Pianoteq 5 impressions (Page 3 of 7)

Thanks very much!

Thanks! It's true. I do have two V-Pianos. I love them. To be honest though, I much prefer Pianoteq's internal sounds. But the V-Piano is so nice to play. It feels amazing.

I've just spent another day with V5 and I think it's absolutely brilliant. In how many ways - well I can
name quite a few.

First - the "brightness" and "metallic overtone" issues. Does V5 generally sound brighter and perhaps more
"metallic" to me? Yes, and thank God! A piano is metal strings, after all. I've spent so much time with previous
versions trying to get brighter and more metallic sounding pianos. This is exactly what I want, I do a lot of
pop and rock production and this sound is what is called for.

But I'll go further - as a recording engineer, I've spent a lot of time brightening up and compressing piano tracks
for pop and rock productions. A LOT of time. As a matter of fact, that's just about job number one in mixing a
pop or rock track with a piano - getting the piano to stand out and cut and stab. Ask any engineer. A piano that
is naturally bright and percussive? That's what works. Here's a very short list of songs that feature (if you listen closely) insanely bright pianos -

Bohemian Rhapsody (Queen)
Death on Two Legs (Queen)
Somebody to Love (Queen)
Pinball Wizard (Elton John)
Burn Down the Mission (Elton John)
Stay the Night (Chicago)
Hello (Lionel Ritchie)
St Elmo's Fire (David Foster)
New Toy (Lena Lovitch featuring Thomas Dolby)
Right Now (Van Halen)
Baba O'Riley (The Who)

This doesn't mean I can't get mellower, smoother, darker piano sounds out of V5 - totally can. Anyone can. Because here's a rule of audio engineering - you can always "cut" what IS there, but you can't "add" what ISN'T
there. Want to make your V5 piano smoother, rounder, darker, less metallic? Start with adjusting the harmonics,
move on to the equalizer, shorten some of the sliders, move the strike point, lower the hammer hardness...in short, get to work programming it. That's the strength of Pianoteq - you can "make" your dream piano. You're not stuck with somebody's samples.

In past, there have been some user complaints about lack of punch in the middle octaves of Pianoteq pianos. This
concern went away largely with the release of V4, and I'd say that now with V5 released we can forget about ever hearing this complaint again. The middle octaves feel solid, punchy, clear and present on all models in V5 (to my
ears).

There has also been some concern expressed about over-emphasized L-modes in certain models. I think to some degree this comes from people who don't have a real piano sitting right next to their mixer chair in their studio.
I do have a real piano (Yamaha C3 Midi Grand) sitting two feet behind my mixing desk chair and let me tell you,
that instrument exhibits loads of sympathetic metallic ringing. If you have just worked with digital pianos, then
you wouldn't be aware of this so much because that type of behavior isn't something that can be sampled, as it is
interactive and actually occurs after notes are struck, and is dependent on which notes are struck, at what velocity, and their position relative to each other. You can't "bake it into a sample". So, this matter of metallic
overtone ringing isn't an annoying one to me. I do see that Modarrt plans to release an update soon that will address it, nevertheless, which I applaud.

Yes, the difference between V4 and V5 is quite dramatic, to my ears. I've had to go back and re-do my presets
from V4, mainly to dial back all the highs I added in equalizing, which is no longer necessary with V5. But the
difference to me is not just a matter of brightness or "metallicness". Across the board, V5 exhibits more concise
imaging, a more transparent high-end, more punch and presence...it just sounds more real. Like I'm in the control
booth and listening to the real thing, mic'd, on the other side of the glass. AND, it is finally a great pop-rock piano
instrument. I've been with Pianoteq since version 1, always believed it would win the virtual piano war eventually,
and I now believe that with V5 it has won.

I love the mic options as well.  Some great condenser and ribbon mic models in that list!  I was almost expecting to see some mic pre models...tube & solid state!     Maybe in 5.5   ( Half joking. )
  I ended up deleting all of my user presets from 4.5 and starting fresh with all the enhancements of V5.  Previously, I removed the onboard reverb in favor of my dedicated 3rd party reverbs. I don't know if the onboard reverb was even updated but it sounds great to me in the factory presets now.  About the only criticism I have is, with my old eyes, I would like to see a bit more contrast in the mic view and the text for the condition slider.  Overall, I think it's a truly great update!
   -Perry-

Well remambered !

A nice tip to help judge the sound, is to always also hear a good recording of a real piano before test pianoteq. If the true piano recording do not sound good in the speaker system you are using, you can't wait pianoteq to sound great.

Headphones can sound ok for some musics, but only a few expensive ones sound good for solo piano.

A little confused, I use ASIO4ALL and although i can set ASIO as 96khz, the max i can set pt is 48Khz. Should the rate in ptq be the same as the ASIO settings?
Also, can you explain Asio Direct x full duplex driver- i thought it was either asio or Direct x.
I'm not particularly knowledgeable of these things, so forgive me if I misunderstand

It's all pretty complex, until you get your head around it.

So the Standard version will only go up to 48k internally (Options->Perf->Internal sample rate).  This means that all PT calculations are done in (say) 48k.  If your sound card (or your DAW project) are set to a different sample rate (eg. 96k), PT will 'resample' its 48k calculations to that, ie. it just converts them.  This way it's compatible with these rates, but if they are higher, you will not get the full quality.  On the other hand, that also saves some CPU (higher rates = more CPU).

The Pro version though can go internally up to (I think)192k.

If you can, match the internal and sound card/DAW sample rates.  Then PT doesn't have to do this translation, which is a little less CPU work and (potentially) a slightly purer sound.  But if you want to work in a higher rate than your version of PT supports internally for other reasons (eg. in a DAW), that's fine too.

I use the software at 48000Hz, and the quality is very very! good.
I think that you should benefit playing at 96000Hz only if you have very high end material.

Hello, Pianotrancer. I use 96 KHz mostly because higher rates of PT makes latency twice lower in my case. It slightly influences sound, but for the quality it is more important, what is the rate of your soundcard. When I switch from 192 KHz to 41 KHz in settings of my s-card, PT becomes poorer. But it depends also on equipment (on my second computer and Audigy 2 I don't have such big differences)

About asio direct x duplex. As I understood it was installed with my Cubase. For my taste, this driver goes better then asio generic low and Asio4All (in my opinion it sounds deeper or, may be, not so harsh and unrealistically clear).

What does resolution have to do with realism?

This always comes up in the forum: "...you must run pianoteq at 192Khz, and the best headphones available for it to sound realistic ... "

This doesn't make any sense - even on the cheapest mp3 player with $5 headphones, you can still identify a real piano when you hear it.
It doesn't sound like you're standing next to it, nor in the concert hall, but it does sound like a real instrument playing through a cheap sound system!

It's the same with computer graphics rendering - low resolution images of real life are usually more realistic than high resolution CG renderings ...

I'd really like to be persuaded otherwise (because resolution is easy to pump up, information inst!)
So dear forum members, what do you think??

B.T.W - Modartt I'm excited about the new version, thank you for it - though I have yet to find the time to give it a real spin. Looking forward.

-- E

That wasn't me.  I use a higher-end soundcard with its own ASIO drivers, so I haven't used the generic ones for a while.  I did try ASIO4ALL once a few years ago, but it wasn't particularly stable on my setup at the time.

Basically ASIO gives you two things - a pure audio path (the sample rate you choose is not changed again by the sound card drivers, something the non-pro audio ones often do behind your back with other driver types like MME or DirectSound), and much lower latency.  The generic ASIO drivers try to give you that for cards that don't have their own, but the underlying hardware (eg. the sound card chip built into your motherboard) may not allow very low latencies, like pro audio cards do, or have other limitations.

Theoretically, one ASIO driver should also not sound any different than another, as the whole idea is to pass the audio without changes.  But with generic drivers maybe you can't guarantee that.

If possible, I would recommend a more pro-audio sound card or sound interface, with a native ASIO driver.  They don't have to be super expensive (although there are some very expensive ones too).  I don't know the current cards/interfaces and prices, but try to look around for recommendations.  Also note that some give you stable lower latencies than others.

BTW, having spent a bit of quality time with PT recently, I find that I now actually prefer a audio driver latency of 3ms + internal PT latency of ~8ms.  Even though I can use much lower PT latencies, on my Ensoniq TS-12 piano-style keyboard (which even has note-off velocity btw) it feels most like I'm triggering a real hammer this way.  In the past I played with the lowest PT latency I could, but now I find this doesn't feel like a real piano.  So maybe getting ultra-low latencies actually isn't that important for PT (although it's very important for eg. guitars).

This made me laugh.  Thank you, vjau.

You play very well, very nice sound.
Do you use a preset here?

Simply beautiful ! Thanks...

I' not sure what my soundcard settings are; It's a Realtec HD card. But what I have found is that seting my ASIO4All to the same 44100 rate as Pianoteq. Routing my audio out to the unbalanced inputs on my KRK powered monitors, and using pianoteq's "Pop" eq on the D4 is absolutely amazing in terms of clarity.
I am using the audio buffer at 64 samples both on the Asio control panel and on Pianoteq- which is giving me 1.5ms latency. Wow--- I'm in Heaven!

To select more than one instrument in the preset manager, you now have to use shift-click (or right button click). This was changed to be more consistent with usual selection methods in other softwares.

Well, just to take a break from the discussion and because Dave Polich mentioned it, an old cover sound alike that's been made by me for an ad came to mind
I did all parts transcription, recording, mixing, mastering and played all keyboards.

Everything is real but the piano and synths.. Pianoteq3

Here it is just for fun

http://www.screenplay.it/showreel/audio/DR%20MOTOR.mp3

Sustain is far shorter, on par with digital pianos. I have to slide the soundboard impedance way up to approximate the sustain of a real piano (and what I was hearing in PT3).

Tone-wise, it's better than PT3 in certain aspects, but still not as convincing as high-end sampled pianos. I think the way to go eventually will be sampled attack + modeled decay. I'm guessing the first few milliseconds are mathematically the messiest and most difficult to model.

Eh, what? Very much no.

The sound really is refined to the point of besting samples in every way. Simply put, my real piano downstairs sounds *faker* than my Bluthner 5, because it is so much easier to hear every nuance in the 4th gen piano (not obscured by room dynamics/listener position/piano type/tuning). So, I am really hearing many more layers of intricacy and detail in the Pianoteq sound - no joke

But then again, I've got an amazing sound card, headphones, and keys, with great care put to calibrating every element..

Have you checked that passage with a real piano?  I recall thinking version 3 sustained unrealistically long when listening to Hugh Sung's Clair de Lune.  But I have never carefully analyzed this.  And it may have not so much been the length of the sustain in v3 that bugged me but, rather, the lack of development of the sound in v3 during a given sustained note.

You intrigued me, so I decided to some tests of my own....

The decay of the D4 behaves and lasts exactly the same as my Steinway sample libraries.

The Bluthner's decay is (slightly) longer than the D4 (just like in real life) and matches up with the decay of my Fazioli, and Bosendorfer sample libraries (give or take a second of decay...very quiet by that point...)

Thank you Joshua for this excellent visual which clearly demonstrates that the double decay is closer to the real thing in version 5 than in version 3, which is precisely a point on which we have been working hard.

Of course one cannot discuss your taste, and I perfectly admit that you prefer the decay of version 3. But then, as you have mentioned, you have several tools that you can mix in various ways:
- increasing the Impedance
- augmenting the Cutoff
- reducing the Q factor
- reducing the Direct sound duration.

A very entertaining read, through these pages of dispute, conviction and analysis.

I've 'been with' Pianoteq since v2 and have watched and listened to the development with interest and a certain confidence.

I've long had a vigorous interest in 'digital' pianos, and have invested quite a bit, financially, in the pursuit of this interest.

I would hesitate to state that I am a professional violinist, since many of my colleagues could care less about 'sound reproduction', declaring themselves more interested in ensemble, expression, whatever. Of course, we are spoiled with 'real sound' every day. This includes, for me, easy access to 2 Hamburg Steinway Ds, 2 New York Ds, a Bösendorfer 290 Imperial (not just any 290 – the 290 previously own and toured by Victor Borge, no less) and a medium-length drive to the Stuart & Sons factory, where I've enjoyed a behind-the-scenes tour conducted by Wayne Stuart himself (and with whom I have discussed the Pianoteq modeling software, much to his interest, by the way). Those big pianos of his, apart from sounding beautiful, are unbelievably exquisite pieces of furniture. I hope we see a model of his work at some point, and would gladly facilitate that achievement.

Since v4, I have felt things where moving along very nicely for this stunning product, not just in sound quality, but in 'fashionability' of the sound.

Now, with v5 (notwithstanding severe comments from some) we have a truth of sound coming from the tone generator engine that is absolutely convincing. I would absolutely challenge even a sophisticated listener to identify the model, as distinct from reality, in a blind listening test, and to a statistically significant degree.

This is to address the accuracy of the sound – not the exact experience of playing the instrument. No doubt this is a yet more stringent test, and dependent upon so many variables as to be outside of the 'responsibility' of the virtual instrument designer/manufacturer. But the interactive luxury of this sophisticated model is incomparably greater than even the most exhaustive efforts from sample-based instrument builders. There is really no comparison.

No doubt, further developments are possible, and will be welcomed, as regards the sound engine. I do feel, though, that the primary remaining challenge is the construction of a truly immersive interface, one worthy of the exceptionally 'true' engine now available, and of those we expect will exceed even this current level of excellence.

It may be that the best way to 'experience' a piano is not the way we are obliged to experience a real acoustic grand or upright. This 'real' experience is simply a function of what must needs happen in the business of activating and generating the sound. I think we all agree it is a compelling experience, but it would be rash to say it is definitive, inherently optimal. We are so busy emulating, we have not quite got to the point of dreaming afresh.

Well, Wayne Stuart has... sort of... and good on him. And I should say, with the extraordinary array of adjustable parameters available in the Pianoteq environment, none of us has the slightest excuse for not jumping on board that voyage to a brave new world.

It's kind of funny. When we hear the instruments Liszt and Chopin had to work with (never mind Beethoven) it is hard to grasp how they were so 'turned on' by that sound to conceive the undeniably powerful masterworks bequeathed to us. I think we are all too easily guilty of seeing things through a very narrow lens. Not intended as a criticism, just a provocative thought. Meanwhile, it is really cool to have the facility of a monster piano playable at full volume, even in the wee hours of the night. This, all by itself, is a singular service to our civilization.

Cheers,
Stephen.