Interesting new modelled piano plugin

[Fogwall]

Concerning release velocity... We intended to implement it but did not as we had never heard of digital keyboards sending this information.

 

However, we will implement this feature if such keyboards will appear. Of course it would be a great improvement to benefit of partial dampers, just like partial pedals.

 

Regards,

Niclas Fogwall

Sales & support

Pianoteq

www.pianoteq.com

[bg]

Kurzweil K2x00, PC1, and PC2 keyboards transmit release velocity.

[mate stubb]

Rhodes Chroma from the 1980's transmits release velocity.

[bg]

Also, Alesis Ion, Micron, and Andromeda transmit RV.

[bg]

..and the Nord G2. The ones I've listed are all currently manufactured.

[mate stubb]

Originally posted by Fogwall:

However, we will implement this feature if such keyboards will appear. Of course it would be a great improvement to benefit of partial dampers, just like partial pedals.

Niclas,

 

I'm curious as to how release velocity would actually be used to control partial dampers. Damper effects is a positional thing, not velocity. If I was accurate enough, I could release the keys on a real piano to a certain depth every time but at different speeds - the transition time of the effect should be a function of the key depth, not how long I took to completely release it.

[bg]

..plus the Kurzweil SP series. Yes, I'll shut up now.

[Jazzwee]

Originally posted by mate_stubb:

Originally posted by Fogwall:

However, we will implement this feature if such keyboards will appear. Of course it would be a great improvement to benefit of partial dampers, just like partial pedals.

Niclas,

 

I'm curious as to how release velocity would actually be used to control partial dampers. Damper effects is a positional thing, not velocity. If I was accurate enough, I could release the keys on a real piano to a certain depth every time but at different speeds - the transition time of the effect should be a function of the key depth, not how long I took to completely release it.

Mate_stubb, if you read our discussions (and we discussed this on another thread also recently), the positional problem is very serious. All keyboards I've tried have one switch off position and that is at about 50% of the travel. So release velocity must be coupled with positional switching off. 50% of the travel is too low. A real piano does not stop the sound until all the way to the top. So this may require some key action redesign.

 

BTW - the biggest effect of this as an added feature is to improve everyone's legato.

[Calumet]

Can anyone confirm or deny that this works on Intel Macs? I couldn't find an exact answer on their site.

 

Thanks.

[Jazzwee]

Site said Intel Mac version is in development.

[Jazzed]

Originally posted by Jazzwee:

Site said Intel Mac version is in development.

Which it states on the bottom of the download page. Should be more obviously stated, perhaps in the FAQ?

[mate stubb]

Originally posted by Jazzwee:

A real piano does not stop the sound until all the way to the top. So this may require some key action redesign.

 

BTW - the biggest effect of this as an added feature is to improve everyone's legato.

That's what is great about this modelled approach - the amount of detail possible is going to cause the boundaries of what we have settled for to be pushed.

[Jazzwee]

I was not a big model believer (at least of piano) before this. This is exciting.

 

Looking at people's comments about this, it seems to be that they need to model more "imperfections" to give it that real world quality. Maybe some of this imperfections are not yet considered in the model.

[ZioGuido]

I haven't read the whole thread but somewhere I read you were talking about note-off velocity capable keyboards...

Well.. the CME UF series do have Note-Off velocity. I own a UF8 and it sends Note-Off velocity (I can see it in MidiOX) and the only plugin I have which supports this kind of midi message is HALion (I still have old version 1.1), infact it's interesting to see how it reacts when using the DustyElectric Rhodes sampleset.

 

Unfortunately Pianoteq seems to ignore this parameter.

 

Btw, Pianoteq is awesome!

[Jazz+]

Need imperfections?

 

Try this:

 

"Try changing hammer hardness (voicing section), combine it with another room ambiance such as the "recording" preset (you may adjust the amount of dryness and spaciousness), the dynamics or velocity curve might also need some adjustments as well. Global resonance can be increased as well as unison width (making the piano slightly out of tune, giving it more acoustic resemblance). All of these changes can

make drastic improvements in the character of the piano sound. "

[Jim Alfredson]

Wow. I'm speechless. That really sounds amazing. I didn't think anyone would be able to do it.

 

I think I found my next plug-in.

[Steve Nathan]

Hopefully the sound of this thing will improve greatly given some time. The artifacts and resonances are impressive, but the basic tone sounds more like a CP70 than a C7, especially in the middle. I'll keep my fingers crossed.

Steve

[Jazz+]

To what extent are samples involved in the Pianotek? Are they using note samples and then applying physical modeling like GEM and Kawai do?

[burningbusch]

Originally posted by Jazz+:

To what extent are samples involved in the Pianotek? Are they using note samples and then applying physical modeling like GEM and Kawai do?

According the PianoTek press release:

 

"Being a truly modelled piano, the sound is

created in real time from scratch through a

very sophisticated mathematical model,

simulating an acoustic piano. It is responding

to how the pianist is striking each key and

how strings are interacting, just like a real

piano does. The result is a remarkably

expressive and vivid instrument, placing it

way beyond the state of the art."

 

No samples, with the exception of a few optional acoustic noises.

 

Busch.

 

Busch.

[vihreamies]

I love this idea of modelling, because it makes possible to create and tune your own kind of sound: this is something that I could use as a tool for creative work.

 

Keep up the good work Pianoteq!

 

-vm

[polysix]

I just installed the pianoteq and I use it with

energyXT - to me it sounds great, plays well and I think the sound is better then my FantomX pianos,

I certainly love the edit features

 

I tried it on top of other synths like a layer,

sounds great too

the demo prog is available only shortly

I wish those guys all the success,

they earn it

[marino]

OK, now that my high speed connection seems to finally work as it should (cross fingers), I have downloaded all the demos, transferred them on audio CD, and listened thru a good system.

 

While the overall impression was still good, I noticed a definite lack of character in the attack portion of the sound. By "the attack portion" I don't mean just the initial transients, but also the very first decay between the intial 'noisy' click and the beginning of the sustain portion - let's say, the first 300 ms of sound. This is valid for all registers.

In short, I kept feeling the urge to substitute those attacks with sampled ones! I know, of course, that the dynamic response wouldn't be so smooth, but I felt the need for a more pronounced beginning of each note. Also, just layering sampled attacks wouldn't work, because it would create every sort of chorused/phased artifacts; I was thinking more of volume-crossfading the sampled attacks with the modeled sustains.

 

Please note that I wasn't looking for a rock'n roll sound; rather, a sound suitable for solo piano. I listened to a few classical and jazz recording, to confirm that my ears weren't fooling me: Those recordings, though soft and dark in character, had lots more 'personality' in the attack.

 

I have no idea, of course, if tweaking the parameters could achieve similar results. I'll try to play with the demo as soon as I can.

 

I like the sustain portion of the sound quite a lot. It could still benefit from a certain increased 'unpredictability' in the inharmonicity of the upper partials (again - maybe this could be achieved by tweaking), but overall, it sounds quite natural and satisfactory to me.

[Jazzwee]

Busch, you have Ivory right? When you get the chance, could you give us a comparison against Pianoteq (after your best tweak)?

 

This is really a battle between Pianoteq and Ivory.

[subtlearts]

Hi. Thought I'd drop my two cents here, though I'm new to these boards. i just posted this over at KVR and I thought it might be interesting here, though I can't offer a proper direct comparison as asked for above. I am also most definitely not associated with Modartt (or any other music software manufacturer) in any way. Before someone asks.

 

Anyway, I spent most of the day today mucking about with the Pianoteq demo, and comparing it to a number of sample sets, both those I own and some I can only hear web demos of. Reason being that I needed to decide whether to spring for the NI anniversary special on Akoustik Piano before it expired. So I was checking out the competition and stumbled, to my very pleasant surprise, on Pianoteq. I doubt I am the only person who has wondered for, well, years, when a real, fully modelled piano plugin or module would appear. Never been happy with static samples 'cause I spend too much time on real pianos and my fingers and my ears know the difference all too well. I used the GeneralMusic RealPiano module for several years just because it was the only thing on the market that did something approximating sympathetic string response; chords sounded vaguely like chords.

 

Background: I am a professional jazz pianist of long standing, studied at McGill in Montreal, have released a number of solo piano albums and made my living this way, entirely, for getting on towards twenty years (played for over thirty). As a composer, piano figures pretty heavily in most of the music I make. I also own two grands, have apprenticed as a technician and have tuned well over a thousand pianos. So I feel relatively qualified to comment on pianistic realism. I have also played 'artificial' pianos of every conceivable pedigree, mostly without much satisfaction (except for real electromechanicals which are a different story of course).

 

So. I was playing the Pianoteq from a weighted-action controller (Fatar Studiologic) and monitoring through an E-Mu 1616m and good studio headphones. I was less than overwhelmed by the sound at first, especially the high end, which seemed tinny and thin... but,as I tried some of the performance things that have always drawn out the weaknesses in sample-based plugs and modules, it began to grow on me.

 

Then I started to tweak the finer details of tuning, voicing and design, and I began to be more and more intrigued and impressed. Within an hour I had created a number of different settings which I felt significantly surpassed the included presets in terms of realism and 'performance enjoyment'. Not to say they're objectively better, but they worked better for me. And, quite honestly, they work better for me overall, as a total playing and listening experience, than any sampled piano I have ever played. And that's after an hour or so with the demo.

 

Now, I should qualify that by saying I have never played Ivory, which I hear nothing but raves about, and I will admit that I have been very impressed with the samples I have heard. To my ears, at least judging by the online samples, it has more warmth and character than any other sample set, including Akoustik Piano, and quite likely more than Pianoteq in its current form, at least 'out of the box'.

 

However the more I listened to the Ivory samples, the more its weaknesses stuck out at me - a kind of unrealistic hype in the bass, the static sound of repeated notes, the 'lifeless' chords - whereas the more I played and tweaked Pianoteq, the more I enjoyed playing and tweaking it. Then I started visualizing the possibility of creating exactly the right piano sound for every track - and not with EQ and effects but with the tools I would use if I were tuning and voicing a real instrument (and had all the time in the world) - hammer hardness at different strike weights, unison width, octave stretch... and, of course, with possibilites well beyond those of a single instrument - control over instrument length, soundboard response, harmonic character, damping... and I started to get very excited indeed.

 

Do I think Pianoteq is perfect, and do I prefer it to my real grand pianos? no, obviously not. When possible or practical I will always try to record the real thing first. Do I think the best sample sets going today likely surpass it 'out of the box'? Yes, though they too reveal their weaknesses with close listening, and then there's not much you can do about it. However, given what has already been achieved with this 'first out of the starting gate' version, and assuming the same level of detail and finesse goes into further development of Pianoteq... I cannot really imagine spending a large bundle on a static sampled instrument at this point, no matter how rich and thick and chocolatey it sounds out of the box, or how many gigabytes and velocity layers and release samples... It would just be too limiting.

 

In the end, the more I listen to static samples the more plastic and lifeless they sound; so far, the more I get into Pianoteq the more I like it and see its potential and want to play and tweak and explore it. Kind of like a real instrument. There, I've said it!

[kanker.]

Originally posted by Steve Nathan:

Hopefully the sound of this thing will improve greatly given some time. The artifacts and resonances are impressive, but the basic tone sounds more like a CP70 than a C7, especially in the middle. I'll keep my fingers crossed.

Steve

Agreed. This is a very impressive piece of software, but it does sound more like a CP70 in the attack to me, and more like a transistor piano in the sustain. It sounds very artificial very quickly. That said, I never thought I would ever hear a modeled piano that sounds anywhere near this good that would run on my 1ghz G4 laptop. Now that I have a MacBook Pro, I'll be interested to check out the UB version once it's ready, but this things still has a LONG way to go sonically before it's ready for prime time. Still, very impressive.

[Jim Alfredson]

Tobias,

 

Nice post and beautiful music on your website. I think this technology is very exciting and I will be buying this software soon. It can only get better.

[Joe Muscara]

Has anyone played with version 2 of this plug? I just installed the demo of it on my MacBook (it's now Universal) and it seems pretty good. I have Ivory, but my ear isn't as good as some of you as far as the qualities of real grand pianos that I should be listening for.

 

I did notice some things.

 

1. At the lightest touch I can muster on my RS4073 (controller), Ivory was much louder than Pianoteq, after getting them both about the same max volume. Neither was perfectly silent, but Pianoteq was close. This is using the default piano.

2. Unlike Ivory, Pianoteq allowed me to create string resonance by quietly holding down the higher octave key and striking the lower octave note hard. Releasing the higher key muted the resonating "string" with accompanying hammer noise.

 

I think there was something else, but I can't remember now.

[kanker.]

Version 2 is a huge leap forward from V1. It no longer sounds like a transistor piano. By V3, I think they'll pretty much have this stuff nailed. I'm very impressed.

[Mark Schmieder]

Version 2 is a huge leap forward, and even the soon-to-be-released point release will be another big step.

 

I am so glad that I held off on buying Ivory. having recently played a friend's GEM PRP800, all I need to decide now is whether I still need the GEM for the road, or a Muse Receptor :-).

 

Pianoteq was Receptorised quite recently, and apparently works quite well in that environment.

[Joe Muscara]

The latest issue of Keyboard (6/2007) arrived in the mail yesterday and Jim Aikin gave it a fairly positive review. He did some comparisons to Ivory. (He also reviewed the Italian Grand Expansion Pack.)