Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Is this now real enough?

This sounds beautiful!

I think a true test of realistic would be in person, rather than a recording. I think DPs and software pianos have come a long way and have made professional sounding recordings very possible. The one thing that DPs haven't quite (as far as I know) been able to accomplish is the feel of playing an acoustic instrument that comes from the sound vibrating through the entire instrument rather than speakers.

It would 'fool' 100% of the population, but there were digital pianos and sample libraries that could do that 20 years ago.

Your playing is flawless and amazing but the end result is clinical and boring IMO. It isn't your fault. All of the attack portions of every note sound the same in that performance. It was like being assaulted by a musical typewriter. To me that is the flaw with all sampled pianos, the inability to alter the attack portion of a note according to how much the string is already vibrating.



Only you know the answer to that question but I would much rather hear you play it again on one of these:

https://www.youtube.com/watch?v=Y7RkwjNplaE

@Philip, bravo for your playing. Terrific in every sense. Also your experiences gave me a lot to think about. But regarding the question of if this kind of recording could "foul" a listener, in my opinion it's difficult to tell, at least at first, knowing in advance how was it made and, without comparing with a quality acoustic recording. A blind test would be a better way.

But the second time I listened to it (with great pleasure, I'd must add), I did it without looking at the video. There were then a few details that, at least for me, revealed some digital artifacts on the samples, specially on the treble. And also there's something missing on every library. I don't know how to express it but it's something like the notes lack cohesion as they were sampled separately, I miss ambiance and whatever... I don't know exactly what it is. By listening only to your recoding one may be fooled but listening immediately after that a quality acoustic recording, I think the enchantment or illusion was broken. That IMHO has nothing to do with the quality of the performances as I think that yours is very good. And the sound take is magnificent, surprisingly good, but there's still something missing, at least in my very humble opinion.

What I can tell, from my experience, is that the acoustic home recordings I do for myself are far superior to those I made with my digital Roland plus Vintage-D software (or the like); even given the fact I own a regular grand, not a concert one and my record equipment is not too bad but is still a home one. But I tell this to add my experience to the thread because of if someone were interested in listening another opinion.

Congratulations again as I enjoyed a lot your playing and the recording.

The dynamic control is superb. At about 1:16 it really shines; in terms of the subtle/gradual augmentation and timbre variation of the sound. Of course, a lot has to do with the player, but the grand piano action and virtual instrument (CFX) set up allow for a more intimate and immediate connection.

That`s a bit extreme! I didn`t quite mean that . .but if a concert piano is tuned before every performance, it would indicate that , the longer the music goes on for, the more out of tune it becomes! Now, some of you pitch perfect guys would notice this, but you`re enjoying the resonances too much to care . . .

...and that's just on headphones, not even speakers.

I'd agree that what happens in the attack portion of a note is critically important and is quite possibly (though not certainly) what is diminishing the realism of this performance.

However, I don't think there's any useful scientific insight relating to the 'already vibrating string/new hammer strike' interplay that you could possibly draw upon.

In any case it would only apply to repeated notes. Are they in abundance in this piece?

In many classical piano concerts, the technician comes on during the interval to retune the piano. Especially if Liszt, Prokofiev or Rachmaninov was on the programme in the first half.....

(Actually, I can't remember the last time I didn't see the tech tuning the piano during the interval of a piano recital, at London's South Bank).

Most if not all sampled pianos actually do produce a changing timbre for repeating notes with the sustain pedal, because of the phasing interactions of multiple voices (i.e, samples for the same pitch) overlapping - the previously invoked voices are allowed to ring for a while (sometimes a very long time) - they are not simply cut off immediately every time a new strike occurs, even on the same note.

Greg.

I have looked at the specs. This thing needs a lot of power.
What spec Imac are you using to run this.
Rach fan by the way......keep the demos coming....seriously impressive.

******"this thing needs a lot of power. what spec imac are you using to run this"

IMac is a late 2013 27 inch - 3.5 GHz i7, 16 GB 1600 MHz DDR3.
Samples are streaming from an external hard drive via USB3 (not a SSD drive), everything keeps up fine.

More demos coming soon.

Excellent- both playing and sound ! Congrats.

Yes agree- the only negative from a player pov are the inferior speakers on the N3. Although I think the AG's speaker system compliment the older AG sample very well. I'd have no reservations having an N3 as it is, as a practice instrument.

But I could see how coupling the internals with a higher quality sample would not be satisfying. So all you need are a pair of these
http://www.sweetwater.com/store/detail/SCM100ASL

Personally I can't play/practice with phones on too long. Reminds me of too many hours/years in recording studios playing music I didb't like, but did it for the $$$...

This isn't a fair comparison as the sound of that acoustic is ambient in the hall. It's like looking at a painting from a distance. If the performance had been close mic'd like the Garritan CFX samples then we could debate about what we are hearing. Also, audio/video compression on youtube is for the birds - but that's a different topic. At least both recordings have suffered similar. I suppose if he wanted to soften and blur things a bit he could slap a Space Designer or other convolution reverb on it and see if that fools 101% of the listeners. I'm joking of course.

Truth be told, this is a great example of how far the sample libraries have come. The sound is very precise and very detailed. I think it's an awesome compromise for recording your performances at home. I'd be more interested to hear Garritan CFX vs. the on board Yamaha samples in the N3.

Your videos are impressive,
I found the sound just normal, as it is expected on a recording (and that is a compliment).
It is crisp and clear. I found this sound perfect for classical music.

May I ask you what is your setting for the samples value and what is the "on screen" latency on your system? It is just indicative for me, to know what kind of latency suits to a pianist (I am just a beginner, 1 year of practice).
Do you feel a difference in the response when you play your Yamaha instead of the CFX library?

Thanks, and bravo for you excellent work.

Of course! Probably 99.9% of the population would get fooled. It's only people like us here who are overly critical and analyse every sound bite.

That's why I posted to this forum Figured that if the community here didn't fall about laughing at my early efforts, then my new setup may yet be a viable way to make videos after all.

I do agree with your comment about "clean" sound as being a problem with sample libraries. I've never played an acoustic piano that didn't have the odd "beating" unison, or quirky mechanical sounds; such things are deliberately and noticeably absent in the digital world, no matter which library you choose.

The same problem hit developers when driving games first appeared on games consoles. The roads were always perfectly and uniformly grey, which made them feel artificial. Now, roads in such games have imperfections—cracks, tyre marks, "patched" potholes, ragged edges—and they feel much more real for it.

That metaphor doesn't quite translate though for music. It would take a very brave sound engineer to deliberately record an out-of-tune or erratically voiced piano....particularly since the quirks would be exactly—and maddeningly— the same every time. It's also not a stretch to imagine that consumers might be wary of such a proudly warts-and-all library.

A compromise suggestion? Maybe the perfection issue is something that could be handled instead in settings. Perhaps there could be a new 0-100 setting called "Maintenance Level", where 100 means "A team of award-winning piano technicians have just spend three months fine tuning this instrument in humidity-controlled zero-gravity", and 0 means "clapped out train wreck that would make John Cage blanch". Pianists could pick whatever level of grunge or pristinity (is that a word?) in between suits their style.

As you may know, this has already been done through modelling, rather than sampling, in Pianoteq. The user has a main-screen setting which allows you to slide between 'perfectly tuned and set up' through to 'warped, ruined and out-of-tune'.

The idea that, unlike with samples, the imperfect piano will not become irritating because the various facets of imperfection are different every time the piano is played, since the whole sound is arrived at by fresh computations.

Despite this, Pianoteq still suffers somewhat from being - or being perceived as being - too clinical in its sound. It is still early days for acoustic modelling and eventually, I think they will engineer the grunge into the equation, literally.

But the whole history of recorded classical music has been a consistent attempt to iron out any imperfections what ever. Therefore, your performances have arrived at an ideal end-point, in a sense. The only direction now is downward.

Dear Philip_Johnston,

great job and fantastic videos!

Just like you, I made several attempts to combine my N1 with VST.
The only positive result so far I reached was with QL Piano (Stainway) Room mics plus "Altiverb" or "Spaces"(East West).

The only two question about the Garritan CFX are:

how about the velocity response from Yamaha AG. I analysed it and the highest level (say over 100) are difficult to be reached. (Personally I have never recorded note than 114)
Is there any possibility on VST engine to compress the velocity.

And the second question: is there re-pedaling option on CFX VST.

Thank you
__________________

At the end talking from my experience and many attempts – AG, recorded with 3 additional microphones (one from the bottom for low frequencies and 2 positioned from a distance on top – left/right (XY) just like a stage set in a normal piano recording) + AG's line out sounds GREAT in the mix as a solo instrument.
Additional reverberation is always necessary.

It's certainly possible to tweak the velocity curve on the Garritan (and is essential to match the repertoire you're recording...I really wish that MIDI offered the same resolution as the native Yamaha format); unfortunately though it doesn't seem to offer repedalling.

Haven't figured out a way to record the AG acoustically yet where it works as a solo instrument - I think the problem is that I'm recording speaker output, which is never ideal I might try your configuration though, if I can get my hands on an additional mike or two.

The N3 also offers three different basic velocity curves on the instrument itself (I'm assuming the N1 does too?), with the "light" touch being more likely to yield those higher velocities.

Not sure if any of this helps or not...

Thank you very much, Phillip!

I have my N1 for more than 2 years and forgot about the "light" option for velocity.

You play fantastically Rachmaninov!
What was on youtube, I heard all performed by yours and it made my day full with joy! One more time – thank you for the pleasure!

________

In EU Garritan SFX is hugely overpriced (VAT incl. €275, NETO €231) but on MakeMusic site appears as $199 (ca. €181 excl. VAT)

_________

And just for the test – part of demo recording with the mixture of N1 and QL Pianos (mentioned in my previous post.
The performer is a friend of mine.

https://soundcloud.com/wess-music/test-n1-plus-ql-piano-room-mic

Best regards,
Wess