of playing a real grand piano.

I completely agree with you!  In fact, I own a small upright piano obtained from the early 1970's, but play Pianoteq almost exclusively. About the only time I play the upright piano is when company is around -- AND -- the computer isn't turned on.

Viva la Pianoteq!

Joe

And neither do manufacturers think that it's unreasonable - Witness products like Yamaha's AvantGrand.  While you and I both may argue that it doesn't accomplish the goal, the goal of the product is to emulate a grand piano through keyboard, action, case, and speakers.

- David

I must confess to not having read every word of every contribution to this thread, so hopefully I do not step on anyone's toes with what I am about to write:

Just for the record, I am NOT dissing Pianoteq in any way. 

Of course, we may discuss flat frequency response, sampling rates, and additions of EQ, compression, FLAC, DAC, MQA and/or third-party reverb, but it seems to me that it may be unrealistic to expect an instrument as tonally complex, massive, and physically voluminous as a piano ... to be reproduced accurately, realistically or forcefully with 1-, 2- or more electrical loudspeakers (to say nothing of headphones) whose own physical volumes occupy the approximate size of a shoebox or a microwave oven.  The largest concert grands exceed 1100 pounds / 500 kg with soundboards exceeding a few dozen square feet in radiating area.  Even moderately sized grand pianos weigh ~50+% that of concert grands.

True, some amplifiers' combined mono-, stereo- or multi-channel outputs can exceed one horsepower (746 watts) in total wattage, which converts to approximately 7 amps of electrical current @110V drawn from the wall outlet (or 25+ amps @ 30V drawn from the amplifier's output terminals).  I have no definitive answer to this observation, but aren't we asking quite a lot of modeled- or sampled piano's sound and audio radiation patterns to be indistinguishable from a large acoustic piano played live?  I think this "something-yet-to-be-thoroughly-defined-and-classified" is what at least partially explains how a person can instantly and unequivocally recognize the sound of a real piano played live versus the sound of its reproduced audio.

Food for thought.

Cheers,

Joe

P.S.   Isn't it also interesting how a person can recognize an electric guitar being played live through a loudspeaker powered by a Marshall amplifier, as opposed to a recording of the same guitar/amp combination played through an electrical sound system? What characteristic is there of the "sound" that reaches our ears that allows us immediately to distinguish a live acoustic- or amplified instrument from a recording of the same performance or even of the same notes being played (in the event of a mediocre performance by a well-intended amateur)?

The short answer:  I don't know.

bm wrote:

It suffices to note the size of a wav file before and after equalization to be convinced factually.

WAV is a pure uncompressed audio format, it will have the same filesize regardless of content for the same sample rate, bitdepth and duration. i.e. one minute of silence will have exactly the same filesize as one minute of white noise, or a sine wave. This is easy to verify.

Sorry for this error, you are right.
In fact my test was in .flac format 96Khz / 24bits Sample on Kirie Mass in B minor, duration 8.5mn-> 92.5Mb size) with over-amplification test of treble and bass for a very old person, versus same piece with equalization much less heavy, even sampling 96Khz / 24 bits but then1 87 Mb size.

Bruno

It suffices to note the size of a wav file before and after equalization to be convinced factually.

WAV is a pure uncompressed audio format, it will have the same filesize regardless of content for the same sample rate, bitdepth and duration. i.e. one minute of silence will have exactly the same filesize as one minute of white noise, or a sine wave. This is easy to verify.

When I hear myself playing Pianoteq, with speakers or with headphones, should the headphones\speakers be as flat as possible?
I mean, is this considered to be the best way to use Pianoteq, sonically speaking?
Is that how Pianoteq designed to be used, by the developers?

If so, should I assume that monitor speakers are the best amplification for Pianoteq?

Hello everyone,

My feedback (with all its limitations: I'm not professional) is that a static frequency equalization performed downstream of a synthesizer (pianoteq) is very quickly lose information in the musical message. It suffices to note the size of a wav file before and after equalization to be convinced factually. On the other hand, it is sometimes possible to better emphasize the musical message with a dynamic equalization obtained with a multi-band compressor, to bring out certain details initially emitted with too low a sound volume. I was seduced by the plugin Tdrnova which there is a free version already very interesting. (https://www.tokyodawn.net/tdr-nova/) There are certainly many other products that do this.
In this regard, it would be interesting if one day Pianoteq allows its standalone mode to be able to use at least one audio module vst the output choice. Today you have to go through a DAW for this, which is a bit heavy when you just want to play for a quarter of an hour ... Obviously the addition of such a vst can cause problems latency.
As for listening to headphones (I prefer the use of an acoustic piano to that of a virtual instrument on speakerphone, (but I also have the chance to do that), I believe (for those who support listening to headphones) that it is preferable (when possible) to invest in a better headset than in an external device to equalize its bandwidth, my feedback is that it is more It pays to first invest in a good headphone, much more than a headphone amplifier and sometimes even a better quality DAC (or sound card), after which the choice of model is a matter of personal preference. (and unfortunately budget for the best models), knowing however that a relatively flat bandwidth is really useful only up to 14 kHz maximum (11 khz for the over 60s) and that the most important is above all harmonic DISTORTION, to get a sound natural (distortion not only 1Khz but 30hz to 5Khz ideally specifying for what sound level, harmonic distortion to which we must add all the parasitic vibrations not always harmonic of the headphones audi in the range 100hz-15khz).
The piano is a formidable instrument to allow very quickly to eliminate in a few seconds of listening headphones audio (including some of the most expensive) on the criterion of parasitic vibrations, as soon as you raise a little sound level.
For the regularity of the bandwidth, this is less critical, the human ear adapts to a less good headphones (same for an aging ear) on this criterion and the brain compensates to a large extent. (as long as you are not constantly comparing several headphones on the same piece of music - hence the permanent dissatisfaction of some audiophiles ...)

Best regards

Bruno

Gilles wrote:

I didn't keep it in the end. I like the unequalized sound better. Even though the result was flatter, I prefer the more detailed sound of the raw response, probably due to a slight peak in high frequencies.

How long did you use it? Because yes, the compensation curve does make everything sound "weird" at first, and yes it does make you think your (possibly expensive) headphones are a piece of crap (and looking at all of their frequency responses, they all are crap because none are perfectly flat ever!). But, this is literally like placing a set of speakers in a completely different room. Of course they're gonna sound completely different, and that a transition phase is necessary to get accustomed to the new sound. That can last for more than a week, for sure...

I used it for maybe 3 or 4 days, comparing with the bypass switch on varied musical material as well as Pianoteq. I also tried with my Grado SR 80 although the supplied curve is for the more recent model SR 80i but I suppose they are similar. Both headphones show a slight high frequency bump that, in my case, makes for better details. But then, due to age, I may have a loss in that region that is compensated in this way. In no way does this slight bump sound piercing or annoying, so it works for me. Maybe younger ears would react differently.

The compensated bass (both my headphones are a bit bass-shy) is not readily evident for most material, and having tested previously with pure sine waves, I know that the AKG goes down to about 30-40 Hz very easily.

The correction did not introduce any bad artifact, but there is much more to flatness in headphone design, so correcting this parameter may or may not be useful, it's all personal choice. For instance the AKG has a large round over-the-ears design that produces a much better soundstage than the Grado.

I didn't keep it in the end. I like the unequalized sound better. Even though the result was flatter, I prefer the more detailed sound of the raw response, probably due to a slight peak in high frequencies.

How long did you use it? Because yes, the compensation curve does make everything sound "weird" at first, and yes it does make you think your (possibly expensive) headphones are a piece of crap (and looking at all of their frequency responses, they all are crap because none are perfectly flat ever!). But, this is literally like placing a set of speakers in a completely different room. Of course they're gonna sound completely different, and that a transition phase is necessary to get accustomed to the new sound. That can last for more than a week, for sure...