Piano Sample Libraries - pedal, sympathetic resonance etc. in 2012...

[Douglas Whates]

RE: convolution. How would one go about making an impulse response of the sustain? That's where I'm a bit confused. I've researched it a bit and I just can't work out how you would capture an impulse response resonance of the piano with dampers up without also capturing the room reverb.

That's a job for an anechoic chamber. Some sample sets for hardware pianos are recorded in them - Kawai used to have photos of the recording for the MP9x00 series where they claimed to have done this. And welcome back Douglas - still a fan of your library.

Ah, I wondered it it may have been something like that. And thanks for the "welcome back"

[Douglas Whates]

So, anyone know the ins and outs of how Galaxy II resonances work? I guess they must have sampled resonances somehow? Maybe an e-bow as Greg suggested (great idea!), or...? I'm quite sure they don't have separate dampers up samples, so it must be that all sustain-pedal stuff is done done by scripting and triggering the appropriate resonances.

[AKJ]

I have no idea how they do it, but imo it works great. for me galaxy II is probably all I'll ever need in terms of piano. the tweaking possibilities are really deep.

[Douglas Whates]

I have no idea how they do it, but imo it works great. for me galaxy II is probably all I'll ever need in terms of piano. the tweaking possibilities are really deep.

Yes, it's definitely one of the better libraries. It would be nice to know what's going on behind the scenes. I'm really curious to know how the resonances and sustain-pedal stuff is implemented.

EDIT: also, just thinking about the impulse response stuff. If I wanted to capture an impulse response of only the effect of the dampers-up, could I capture two impulse responses - one with the pedal-up on the piano and one with the pedal-down - invert the phase on one of those, which would then leave me with only the difference between the two, which would be the effect of the sustain. Does that make sense and would that work in practice?

[Gregjazz]

also, just thinking about the impulse response stuff. If I wanted to capture an impulse response of only the effect of the dampers-up, could I capture two impulse responses - one with the pedal-up on the piano and one with the pedal-down - invert the phase on one of those, which would then leave me with only the difference between the two, which would be the effect of the sustain. Does that make sense and would that work in practice?

That seems reasonable to me, though in that case it's imperative that you use a static sound source to create the impulse. For example, running a sine sweep (or noise burst) through a flat speaker. That way you can recreate the exact impulse conditions.

[Douglas Whates]

also, just thinking about the impulse response stuff...

That seems reasonable to me, though in that case it's imperative that you use a static sound source to create the impulse. For example, running a sine sweep (or noise burst) through a flat speaker. That way you can recreate the exact impulse conditions.

It would be great if it works in practice, right? From what I've read, a long sine wave sweep is the best solution for the lowest noise. Hmm, I'll definitely have to experiment with that idea. Where best to place the speaker(s)? Any ideas?

[Gregjazz]

In terms of placement, that would take some experimentation. Whichever placement results in the most string excitement (and as evenly exciting all the strings when the damper pedal is down as possible).

[Douglas Whates]

In terms of placement, that would take some experimentation. Whichever placement results in the most string excitement (and as evenly exciting all the strings when the damper pedal is down as possible).

Yes, I'm going to experiment with it. Should be very interesting to see how it works out.

With regards to sympathetic string resonance for non-sustain pedal stuff, I'm still wondering how best to sample that. For integration with the rest of the sound, I think it's definitely necessary to *sample* it somehow (rather than create it synthetically), but I'm trying to work out the best way to do that. I'm not sure if the e-bow would work afterall, as I would need to be able to capture different partials, and wouldn't an e-bow just give me the fundamental of any one string?

[meldavid]

Just interested to you know how you rate Roland's take on physical modelling the piano. eg. the V-Piano or 700NX?

Sounds pretty convincing to me, but waaaaaay out of my budget.

[egbert]

With guitars and similar instruments with a handful of strings over a bridge (eg double bass, cello, violin, mandolin etc). There are methods where the response of the whole instrument to an impulse are recorded - you might excite the instrument to resonate by tapping the bridge with the end of a small ring spanner or similar and recording the response using a mic and (if fitted) a piezo pickup simultaneously.

There was a plugin from a scanadanavian developer - name slips my mind just now - that used a collection of these sorts of impulses and a convolution engine to apply the body resonance to piezo pickup based instruments.

This sort of approach might work with a piano but the bridge is several feet long and has 88 sets of strings over it. It might be that the right sort of blow on the soundboard from the bottom might work - eg using a metre rule or something of the right length to get the whole system in motion simultaneously. Assuming that worked, doing that with the sustain pedal down would get you the response for pedal down and doing it with the pedal up would get you the reponse with just the undamped upper octaves.

You might need to do additional examples when other notes or chords were held down to capture the results when a chord is played. Anyway - just some ideas. Getting hold of the developer of the plugin I mentioned - who definitely knew about all of this - would probably be a big help.

[Douglas Whates]

Just interested to you know how you rate Roland's take on physical modelling the piano. eg. the V-Piano or 700NX?

Sounds pretty convincing to me, but waaaaaay out of my budget.

The VPpiano is the best synthesized piano I've heard. I haven't played it, so I couldn't comment on that. However, the Achilles heel (as with all synthesized pianos): when you hear demos of the dry piano (i.e. the direct output with no reverb), it is not nearly as convincing. It just doesn't sound like a real instrument in a real space. The beauty of the best sampled instruments is that what you are getting is a real sound recorded in a real place.

[fas1piano]

hi douglas, welcome back! yes ns_kit7 is excellent, it's just my fault that it doesn't get used as much as i should since i have barely time to practice.

i agree on the value of a sampled piano.

i really enjoy playing Alicia's Keys, sampled by Thomas Skarbye. i think the resonance there is excellent, you can dial in just as much as you want and also control the amount of voices used for that. in addition, there is convolution used for reverb but not for resonance AFAIK. i think, pedal-down is another matter, there are separate samples for that.

your project sounds good. I like AK, but having another sound/timbre would always be welcome.

just make sure to test it in all kinds of piano players' real world scenarios, e.g.

- for live use on a laptop with restricted power (something like a core duo or core2duo generation, lots of people don't replace their laptops all the time) and provide a way to tame the instrument for that.

- in a band setting (e.g. i lkie Virtual Grand Piano but in the band, it's too much recorded ambience, so the sound is indirect!)

- with live speakers (lots of sounds get hard and screechy even on good PA speakers)

- in mono

- and with different types of master keyboards!

Because ultimately there is tons of great giant studio pianos (quantum leap and what not) but very few that piano players actually like and use everyday, i think that's for the reasons just described.

[Lode_Runner]

Check out the following (if you haven't already):

Imperfect Samples Steinway and Sons 1908 Walnut Concert Grand
http://www.imperfectsamples.com/website ... /demos.php

8Dio Legacy 1928 Steinway Grand
http://8dio.com/?btp_product=legacy-1928-grand-piano

Cinesamples Piano in Blue (another Steinway - same one Bill Evans played on Miles Davis's Kind of Blue)
http://cinesamples.com/products/piano-in-blue/

[kbaccki]

Check out the following (if you haven't already):

Imperfect Samples Steinway and Sons 1908 Walnut Concert Grand
http://www.imperfectsamples.com/website ... /demos.php

8Dio Legacy 1928 Steinway Grand
http://8dio.com/?btp_product=legacy-1928-grand-piano

Cinesamples Piano in Blue (another Steinway - same one Bill Evans played on Miles Davis's Kind of Blue)
http://cinesamples.com/products/piano-in-blue/

Also:

http://www.soundiron.com/instruments/pi ... onal-piano ... doesn't do anything specific w/ resonance, but it's a nice sounding sampled piano...

http://acousticsamples.net/kawai-ex-pro http://acousticsamples.net/keys/oldblackgrand

The acousticsamples pianos are fantastic, but (unfortunately) no longer available in Kontakt format... UVI only... but props to them anyway, they sound fantastic... I think OBG is my favorite character piano.

And of course, on the "modeled" side... http://www.truepianos.com/demos.php

I believe TruePianos uses sampled attacks and synthesizes the sustains and resonance. It's not perfect by any means, but I think it doesn't suffer from the plasticy sound of Pianoteq (which IMO has been there since v1). Many people swear by Pianoteq, but to me the attacks and sustain in the mid to upper range just don't sound right... In any event, TP is nice to have around at least as a lightweight (both CPU and memory) practice module...

[synchronizer]

I have the Galaxy Vintage D (an extra to the Galaxy Pianos line), and I can say that I love how it possesses a certain warmth and attitude that other sampled pianos lack.

I hear that Ivory is good as well, but I'm rather pleased with the Vintage D.