Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Pianoteq 6!

Painted faced football fans shouting fight songs.

Shish Boom Blah!

I don't think it is necessary (nor courteous) to refer to any particular individual(s) specifically.

That's nice. But what's the answer to my question? In my reading of this thread only one person (pro PianoTeq as it happens) has strayed just slightly towards aggression or fierceness. But who do you think has been aggressive or fierce?

If I had to choose only one piano VST. It would be Pianoteq without question or hesitation.

I read intently criticisms of Pianoteq on this forum, and I personally find them useful in that I try to get a sense of what is being reported on. Since I own sampled VSTs that are highly thought of, CFX, Ivory II, ACD, Ravenscroft, I would play them, and it always ends up the same, I always go back to the Pianoteq D4 in particular.

The sampled VSTs I own could not be more different from each other, so Pianoteq to me is just another piano. The way how my setup is I do forget whether it's Ivory II German D or the Pianoteq D4 I left it on, I can't tell the difference unless I check.

It is interesting, and laudable that Bluthner, Grotrian and now Steinway presumably thinks highly enough of the Pianoteq models to put their imprimatur on them. However I did not need the Steinway stamp of approval to enjoy using and choosing the Pianoteq D4, which since I ugraded is now the Steinway D.

That's how I feel as well, in no small part because of Pianoteq's flexibility and the detailed configurability allowed for and implemented in Pianoteq's modeled piano approach as contrasted with sampled piano libraries, and its high quality, especially considering its modest demands upon computer resources.

Fiddle faddle.
There's been nothing fierce here.
And your psychiatric viewpoint is suspect. That's just silly.

Of course I don't wish to discourage anyone's freedom of expression, and I'm happy to see that in your case at least, I was successful. And after commenting about the discussion being diverted away from the primary topic, I don't want to contribute too much to that very problem myself.

I think a new major revision always merits scrutiny and re-evaluation, and some folks won't like it (or will continue to not like it). Who cares? Some people seem to perpetually hate on the Garritan CFX I really like, but that's fine too. Nothing is perfect (yet), but if it brings more joy and music into your life, it probably shouldn't matter what other people say.

After playing with the update myself for a day I have to admit, I'm with Phil.

https://youtu.be/c_7_v9gsvsA

Personal preferences aside,

which other company or producer of digital pianos does publish a straight comparison on their website,
where you can directly and comfortably compare the differences of an "old" against the new product.
IMO that is speaking for itself.

I guess the "sampling industry" with their oldish ROMplers is in uproar about the serious competition arising from a small company in France!

Roland went fully modeled but as far as I can see people are still buying Kawai and Yamaha pianos. You should realize not everyone is satisfied with the current state of modeling. Each to their own. Modeling isn’t there yet to replace sampling entirely. Maybe that will happen some day and I’ll be more than happy if I have an entirely modeled piano that can bring me as much joy as the best sampled ones today, so I am expecting it eagerly.

It'd be interesting to see if those on the real piano forum have fierce opinions concerning their instruments . . . .I read of one who got tired of his Kawai grand, and another who intensely disliked a Steinway . . . .
Maybe we all need our hearing standardised, regulated, cleaned up and the batteries replaced.

Indeed, bass from Rolands is still boomy and boxy, while the highs and mids are best in class of all pianos. However, modeled pianos have some kind of strange dullness in them - it's not the fact being them dark like German pianos or bright as Fazioli, but they can be dull and metallic/bit harsh together. Totally opposite of what real piano attributes are - even if they are quite bright, there is not at all metallic sound, harsh, or wiped out tonality richness, which is still normal thing in modeled pianos.

I think there are still years to come unless modeling will be real competition. I was listening to V-grand few days ago, and while Roland get rid of metallic characteristic, so the bass response is worse. But, in the end, it all depends, in many areas they achieved great development, but in the same time in few it's a step back, or the best, sideways. Hopefully in the next generations they will be able to slightly improve those things.

And, finally, all pianos sounds different in real conditions, and while for Example Casio GP sound really good and Kawai SK-EX a bit dead via youtube, in real life it's the opposite. Casio sounds good only with the lid open, and Kawai is like real compared to Casio, tested CA67 and 97 vs GP400.

Wooden pianos also have limitations, although their supporters/owners usually describe that as "unique character".
They claim to love it, "warts and all".

To some extent you can emulate the imperfections of wooden pianos with the condition slider in pianoteq.
Several users have commented that they PREFER a little bit of degrade in that setting to better represent their wooden pianos that have settled a week or month after tuning.

It is all too easy to underestimate the pace of progress in computing - autonomous vehicles being a good example.

There is SO MUCH subjectivity in all this, bias/prejudice too

Modelled piano's , sighhh,,, have tried them all and own several modelled software AP , including PT.

Problem is in the attack. Always in the attack (transient).
The attack has something dull /thin / unnatural / too much rounded quality to it. The rest I LOVE ; dynamics , control , sustain , resonance , intonation on modelled sounds - alle very enjoyable ! But the initial transient / attack has always been why I keep on going back and forth and back and forth between sampled and modelled.

I don't really understand why they don't simple put a very small /tiny sample at the start off the sound, just neatly in the initial attack and be done with it.
I think it woud solve so much of the always noticable 'modelled' character of the sound, that it would become instantly perfect instead of almost there, but still something is 'wrong'.

At first touch , the new PT piano's always impress en are very nice too play, but along the line this little creature in the back of your head keeps on telling you " why does it work so well and sound basically so good and STILL keeps on bothering me as if something is not right and becomes annoying in the long run".

This is not bashing PT - it's really good and plays nice , but if I have t pinpoint where the underlying problem with long term use starts to creep up , its that tiny little transient part of the sound. IMHO I think that the way that start is modelled / calculated and the algorithms used for that , have not dramatically changed over the years . Neither at Roland , Arturia or whatever other modelling attempt. So I would suggest - either develop or try a complete different approach for the attack modelling , or (perhaps easier) throw in a tiny attack sample, mary that with modelling and be done with it. (BTW isn't that what the 'old' supernatural engine of Roland also attempted , although that had its own flaws here and there ?)

---
Many of the comments and complaints in this discussion against modeled pianos in general and Pianoteq specifically, seem to be from people who have little or no experience with the benefits and advantages of the extensive control one is offerred of nearly every aspect of the sound, from adjusting the inharmonicity of the bass strings, changing string length (Pianoteq allows one to create a modeled piano with strings much longer than the largest concert grand), hammer hardness and condition, damper response, the relative strength of the partials to the fundamental tone, the strength and brightness of the higher frequencies of each string, how fast they decay, the duration and prominence of the various components of the sound produced by each string-strike, equalization applied to each string individually rather than merely an overall equalization of the composite, combined sound output, customization of sound and model parameters applied variably or in graded form to the entire keyboard or only to specific ranges of the keyboard or even on a note-by-note basis (in Pianoteq Pro), and numerous other aspects of the sound that can be customized in fine detail to suit one's particular tastes.

If these people are satisfied with their sampled pianos, that's fine. For others of us, the limitations, idiosyncracies and deficiencies of sampled pianos make them less desirable.

I've read somewhere that Pianoteq uses samples of attacks because it is too difficult to model and calculate them in real time. Not sure where I read it so it might not be true.

I am not current, but it was my understanding a couple or three years ago that an "attack sample" is typically added as a preface to modeled sound.
From memory (unreliable at best) this was on Roland's offering.

ONE of my reservations with sampling is that it is done on ONE SAMPLE of the wooden piano, which may or may not be a GOOD sample of that maker's work, may or may not be the best maintained sample, but is an available sample, it may not be the piano you would buy in a store, not necessarily better (however measured) than you could buy.
Steinways vary a LOT, buyers typically visit a showroom to make their choice of what appear to be the best on the floor that day (or "least bad" if you are a cynic).

Another is the extraordinary amount of time it must take to gather all the samples at all the velocities and record them with .....all the artifacts of the recording process including whatever particular room/hall.
All the time it is going out of tune, or a tech/tuner is standing by to tweak things.

I don't think interpolation is such a HUGE problem as some may imagine.
Things are almost certainly not linear between velocity levels, but some clever curve fitting can probably get very very close to where the wooden piano would behave at all the in between points.

I know a little about post processing, i.e. the attempts to get rid of the room and produce "dry" samples so that the user can add room/hall ambiance - here we are REALLY getting into "synthetic sound".

So yes, I mostly reject the "realism" claims of sampling.

OTOH, when in the right mood I can enjoy trying to make music on almost anything in almost any state of tune
Music isn't ABOUT the instrument.
The archer, not the arrow. The singer, not the song, etc.

This thread has been nothing but FANTASTIC for Pianoteq.

1) They've had great exposure for v.6
2) They've had constructive criticism in a number of aspects from experienced VST users
3) Everybody is talking about modelling vs. sampling, which means those who don't own one or the other are now curious.

It's clear that people have very strong views about piano modelling vs. sampling.

What worries me is that everybody using these VSTs has their own system, settings, headphones, audio-equipment etc.
Perhaps as a community, we should start threads sharing Pianoteq settings etc., and start to talk about how to make PT and other VSTs function optimally.

Also, I appreciate the various strong view points etc.; however, for those who are confident players, it would be more useful to compare/contrast actual recordings of various VSTs in a dedicated thread, with each performance showing the settings used etc.

Modartt has always been up-front that the mechanical noises of the piano use samples, while all of the strings, sound board, and resonant aspects of the piano are modelled:

http://www.forum-pianoteq.com/viewtopic.php?pid=12240#p12240