How to create this warm sound?

Hello EdenResident and welcome to this forum,

Here are some of my listening impressions of the Fur Elise recording you submitted.  I feel we need to get ourselves on a common ground as to what we are listening to, such that the various effects can be replicated via Pianoteq.

The recording sounds like a commercial recording, with stereo microphones placed fairly close to the curved side of the piano case.  Here is why I believe this is so: 

1)  Microphones are fairly close because you can hear the dampers' "hiss" as the damper pedal is depressed and released.

2)  It is possible to hear the hammers impact the strings, a feature that would have been lost, had the microphones been placed too far away from the piano's case.

3)  The sound is NOT panned from left to right, corresponding to any player's position.  Otherwise, the low notes would originate from towards the left channel, and treble notes would sound from towards the right channel.

4) The reverb sounds to me as though it was applied after the recording was originally made.  Why?  Because the actual sound of the piano is quite clearly rendered (perhaps in a dry recording studio), yet the reverb is quite lush.  These two qualities do not usually occur simultaneously in a commercial recording.  In a reverberant space, the initial piano's sounds will sound as though they are in a heavy "wash" -- not so with this recording.

5)  There appears to be some "gain riding" going on by the recording engineer.  Why?  Because the piano gets louder in places -- but the BRIGHTNESS of the piano does not increase in relative proportion to the increase in loudness.  Quite possibly a multi-channel compressor is at work here in the commercial recording.

Now, let's discuss how some of these features may be implemented in Pianoteq:

A)    This particular performance of Fur Elise is deliberately not percussive; restated, when the piano gets louder, the brightness does not increase as much as one might anticipate.  My particular choice of piano among the C3, M3 and K1 would be the C3 if you were to replicate the sounds of this performance.

B)   The nature of the sound field (e.g., panning) would suggest that you select the "C3 Solo Recording" patch as a starting point (or one of its variants), and let the microphones be disposed to the right of the piano, as you face the keyboard.  Again, this recording sounds as though the engineer desired a "listener's perspective" than a "player's perspective".

C)  Regarding reverb, I would suggest that you listen to the "audience perspective" of either the Concert or Hall impulse responses.  This was a rather "wet" sounding reverb in the submitted recording.

D)  Since the dynamic range sounded rather compressed in the commercial recording, I might suggest you either leave the Dynamics alone, or even slightly decrease the Dynamics slider.

E)  Since the brightness of the piano did not increase very much when the piano's sound got louder, I would also slightly reduce the hardness of all three hammer settings.

F)  Since one can hear the dampers' characteristic "hissing" sound, I might suggest that you slightly increase the Sustain Pedal Noise in the ACTION window of the effects section.

G)  Lastly, although this was a very pleasant sounding recording, I was a bit bothered by the contradiction of the equalization of the bass and midrange notes (sounding slightly elevated) -- something that is NOT usually heard in a live recording in a large hall.  In other words, the recording engineer may have applied about 1-2dB of EQ in the bass and lower midrange frequencies (up to about 200 Hz) -- OR -- he applied a small amount of "WARMING", a post-processing effect that some software plug-in manufacturers offer.  One that comes to mind is Antares WARM, a Tube Saturation generator.  This is optional, of course, in your venture.

These are only starting points, for sure.  But as you gain experience with Pianoteq, you will quickly learn how to listen for various sound characteristics in desirable recordings (i.e., panning, reverb, dynamic range, etc., etc.).  It is only after you are able to ascertain what characteristics of the sound you consider to be desirable, .... then you can go into Pianoteq and start moving sliders with some sense of purpose.

If you simply start yanking on the sliders in some random way, I assure you that you can become lost quite quickly.  The program is SOOO easy to modify, that is also easy to get lost in the beginning.

Happy playing.

Cheers,

Joe  <jcfelice88keys>

Without a doubt, this is some of the best advice (well-written, to boot!) that a relative "newbie" to Pianoteq can receive.  Someone should put this text on a virtual brass plaque somewhere on the website...  (I'm NOT joking.  ;^)

IOW, easy on the accelerator, cowpoke!

:-)

(And an excellent dissection, Joe, especially regarding the reverb & slight compression!)

Indeed!

About recrate such warm sound in details, I would say that ist's not possible yet, but you can get a beautiful sound, speacially with the new K-1 model.

Give some time, to more pianoteq updates, new versions, and I'm sure it will be possible.

Joe, what great piano sound recording analyze you did.

Another go at a warmer sound (not withstanding Joe's excellent comments which should be kept in mind at all times):

http://www.box.net/shared/p1ti1pvx9f

And obviously one must keep in mind that the dynamics are quite different as the two versions were played by different pianist.  And recalling Joe's previous comments (quite valid) - my rendering was done with a midi file that was recorded using a probably very different sound and dynamics.

Glenn

Beto-Music,
Glenn hasn't provided the FXP (right?) - are you sure you don't mean the FXP that I uploaded?   I was quite drastic with my hammer hardness reduction.  With my controller, I don't have to play very hard before the sound becomes brighter though.

Greg.

No Greg, I didn't provide an fxp - I just provided a link to an mp3 of Fur Elise (on box.net).  So I'm not sure what Beto is referring to as "cotton".  Perhaps my mp3 sounds like cotton.

In any event I posted a link so that EdenResident could do a listening test to compare the two versions (his and mine).

BTW - how do they compare on other people's systems?

Several times, we've discussed how our sound systems can affect our perception of sound - yesterday this was made very clear to me:  I took my Appassionata snippet to my piano re-builder friend and he played it on his system.  On my system, I can detect only one occasion where Schiff breathing can be heard - on his system I clearly heard three!!  I was shocked.

Glenn

PS - if anyone want the fpx, I'll post it.

How does it sound?  Is it warm enough?

Glenn

Don't apologize - listen to updated version.

Glenn

This is coming in late as I couldn't keep up with the forum lately, but wow, I have to join the chorus and praise jcfelice88keys for such an incredible post -- so much to learn from it, thanks!

Tell me what you think is lacking - this may be fixable.

Glenn

Ted, you've misunderstood what was going on there.

jcfelice88keys was writing about a recording of a REAL piano, with the aim of helping someone who wanted to make Pianoteq sound similar.

Well, now, I haven't bought Pianoteq yet because, for what I want, it doesn't sound enough like a real piano.  I prefer the sound of a good sampled piano, even though - judging from Pianoteq users' comments - Pianoteq's response to touch is way better than the response of sampled pianos.

In other words, it seems to me that people who prioritise a good playing experience over the realism of the sound prefer Pianoteq, while those who prioritise realism of the sound will put up with the less good playing experience of a sampled piano in order to get the desired sound in a recording.

---

BTW, there's something strange about this forum ...

Some Pianoteq users seem determined to deny ANY inadequacy in the sound, sometimes quite aggressively.  As far as I'm concerned, listening objectively on good equipment exposes weaknesses in the sound - and it's certainly not just I who think that.  Admittedly, I find it hard to describe exactly what's wrong - in fact, I find it hard to describe (differences between) sounds at all.  But, even so, I can tell it's not right - just as I can tell if a chair's unfomfortable, whether or not I can pinpoint the reason for the discomfort.  (And no amount of being told by someone else that the chair's comfortable will make it so.)

I find it a little strange that some Pianoteq supporters don't seem to be able to hear the deficiencies in its sound AT ALL (perhaps in some cases, the equipment they're listening on doesn't reveal enough?). 

Actually, I expect that those most aware of the deficiencies in the sound will be the developers - without that awareness, I don't imagine they could have made the improvements that they have made so far.

I'm looking forward to the day when I listen to the Pianoteq demos and hear a sound that I find realistic enough: then I'll buy it and enjoy the playing experience in the way that other users already do.  IMHO, what's been done so far is a remarkable achievement.  But the sound isn't yet what I want.

Hello Glenn and Cheq,

First off, many thanks to you, Glenn, for your kind comments and for stating my own case so effectively about the current status of Pianoteq.  I would like to add my own opinions and comments to add credibility to your writing.

Regarding pianos and Pianoteq,
I for one have daily access to a Steinway Model M and K Kawai grand pianos.  In fact, I also tune and regulate these two particular instruments at least four times per year, and I have been performing weekly or bi-weekly touchups on these very two pianos since 1994.   As a sideline career as a professional tuner (among other things) I have tuned at the rate of 4 to 6 different pianos per month on average -- ranging from several pianos per week to only a few pianos per month over the last ten years or so -- translating to approximately 600 different pianos over this time, with many additional instances of repeat customers.

The reason for stating these credentials is that I do know how pianos of various sizes and shapes are supposed to sound, and how they respond to one's touch as a classically trained pianist.  Having had this much exposure to two rather fine grands, has allowed me to expand my repertoire and musicality -- to a far greater level of competency than some 40 years ago, when I was contemplating a concert pianist's career.

Now, this is where Pianoteq comes in:
Admittedly, when I downloaded the trial version of Pianoteq in its infancy, I played with the demo version for only about five minutes -- and never touched it again for two or three years.  The original version of Pianoteq was "interesting", but didn't "play well" ... in my estimation.  The sound was "clever", but it didn't sound like a piano to my ears.

Backtracking for a second, beginning about 2005, I had purchased Synthogy Ivory, and was quite impressed with it, as compared to previous hardware piano sounds from Roland and Emu.  However -- and this is quite important -- the software was controlling ME (by having to make compromises), rather than I playing the software!

To be honest, I believe it was my classical training and technique that allowed me to make rather musical sounding performances with Ivory and its subsequent upgrades.  In addition to Ivory, I had acquired BDMO (the Bluethner Digital Model One) piano, having heard its audio demos -- but that piano software did not match MY technique; it was frustrating to be spending money and not getting the desired return of fulfillment on my investments.

After Nick Phoenix of EWQL had heard some of my performances on Ivory (Nick was soliciting live performers who would do justice to their then upcoming EWQL Piano Library), I made some demos for him that were considered as worthy to be included in the EWQL website demos for the Bosendorfer, Steinway and Bechstein libraries.

In my own opinion -- and this is not to denigrate the sampled library vendors -- all of the sampled pianos gave me the impression that I could make a performance that was "nice sounding", but none of these expensive libraries allowed me to express myself, musically, in a live performance situation.

To paraphrase Keith Emerson of ELP, who was originally talking about electronic organs as opposed to pipe organs:

"Playing on sampled pianos .... is like f***ing through a rubber."

When I re-discovered Pianoteq Version #3 -- for the first time -- I found that the instrument "played" much better under my fingers than any sampled library had done so before.  Of course, this is hard to describe to someone who has not experienced Pianoteq for one's self.

While I am not fooled into believing that Pianoteq sounds indistinguishable from a real concert-quality grand, it does give me hours of musical pleasure performing mainly the classic piano repertoire.  I enjoy voicing and "tuning" the piano note by note in the Pro version.  I do not use the octave stretching parameters at all; rather, I prefer to tune the octaves by ear (having been born with absolute pitch helps with this task) to achieve a beautiful sounding instrument.

* * * * * *

Now, back to the original reason for beginning this lengthy post:

I wish to address what I believe is behind the Pianoteq users who are in love with the instrument, and the skeptics who are rightfully wary of such skewed positive praise of Pianoteq:

There have always been concerns about the trustworthiness of online ratings of all products and services featured in website forums on the internet.  One would certainly hope that the truest and most accurate evaluations of Pianoteq will come from aggregating the opinions of a large and diverse group of people. Yet a closer look might reveal that the wisdom of crowds may neither be wise nor necessarily made by a crowd.

All online ranking systems, including Pianoteq, suffer from a number of inherent biases.  The first is deceptively obvious:  people who rate the product have already made the purchase.  Therefore, most of these well-meaning souls (Yours Truly included) are disposed to like the product.  The high ratings draw people who would never have considered a modeled piano in favor of a sampled library.

.... And if they hate the product, their spite could lead to an overcorrection, with a spate of very low ratings.  Such negativity exposes another, perhaps more insidious bias: namely, people tend not to review things they find merely satisfactory.  They evangelize (or appear to evangelize) what they love ... and trash things they hate.  These feelings lead to a lot of highly negative and highly positive reviews of the same product.

Self-selected online voting has the potential of creating an artificial judgment gap:  as in modern politics, where only the loudest voices at the furthest ends of the spectrum seem to get heard.

Enough of my rambling for now,

I hope this information was helpful or at least thought provoking to you.

Cheers,

Joe