Best piano sample library? especially Steinway?

[Shooshie]

I'm liking the Authorized Steinway piano. But I can't tell for sure until I play it. In the demos, however, I've noticed an annoying range of samples at a particular velocity layer (probably), which seem to decay too quickly. Not sure what's going on with it, but I heard it in several demo recordings. It's not pervasive -- you only hear it in spots -- but it sounds very unrealistic when it appears. Not sure what to make of that. I hope that's not some kind of goof that ruins an otherwise fantastic set of samples. I was surprised they used a live player to do that. I'd have chosen a mechanical player over live any day, just for accuracy and consistency. This is not music; it's notes on an instrument. You use those notes to MAKE music, but adding a live player at that stage of the process diminishes the returns. Maybe if someone is watching the meters and keeps saying "do it again" until they get the exact velocity they're after... I don't know. I assume they know what they're doing, though, so I won't question it too much. I'd sure like to try it out.


Shooshie

[jcfelice88keys]

Hello Shoosie,

I agree with you that the Garritan Steinway has a highly musical sound about it. At the same time, while listening to the demos through earbuds, I hear extra-musical noises that sound like the dampers falling down onto the strings after releasing the sustain pedal. I am guessing at the origin of these mechanical sounds, because they occur at the same times within musical phrases when the pianist would be manipulating the pedals.

In my opinion, these extra-musical noises in the demos are too loud in relation to the music, especially in microphone positions meant to emulate a room-like perspective. I would have no trouble with this feature if the volume of these sounds is user-adjustable, as in Synthogy Ivory.


Cheers,

Joe

[Shooshie]

Hello Shoosie,

I agree with you that the Garritan Steinway has a highly musical sound about it. At the same time, while listening to the demos through earbuds, I hear extra-musical noises that sound like the dampers falling down onto the strings after releasing the sustain pedal. I am guessing at the origin of these mechanical sounds, because they occur at the same times within musical phrases when the pianist would be manipulating the pedals.

In my opinion, these extra-musical noises in the demos are too loud in relation to the music, especially in microphone positions meant to emulate a room-like perspective. I would have no trouble with this feature if the volume of these sounds is user-adjustable, as in Synthogy Ivory.


Cheers,

Joe

Hi Joe! I'd really like to hear YOU play something on that Steinway. And tell me if you hear that little premature decay problem. Maybe it was just the way someone was pedaling, but it sounded wrong.

Also of interest is that at least one of the demos is actually a combination of the virtual instrument plus the REAL piano played live in the same hall where the samples were recorded. They sound very much alike, not surprisingly, but you can easily tell where it switches, and which is which. If I remember correctly, it's one of the Mendelssohn SWW.

Shooshie

[jcfelice88keys]

Hello Shoosie,

Regarding my performances on the Authorized Steinway, I have been in email contact with Gary Garritan, offering my services as a demo performer. He replied that he enjoyed my playing on an mp3 of the EWQL Bosendorfer and Steinway that I sent him. Gary asked for a little more information from Yours Truly, which I promptly sent him.

The ball is in his court.

In the meantime, Nick Phoenix has been kind enough to add two more demos of my performances on his EWQL website: Rachmaninoff's Liebesfreud on Steinway, and Beethoven's Hammerklaver Sonata, last movement, the tour de force fugue of Opus 106. I would be interested in your opinions and/or comments on these performances.

Cheers,

Joe

[David Polich]

I had taken almost a year off from using Ivory, then decided yesterday to install it again on my new PC (whioch I was forced to buy because my old Sony Vaio died in January). For much of the past year I've been using Pianoteq.

And, it's pretty much back to Pianoteq for me. The trouble with all these sampled pianos is that they sound to me like collections of individual piano snapshots, they don't interact like a real piano. Once I got Ivory installed and running, I played a few notes (using my Disklavier as controller) and it all came back to me why I didn't like it. Particularly, in every one of the Ivory pianos, there are certain ranges where the samples are just plain lame, in all velocity ranges, and there's nothing I can do to fix them.

I can think of busy mixes where I'd go back to using Ivory, because it does cut through better than Pianoteq, and the attacks are more realistic. But as to whether Ivory responds like a real piano - well for me there is no contest, Ivory just doesn't "feel" like a real piano, whereas at least Pianoteq does.

Of course, I do have a really fantastic piano "plug-in" - my Disklavier MIDI grand, and it gets the nod when I've finally decided on the best arrangement for the tune and it's time to cut the piano track "for real".

[mhschmieder]

I came to the exact same conclusions, and thus am using some sampled libraries in thicker mixes so the piano cuts through better (where it isn't being soloed), but was curious whether the Italian Grand uses some programming tricks that improves the intra-note continuity, as many reviews say it sounds more fluid than the other Ivory libraries (and the on-line audio demos tend to demonstrate this as well).

I'm able to get pretty close to the Steinway, Fazioli, or Yamaha sounds with Pianoteq, when doing soloed parts, and certainly feel more connectedness when playing it. I also like that I can more easily emulate other piano brands that don't get attention from the samplers . It's also great for uprights, but doesn't do so well at electromechanical pianos (as others have mentioned, those devices are more like percussive single-strike instruments and may not benefit as much from modeling vs. sampling).

Due to who is behind Ivory, and their track record, I keep it and its spin-off libraries in my sights. If I had bought a new computer before Pianoteq came out, I doubtless would have bought Ivory, but as my computer is underspecced for running it, Pianoteq came at just the right time .

As the modeling vs. sampling discussion isn't likely to win any new converts at this point, probably the most useful comparison now will be between Ivory, Akoustik Piano (heavily tweaked since it stinks coming right out of the box), and the new Steinway from Garritan. I can't remember if they all sampled the same model or not.

If you're used to the Steinway sound but are doing some heavily orchestrated works that include a piano (but not as a soloed instrument), might I suggest trying a Boesendorfer sound for those purposes. You may or may not like it soloed, but I find it tends to sit better in an orchestral mix than a Steinway, in most cases. I would even say the same about a Yamaha, even though that too is not my preferred soloing piano.

[toodamnhip]

People say that hearing a real piano is different than hearing even the best piano sample, but they are comparing apples and oranges.* The real comparison is between the piano samples and a RECORDING of a real piano. The Ivory samples were very well-recorded. It would take some sophisticated mics, preamps, and a really good room to surpass Ivory. Not to mention a top-notch piano perfectly tuned. Factor in the number of takes and punch-ins, and then you're looking at a tremendous difference in cost, not to mention the possibility that Ivory's recorded samples may still sound better than your recorded piano.

Also, if you get Ivory's Italian Grand, it's going to up the bar even more. The thing to do would be to do a pair of test recordings where the client does not know which is which, play them for the client(s), and then let them choose the sound they prefer.

Shooshie

*a more accurate analogy is comparing apples with pictures of apples. In the end, it will be pictures of apples vs. pictures of apples, and you need to see whose pictures look more realistic.

Hey SHoosh..

I just thought of a vital need for qualification of your post and the discussions herein...

Like all things in life, "truth" is multi dimension and viewable from many angles...at least until we get to total truth..which is very very simple..perhaps even a singularity!
Anyway, in terms of normal physical universe truth, there is ONE thing a real piano does better even if recorded not as well a Ivory..

It PLAYS REAL!!!!...

You see my brotha, the truth of all of this is the controller heavily enters in, and how the controller interfaces with the sample library heavily enters in...Thus, one guy might have a controller work great with Ivory, and another guy might get a better performance with more feel and magic on..oh lets say a different controller and a Motif ES....And another guy might record a REAL piano 1/2 as well as the samples in IVory...but damn if that guy was effing wailing due to the feel of actually playing a real piano...So a great sample library can actually sap the life out of a performance that would have been better with the tactile feel of a real piano...thus, the inferiorly recorded real piano outshines the shiny Ivory..or whatever your sampled piano is.

I myself realize that my sucky M-Audio Keystation 88 does not live up to the quality of samples I have in Ivory..and that interfacing HUGELY affects the final performance and sound...It might even be nice if these companies told us what controllers they designed their software with so we can know this perfect controller to library match of which I speak...
But of course, that would make the software maker enemies with whichever controller they said "sucked" with their libraries..such as M Audio..lol

If we all had a standardized controller we could then truly compare sample libraries..but we don;t...so, we cant truly compare the libraries on an individual basis...

There are times I get a way better sound on my Motif, as compared to Ivory, because it seems to like my controller..or it's own keyboard...
This can even be a song by song basis where the lack of dynamics make a more un realistic, stagnant piano sit in the mix better,,

There are many angles to this equation if one truly takes into account the realities of the real world use of piano libraries..

Food for thought...

[Frodo]

In the meantime, Nick Phoenix has been kind enough to add two more demos of my performances on his EWQL website: Rachmaninoff's Liebesfreud on Steinway, and Beethoven's Hammerklaver Sonata, last movement, the tour de force fugue of Opus 106. I would be interested in your opinions and/or comments on these performances.

Cheers,

Joe

Hi Joe--

I know that your post was addressed to Shooshie, but I wanted to commend you first of all for choosing Beethoven's Op. 106. Only someone who really "knows" would earmark the "Hammerklavier" at all, and only the bravest would tackle it.

Masterful job. Thank you so much for that. I can see why Nick included it in the demos.

When I first heard the EW pianos, I thought they sounded nice but didn't make them a priority for purchase. Thanks to your performance I now have them on my "red list" of items to seriously consider.

Gratefully,
Frodo

[mhschmieder]

Well, a standardised controller is not an abstract concept; it is one that properly matches the MIDI spec and consistently provides all values between 0 and 127 for velocity, aftertouch, etc.

The ONLY weighted-action controller on the market that does this is the Studiologic VMKplus series (all of which have graded weighted action).

Escapement, weighting curves, etc., are subjective and can never be standardised. And they DO affect performance, but differently for each player. Depending on style of music and/or technique of the player, a controller such as M-Audio or CME that does not fully meet the MIDI spec may matter less than other factors.

But from the strict point of view of standardisation, which was your point, this is easily measured and proven, using MIDI monitor tools.

Anyway, I agree that we all listen differently and in a blindfold test we may focus on different aspects of the sound. I have yet to feel connected to any sample-based piano technology in a blindfold test, but routinely pick physically modeled (or hybrid) choices that aren't as up to snuff timbre-wise, because of the connectedness of the notes and the full range of expressivity.

Back to the standardisation issue, the sound source should be programmed with the assumption of an ideal MIDI controller such as the Studiologic VMKplus series, and I believe all libraries are programmed with this assumption. I find it unlikely that better results could be gotten with an underspecced controller, but adequate or acceptable results might be possible in spite of only every eighth or sixteenth MIDI velocity level being communicated by such a controller.

I'm not specifically promoting the Studiologic controller in this thread, as it has a fatal flaw in that it only is capable of one MIDI zone (rendering it virtually useless for controlling multiple rack modules), but am simply using it to illustrate a point about standardisation being defined by the MIDI spec itself.

In the realm of semi-weighted and unweighted controllers, there are far more choices for keyboards that properly transmit all MIDI values, but of course for acoustic piano it will be a rare individual that can muster a convincing performance from such a light action keybed.

[Shooshie]

Hello Shoosie,

Regarding my performances on the Authorized Steinway, I have been in email contact with Gary Garritan, offering my services as a demo performer. He replied that he enjoyed my playing on an mp3 of the EWQL Bosendorfer and Steinway that I sent him. Gary asked for a little more information from Yours Truly, which I promptly sent him.

The ball is in his court.

In the meantime, Nick Phoenix has been kind enough to add two more demos of my performances on his EWQL website: Rachmaninoff's Liebesfreud on Steinway, and Beethoven's Hammerklaver Sonata, last movement, the tour de force fugue of Opus 106. I would be interested in your opinions and/or comments on these performances.

Cheers,

Joe

Do you happen to have a link? I am not sure where to look for this. But I'd like to listen. Sorry I did not see this until today.

Shooshie

[Shooshie]

Well, a standardised controller is not an abstract concept; it is one that properly matches the MIDI spec and consistently provides all values between 0 and 127 for velocity, aftertouch, etc.

The ONLY weighted-action controller on the market that does this is the Studiologic VMKplus series (all of which have graded weighted action).

My Fatar (which is the real name behind the Studiologic line) does all this. I've specifically tested to see if I could get all values from 1 to 127 (I don't think I tried for 0), and I succeeded. Nevertheless, I've long since given up on a MIDI standardization, for it's just too subjective and will never happen. We don't even have to test it to prove that. A quick "proof" is readily at hand. Just look instead at the myriad concert pianists out there. You will find devoted fans for each of them. Those people rave about various facets of their playing -- tone control, pianissimo, dynamics, phrasing, musicality, and so forth. I hear all these superlatives about some of the lesser-known "greats," so I go listen to recordings of them, and watch videos of them on YouTube, and at least half the time I come away absolutely astonished that someone could find something to like about that pianist. I hear such widely off-the-mark comments about virtually any aspect of performance, technique, or even about the pianos themselves. From my own experience and listening, my disagreement will be so profound that I am at a loss for what to say to that person next time I talk to him or her.

So, I have to conclude either of two things:
1) most people in this world are just stupid
or
2) people's tastes evolve for complex reasons that I do not fully understand.

Of course, reason tells me that #2 is closer to the truth. Since tastes, then, can vary so dramatically, any hopes for a "MIDI Standardization" are out the window. This reasoning also is what causes me to back away from statements like The ONLY weighted-action controller on the market that does this is the Studiologic VMKplus series. Heck, one of the finest MIDI controllers ever created is the Yamaha Disklavier Grand. We tend to forget about it, because it costs so much. But I had the luxury of 16 years with SIX of them at my disposal. I did learn a lot!

Anyway, I know your post was well-intended, and I do not mean to come off non-supportive. I've just had to come to different conclusions about a lot of things, one of which is MIDI controllers. As with audio monitors, they become products of our tastes, technique, and familiarity. One adapts to such things. No amount of standardization will ever change that.


Shooshie

[Frodo]

Hello Shoosie,

Regarding my performances on the Authorized Steinway, I have been in email contact with Gary Garritan, offering my services as a demo performer. He replied that he enjoyed my playing on an mp3 of the EWQL Bosendorfer and Steinway that I sent him. Gary asked for a little more information from Yours Truly, which I promptly sent him.

The ball is in his court.

In the meantime, Nick Phoenix has been kind enough to add two more demos of my performances on his EWQL website: Rachmaninoff's Liebesfreud on Steinway, and Beethoven's Hammerklaver Sonata, last movement, the tour de force fugue of Opus 106. I would be interested in your opinions and/or comments on these performances.

Cheers,

Joe

Do you happen to have a link? I am not sure where to look for this. But I'd like to listen. Sorry I did not see this until today.

Shooshie

http://www.soundsonline.com/product.php ... tid=EW-171

"Hammerklavier" is under Bosendorfer 290 and "Liebesfreud" is under Steinway.

[Shooshie]

People say that hearing a real piano is different than hearing even the best piano sample, but they are comparing apples and oranges.* The real comparison is between the piano samples and a RECORDING of a real piano. The Ivory samples were very well-recorded. It would take some sophisticated mics, preamps, and a really good room to surpass Ivory. Not to mention a top-notch piano perfectly tuned. Factor in the number of takes and punch-ins, and then you're looking at a tremendous difference in cost, not to mention the possibility that Ivory's recorded samples may still sound better than your recorded piano.

Also, if you get Ivory's Italian Grand, it's going to up the bar even more. The thing to do would be to do a pair of test recordings where the client does not know which is which, play them for the client(s), and then let them choose the sound they prefer.

Shooshie

*a more accurate analogy is comparing apples with pictures of apples. In the end, it will be pictures of apples vs. pictures of apples, and you need to see whose pictures look more realistic.

Hey SHoosh..

I just thought of a vital need for qualification of your post and the discussions herein...

Like all things in life, "truth" is multi dimension and viewable from many angles...at least until we get to total truth..which is very very simple..perhaps even a singularity!
Anyway, in terms of normal physical universe truth, there is ONE thing a real piano does better even if recorded not as well a Ivory..

It PLAYS REAL!!!!...

You see my brotha, the truth of all of this is the controller heavily enters in, and how the controller interfaces with the sample library heavily enters in...Thus, one guy might have a controller work great with Ivory, and another guy might get a better performance with more feel and magic on..oh lets say a different controller and a Motif ES....And another guy might record a REAL piano 1/2 as well as the samples in IVory...but damn if that guy was effing wailing due to the feel of actually playing a real piano...So a great sample library can actually sap the life out of a performance that would have been better with the tactile feel of a real piano...thus, the inferiorly recorded real piano outshines the shiny Ivory..or whatever your sampled piano is.

I myself realize that my sucky M-Audio Keystation 88 does not live up to the quality of samples I have in Ivory..and that interfacing HUGELY affects the final performance and sound...It might even be nice if these companies told us what controllers they designed their software with so we can know this perfect controller to library match of which I speak...
But of course, that would make the software maker enemies with whichever controller they said "sucked" with their libraries..such as M Audio..lol

If we all had a standardized controller we could then truly compare sample libraries..but we don;t...so, we cant truly compare the libraries on an individual basis...

There are times I get a way better sound on my Motif, as compared to Ivory, because it seems to like my controller..or it's own keyboard...
This can even be a song by song basis where the lack of dynamics make a more un realistic, stagnant piano sit in the mix better,,

There are many angles to this equation if one truly takes into account the realities of the real world use of piano libraries..

Food for thought...

I've done an awfully lot of comparisons between real world pianos and VI, because for 16 years I arranged and composed for 5 Yamaha Disklavier Grand pianos. We had 6 at our disposal. (one was a spare which we rotated in and out of service) I actually learned a lot about piano during that time, and I was a piano aficionado beforehand, so my job was like turning me loose in a playground of my favorite toys. The result? I have dozens of CD-quality recordings of piano music for which I also have the MIDI of that same performance. This means I have the ability to match VI's to the real thing using a standard common-denominator: the Disklavier MIDI keyboard.

I've learned a great deal about doing that. The single most important thing that I've learned -- and this applies to all instruments, real or VI, in all situations -- is that we adjust to what our ears tell us. If a VI does not play like the "real" piano, we adjust. If our audio monitors don't sound like the real thing, we adjust both the audio file and we adjust our hearing.

There's a lot to all this, but in the end, I think we have to concede that there will never be any standard. At all times, we adjust to what we're playing or hearing, and our tastes change. One day we might like the bright piano -- for Mozart of Prokofiev. The next day we might like the dark piano for Rachmaninoff or Chopin. There just isn't a one-size-fits-all instrument. Not real, not virtual.

Shooshie

[Frodo]

I have to step back from all of this and consider that the day I chose to invest time and energy into a computer and a DAW for the sake of replacing anything other than a multitrack tape deck, the horse was out of the barn.

Real instruments can be substituted, but perhaps they cannot be entirely replaced with 100% confidence. Heck, they my not even make the 50% mark, if that.

As I see it, musical technology is just reaching its adolescence. So much of what was out there 10 years ago sounded bad-- plain and simple. The rest was unaffordable if it sounded above the status quo at all.

The trade-offs are obvious, but we already *know* what those trade-offs are with virtual instruments even as we embrace the benefits (or evils, depending upon one's perspective). Virtual Reality remains closer to being "virtual" than it accomplishes the tall order of fully competing with "reality", sonically speaking. The best of what's out there now is tons better at a relatively affordable price than could ever be thought of a decade ago. Major libraries once going well into the five figures are now old hat for the way they sound. Most of them couldn't even be given away for free today.

Personally, I appreciate where VI technology is in its own right, considering its history.

As for real pianos, most people have no idea how much TLC such an instrument requires. It opens up a whole different ball of wax that most people don't even want to get into. There really is no comparison-- or it's nearly futile to do so.

[toodamnhip]

I have to step back from all of this and consider that the day I chose to invest time and energy into a computer and a DAW for the sake of replacing anything other than a multitrack tape deck, the horse was out of the barn.

Real instruments can be substituted, but perhaps they cannot be entirely replaced with 100% confidence. Heck, they my not even make the 50% mark, if that.

As I see it, musical technology is just reaching its adolescence. So much of what was out there 10 years ago sounded bad-- plain and simple. The rest was unaffordable if it sounded above the status quo at all.

The trade-offs are obvious, but we already *know* what those trade-offs are with virtual instruments even as we embrace the benefits (or evils, depending upon one's perspective). Virtual Reality remains closer to being "virtual" than it accomplishes the tall order of fully competing with "reality", sonically speaking. The best of what's out there now is tons better at a relatively affordable price than could ever be thought of a decade ago. Major libraries once going well into the five figures are now old hat for the way they sound. Most of them couldn't even be given away for free today.

Personally, I appreciate where VI technology is in its own right, considering its history.

As for real pianos, most people have no idea how much TLC such an instrument requires. It opens up a whole different ball of wax that most people don't even want to get into. There really is no comparison-- or it's nearly futile to do so.

Add to that the space the piano is in...

Sometimes the space becomes part of the Piano..

There's a piano in studio C or B at westlake in Santa Monica that I composed one of my favorite pieces on while on break...

Why? The room is magic..the room plus piano is INCREDIBLE magic...

NO TLC could change that if you removed the piano from that little corner it would not be 1/2 as nice...lol