New Yamaha Motif XF

[Dr.Innovation]

Anyway the problem with Kurzweil piano lies in the middle octaves. The higher keys sound good and the lower sound ok too.

 

 

 

 

The XS doesn't sound like that stock.

It's an addon right?

[DmitryKo]

 

The XS doesn't sound like that stock. It's an addon right?

Full Concert Grand is a factory voice PRE1 A01, a sample of Yamaha CFIII piano; Dream Grand is a performance USR3 F12; S700 piano from S90ES is a free sample voice download for Motif XS and XF (requires sample memory). Motif XF also includes Natural S6 grand from S90XS, voices USR1 A01-A04.

 

Sound demos:

http://www.yamahasynth.com/products/synthesizers/motif_xf/

http://www.yamahasynth.com/products/synthesizers/s9070_xs/

http://www.yamahasynth.com/products/synthesizers/s90_es/

 

 

Kurzweil, with its 4 Mbytes piano ROM from 1990, destroys Motif XS/XF with their 20-something Mbytes CFIII, 53 Mbytes S700 and 142 Mbytes S6 piano samples? I don't think so. Yamaha CFIII sample has been used in Yamaha P-series, S-series and Clavinovas which have been the preferred digital instruments for many classically trained pianists, with Roland probably as the second most popular. On the other hand, I have not seen as many people play Kurzweil or Korg pianos.

[Dr.Innovation]

Full Concert Grand is a factory voice PRE1 A01, a sample of Yamaha CFIII piano; Dream Grand is a performance USR3 F12; S700 piano from S90ES is a free sample voice download for Motif XS and XF (requires sample memory). Motif XF also includes Natural S6 grand from S90XS, voices USR1 A01-A04. Sound demos: http://www.yamahasynth.com/products/synthesizers/motif_xf/ http://www.yamahasynth.com/products/synthesizers/s9070_xs/ http://www.yamahasynth.com/products/synthesizers/s90_es/ Kurzweil, with its 4 Mbytes piano ROM from 1990, destroys Motif XS/XF with their 20-something Mbytes CFIII, 53 Mbytes S700 and 142 Mbytes S6 piano samples? I don't think so.

While I understand what you meant
a newbie reading it may not.

The posted clip is of the loyalty program S700 sample
that came on a disc, and isn't (legally) available for download.
If it is, please post the link thanks because I've yet to find it.


And btw, it's not the amount of MB or GBs that make a piano
patch sound better than the other, its in the programming
and sound design architecture.

My Alicia's Keys VSTi outperforms and outdoes East West Quantum
Leap pianos handily and its easily a difference of 50GB between
AK's and just one piano of QL's 4 piano library.

[Aanalogaddict]

Kurzweil, with its 4 Mbytes piano ROM from 1990, destroys Motif XS/XF with their 20-something Mbytes CFIII, 53 Mbytes S700 and 142 Mbytes S6 piano samples?

No, but they're definately in the same ballpark. Which - by logic - tells us that the next generation of Kurz samples will outdo the next 20 years of Yamaha samples...

Physical modelling is the way to go IMO. In just a few computer generations we'll have physical modeling in workstations and a whole new ballgame.

[DmitryKo]

 

they're definately in the same ballpark

Judging by sound demos of PC3K8, I'd say they are definitely NOT in the same ballpark.

 

 

Physical modelling is the way to go IMO

For the price of Roland VPiano, the difference is not that dramatic. PianoTeq is not that good either.

[Dr.Innovation]

S700 piano voice was made a free download as a part of the XSpand Your World promotion. http://www.motifator.com/index.php/main/comments/xspand_your_world_content_updated_with_new_free_downloads/ True, but no programming can compensate for the lack of velocity layers or key zones, and 4 Mbytes is well below any acceptable limit for the year 2010.

 

 

Thanks for the link, a buddy of mine was looking for the sample

and I was looking for more sounds to try:thu:

 

2nd, I'm confused with your point.

I understand your POV, and with all the sandbaggin these ROMpler

manufacturers have done to use enthusiasts and musicians

Kurzeil was basically resurrected and had limited funds from what I

had read a couple of years back or so...recycling old sounds

is common. -With the exception of the Motif Es to Xs, though.

 

This is why I jumped ship and went sample based.

As for the Motif XS/F I feel these sound better than most

synths out there, and while I did own a K2600XS and really enjoyed it

I don't think the newer Kurz synths sound any better, but

they did add more variety and tweaks which I think sounded pretty good.

[DmitryKo]

2nd, I'm confused with your point. I understand your POV, and with all the sandbaggin these ROMpler manufacturers have done to use enthusiasts and musicians Kurzeil was basically resurrected and had limited funds from what I had read a couple of years back or so...recycling old sounds is common. -With the exception of the Motif Es to Xs, though.

I beg to differ. No other manufacturer that remains in MI business still uses their sound ROMs as developed in the year 1990. Not Yamaha, not Roland, not Korg. Probably not even Casio. Even E-mu does not do that with their Emulator/Proteus software instruments. They all made a significant progress and now operate on 0.5-1 GBytes of samples, that is 2-2.5 orders of magnitude (100 to 250 times) more.

 

If Kurzweil realy managed to successfully use their complex synthesis architecture to improve on the limited sample size, I would have no complains. However what I hear in their demos does not impress me at all. Their piano has a distinct sound of a 4 MB board, I've had a couple of ones made by Alesis and E-mu. It may hold up with 16 Mbyte Korg Triton piano boards or pianos in the Yamaha S80/S08, and maybe earlier Motif, but it is NOT in the same ballpark as modern Yamaha (ang arguably Roland) gear. I own an Yamaha P-90 which 8 year old now and it sounds agruably better than this Kurzweil; Motif ES has the same CFIII piano sample. Not to mention the newest CP-1/5 and S90XS.

 

Just listen to On Cloud Mointain sound demo for the S90XS, which uses Natural S6 Grand voice, with good quality headphones; there is NO comparison to Kurweil pianos in the PC3 series.

 

When (and if) they will finally release K3000 and upgrade their samples, I will surely evaluate it as well. In fact, I'd think anything they would put up to replace these 1990s sounds will be significantly better than PC3 series

[kilon]

I once did a quite faithful emulation of EWSQL's SILVER EDITION single french horn with a single oscilator in Absynth 3 . No samples were used or abused.

It show how a powerful engine can render samples unecessary. This the true power of PC3X that it does not need samples, it just use them to make things abit easier for the user. "Need" meaning that you can create sounds without using any samples because there is a VA engine inside.

I can tell you here and now that I have played with K2660 which happenes to be owned by a friend of mine , he tought me what VAST is really is and why its called VAST. Its pianos and accoustic sound where not even close to my Motif's ES sounds. I own an ES and know its sounds very well. I have played two times with XS . I dont talk about expansions, edited presets or anything else, I am talking about 100% on board FACTORY presets. All other pianos of the above models do not even dare to compete with PC3X piano.

And even if XS piano sounded lot better , pc3x will have no problem compete with it, its just a question of programming. MOTIF engine on the other hand is very limited and largely sample based.

But of course you have the right to disagree with me , absolutely and say that XS piano is much better than PC3X piano.

But please dont compare the PC3X rom with motif ROM , its 2 entirely diffirent synths . By the way I like to inform you that MOTIF XS piano is 8 layers , PC3X piano is 31 layers (whatever that may mean to the average musician, it means to me alot because it helps the piano to carry alot of expression ) and by the way pc3 layers is completely unrelated to motif's layers and flexibility.

Night and Day.

But as I said I am not a piano player, just a guy that likes to create music and sound in his free time and all I am stating here is nothing more than personal opinion. Well except the fact that k2660 , es, xs and pc3x have diffirent pianos and that pc3x and generally VAST has at least 10 times more sophisticated engine than Motif.

[TechEverlasting]

No other manufacturer in the synth industry has a useable set of orchestral strings in its factory ROM - not Yamaha, not Roland, not Korg.

With all the talk on this thread about this synth slaughtering that synth I'm surprised that no one has mentioned that the new set of strings in the PC3X ROM (brand-spanking-new, not from 1990) not only viciously murder the lame string samples in the Motif XF, they also insult the XF's Mom, steal its dog and have their way with its girlfriend.

I will grant that the new accordion sample in the XF's updated ROM is awesome.

At the end of the day these are two very different instruments, and I'd be happy to own one of each. Yes, if you have a 14 layer deep program in a Kurzweil each layer's envelope and filter parameters are individually adjusted. For me that's a source of endless joy, for someone else it might be a PITA.

[DmitryKo]

No other manufacturer in the synth industry has a useable set of orchestral strings in its factory ROM

Interesting, though I don't think I ever heard convincing strings since I owned the Alesis QS... but we were talking about piano, weren't we?

[kilon]

Are you saying that Kurzweils can a make faithful acoustic piano sound using no samples but just its VA engine? That's quite a bold statement, which is pending a demonstration.

 

 

Bold statement ? hardly.

 

additive theory says that a sound can be faithfully be emulated with a collection of sine waves. This theory gave birth to additive resynthesis.

 

of course in real life piano emulation of any accoustic sound ,an additive synthesizer might need more than 1000 sine waves. But in the VA world things are abit more complex than simple sine waves.

 

Now that we have set straight the theory now its time for the proof. While I wish I had kept that french horn Absynth preset. But rest assured proof is not far away. For first you can feed your pianos samples to additive resynthesis softsynths . For example I have used alchemy to emulate many accoustic sounds, its quite easy really, you load the sample it analyze it and in 2 seconds it gives you an emulation with sine waves.

 

There is this guys that emulated many orchestral sound with analogue synths. He used Alesis Andromeda for this one.

 

 

 

Of course emulating accoustic sounds with VA engine is not actually easy. And that is why PC3X comes with samples, thus making it easy for the average user to emulate and create easier his sounds.

 

I can not say I understand the other 2 points you made (about the layers or ES sound vs the k2660 remark of mine) and I cannot persuade you to like PC3X piano, all I am saying that if you dont like the pianos on motif you have to compromise with that if we do not take into account that one can load his own samples, but if ones does not like the pianos on PC3X he has alot more space to move around till he find the sound he really likes. But then that works on the assumption that one is also a tweaker and not a pure piano player.

 

However if you seriously into piano sound the I strongly advice either a physically modeled piano or a big soft rompler. I dont think that sample orientated workstations can compete with that.

 

Actually I knew a person that he was once turned down for commisioned work for a classical soundtrack when he mentioned that he would use a workstation.

 

As for your demonstration just do what I suggested, feed a good piano sample or your own sample if you so wish to Alchemy and see how good it emulates it. You will be surprised how far a substractive can go. Also alchemy has a "partials" setting that you can use reduce the amount of sine waves even up to 1 oscilator / sine wave to see how the amount of sine waves affect the emulation.

 

Have fun

 

PS. I can post example waves of an accoustic piano and the emulation by Alchemy (which I happen to own) if you wish so. Or you can send me your sample as well, just upload it to http://www.mediafire.com and pm the link.Unfortunately I dont have a PC3X to guide you through the process of making a faithful representation of an accoustic pianos. I find additive resynthesis far easier in these matters .

[DmitryKo]

additive theory says that a sound can be faithfully be emulated with a collection of sine waves... of course in real life piano emulation of any accoustic sound ,an additive synthesizer might need more than 1000 sine waves. But in the VA world things are abit more complex than simple sine waves.

Show me a hardware instrument capable of generating even a single voice that uses 1000 oscillators. This is certainly well beyond the capabilities of any V.A.S.T board, including PC3X.

 

Of course emulating accoustic sounds with VA engine is not actually easy. And that is why PC3X comes with samples, thus making it easy for the average user to emulate and create easier his sounds.

Exactly, with the exception that it does not make it "easier for an average user", it makes it possible at all.

 

this guys that emulated many orchestral sound with analogue synths. He used Alesis Andromeda for this one. http://www.youtube.com/watch?v=73iYaoXBzVY

Nice analog orchestra sounds (which were probably only possible with the most expensive old-school modular synths of the past), but it's not nearly convincing enough to be mistaken for a real symphonic orchestra. If you think it is, I suggest you should go to your local music store and buy any compilation of J.S.Bach

 

For example I have used alchemy to emulate many accoustic sounds, its quite easy really, you load the sample it analyze it and in 2 seconds it gives you an emulation with sine waves.

Again, how does it relate to Kurzweil, whose practical limitations are well beyond this use?

 

I can not say I understand the other 2 points you made (about the layers or ES sound vs the k2660 remark of mine)

I'm just saying that most layers in VAST are not sample layers, like on the Motif, but are essentially onboard oscillators which generate simple waves or noise from a physical model. And if you didn't notice, the recent discussion deals with Motif XS vs PC3X, not Motif ES vs K2660 which are different boards.

 

he was once turned down for commisioned work for a classical soundtrack when he mentioned that he would use a workstation

Quite rightfully so, he should have at least used a dedicated piano board or a Clavinova.

 

However if you seriously into piano sound the I strongly advice either a physically modeled piano or a big soft rompler. I dont think that sample orientated workstations can compete with that.

I did try GigaSampler and lately Pianoteq, but unfortunately they didn't live up to the hype as they ddn't really feel and sound like a real piano when played from a good graded action keyboard, unlike Yamaha P-series. Any other suggestions?

[Dr.Innovation]

I never understood why people would limit themselves
to namely Hardware Synths for acoustic instruments?

It's not a fair game to compare a real instrument with
a replica, however it's also not fair to compare
a hardware synth (which is immensely limited in terms of memory)
to a sample library?

The best acoustic guitar ever made isn't in a Kurz, Roland, Yammy...
It's the one made by Pettinhouse and its remarkable!
A league of his own, and his new ukulele is phenomenal!

Sample based software is where it's at, no question, no exception.
The Alesis Andromeda video Kilon posted had to have been a joke, right?
There wasn't a single authentic sound in that video.

I also think Pianoteq is a joke is right there with true pianos
as being a waste of good money for true sounding piano copy.

If we're looking for the best piano patches
either derived from sample based software or ROMplers?

Ivory 2 is has gotten better.
Alicia's Keys and Tonehammer's Emopiano are both exceptional
playing, and sounding.

This one is also very top notch:
http://www.acousticsamples.net/index.php?main_page=product_info&cPath=&products_id=29


The best in my mind would have to the VSL's Bosi library.
Not one single Hardware synth comes to mind for a piano.


If we're to talk about strings?
EWQLSO Gold edition (16bit strings) this decimates
anything Kurzweil could ever come up with.

Still I am interested in hearing how Dmitry feels
how the PC3X is outdated?

[ChristianRock]

Well I don't think anybody's saying hardware can compete with giga-libraries. But there are reasons to use hardware, especially if you gig.

 

But still, that Yamaha S700 sample is perfectly fine for most uses.

[kilon]

Show me a hardware instrument capable of generating even a single voice that uses 1000 oscillators. This is certainly well beyond the capabilities of any V.A.S.T board, including PC3X.

 

you are missing the point. A va does not need 1000 sinewaves because it can generate waveforms far more complex than sine wave which is the simplest form of wave.

 

Vast can do both fm and additive. You have to remember I talked strictly sine waves. If you take a sine wave and fm by itself you can come up with sound that are equal to 1000 sine waves. Add abit of effects and viola.

 

 

Nice analog orchestra sounds (which were probably only possible with the most expensive old-school modular synths of the past), but it's not nearly convincing enough to be mistaken for a real symphonic orchestra. If you think it is, I suggest you should go to your local music store and buy any compilation of J.S.Bach

 

 

I already told , it was made using an Andromeda , its not expensive neither modular , but it is a complex and deep synth. I happen to own an Andromeda and costed me just abou the same a Pc3x cost. Are those sounds realistic enough to trick a person into thinking its the real thing? no but if he combined it with samples he would certainly went alot further with his emulation. That was my point.

 

 

Again, how does it relate to Kurzweil, whose practical limitations are well beyond this use?

 

 

Aha... then you found already the relation .If an analogue with only 4 diffirent sounds can do this , imagine what a synth with loads of samples , a complete modular engine and a va engine can do.

 

 

 

I'm just saying that most layers in VAST are not sample layers, like on the Motif, but are essentially onboard oscillators which generate simple waves or noise from a physical model. And if you didn't notice, the recent discussion deals with Motif XS vs PC3X, not Motif ES vs K2660 which are different boards.

 

 

incorrect my friend. Layers on motif are basically a one oscilator timbres containing the full substractive chain of filter and an amp with corresponding envelopes and lfos. Layers on the VAST engine , which is the same for k2660 and pc3x are completely diffirent. They can be anything and they can be nothing.

 

A VAST layer is a collection of algorithm slots, now these slots have diffirent setups, there are single slots , double slots and triple slots. The reason for these distictions is because each algorithm uses diffirent amount of processing power.

 

Now because of this modular setup , the algorithm can be anything you want, an oscilator, an amp, a filter, a ringmodulator , an additive oscilator etc. So its possible to have layers that make no sound at all, so for instance you may have a layer that is 3 filters . This happens because VAST not only allows you to choose what makes a layer but you can route your layers. So you can say to vast "I want my layer with oscilators to pass through the layer with the 3 filters " . Or you could make 10 layers with 3 filters each and make a sample layer pass through all the 30 filters. So basically VAST is not a VA , is a modular synth altogether.

 

Quite rightfully so, he should have at least used a dedicated piano board or a Clavinova.

 

 

He did not use a piano.

 

 

I did try GigaSampler and lately Pianoteq, but unfortunately they didn't live up to the hype as they ddn't really feel and sound like a real piano when played from a good graded action keyboard, unlike Yamaha P-series. Any other suggestions?

 

 

not at all, I respect your taste.

 

anyway I dont want this to go on forever. I just said what I wanted to say , I dont want to monopolise this thread or derail it.

[Dr.Innovation]

Well I don't think anybody's saying hardware can compete with giga-libraries. But there are reasons to use hardware, especially if you gig. But still, that Yamaha S700 sample is perfectly fine for most uses.

Definitely.
My only thing to really point out is once you've
played a truly superior piano patch, coupled to an exceptional controller
the connection between keyboard and player become one

A controller is a need for every keyboard player anyway, right?
I do understand some or even most opt for a 61 or 76 key
because they don't want the weighted action of an 88 key controller/synth.

However if you're on a budget, truly want the closest experience to
playing a piano it can't be done on light, plastic feeling keys.

I feel a serious player or even a compelled enthusiast should
have 2 boards...

-An 88 weighted controller
-their 61 or 7 key synth.

Of course there are those that have more but
since I've discovered things like Omnisphere, I'm ready to
live without my old synths of yesterday:thu:

[DmitryKo]

once you've played a truly superior piano patch, coupled to an exceptional controller the connection between keyboard and player become one

True, and for piano sounds, Yamaha P-90 is this exact combination for me.

 

I feel a serious player or even a compelled enthusiast should have 2 boards... -An 88 weighted controller -their 61 or 7 key synth.

Cannot agree more, that's why as I had Yamaha P-90 in addition to Motif ES7 (soon to be Motif XF7).

[DmitryKo]

 

you are missing the point. A va does not need 1000 sinewaves because it can generate waveforms far more complex than sine wave which is the simplest form of wave.

If you are talking about virtual analog in general, most VA synths are limited to 2 oscillators per voice, and those oscillators produce sine, saw, triangle, square, PWM, and noise, just like their analog counterparts. Yes there are filters, modulators, and what not, but in the end, pure VA fails to produce a convincing emulation of real instruments.

 

 

If you take a sine wave and fm by itself you can come up with sound that are equal to 1000 sine waves. Add abit of effects and viola

I have never heard a FM synthesis implementation that sounded better than samples in regard to emulation of acoustic instruments.

 

 

If an analogue with only 4 diffirent sounds can do this , imagine what a synth with loads of samples , a complete modular engine and a va engine can do.

If 6-operator FM synthesis is not able to faithfully recreate all the harmonics of a real instrument, I don't really think 4-oscillator VA synth will be able either. I would appreciate a sound example, if you please...

 

 

A VAST layer is a collection of algorithm slots... the algorithm can be anything you want, an oscilator, an amp, a filter, a ringmodulator , an additive oscilator etc... you could make 10 layers with 3 filters each and make a sample layer pass through all the 30 filters. So basically VAST is not a VA , is a modular synth altogether.

Interesting, I need to browse through the manual.

 

Still, even if you could program it to act as piano physical modelling, 32 oscillators per note will be limited by only 128 total voice polyphony - which brings us back to the idea of using samples for piano, and current piano samples in PC3X fail to impress me.

 

 

not at all, I respect your taste

I mean, you could suggest me a softsynth piano library for evaluation... maybe there actualy exists a softsynth piano that is not an utter disappointment.

 

 

if he combined it with samples he would certainly went alot further with his emulation. That was my point.

Sorry, I'm lost in your argumentation, it's all theory that's yet to stand a practical test...

[TechEverlasting]

Why not? You can load a 2 Gbyte piano sample to the Motif XF if you have flash memory, that's practically what many 32-bit software libraries offer. .

 

 

2 Gig is nowhere near enough to load any of the current mega piano libraries. I believe the Imperfect Samples Fazioli requires 50 Gig, maybe you could get an octave of it into a Motif XF, assuming you could convert it. If you want that sort of thing for a live performance you'll need to use a laptop or maybe a Receptor. You could also have an actual Steinway D trucked in to all your gigs. Beyond that you'll be making a compromise if you're trying to play piano live. I'm personally much more excited by the prospect of processing a well crafted 64 or 128 MEG piano library within the PC3K8's superior synthesis environment (which has already done amazing things with that small 1990 piano sample) than seeing what I can do with 2 GIGs in a Motif.

[Dr.Innovation]

The best acoustic guitars are custom made by luthiers, and only then you can try and sample it, so lazy people like me who never bothered to learn guitar can make something out of it Didn't impress me at all, severe lack of dynamics IMHO I'm yet to try Alicia's Keys, Emopiano and VSL Imperial Bosendorfer. In regard to piano, by using outdated piano samples, see above for details. Well, because a digital piano like costs a fraction real grand piano, something say $1300 for P-90 vs $130 000 for Yamaha CF-IIIS, and you can practice in headphones without bothering your neighbours... and because those $100 multi-gigabyte piano libraries sound and feel like crap. Why not? You can load a 2 Gbyte piano sample to the Motif XF if you have flash memory, that's practically what many 32-bit software libraries offer. True, and for piano sounds, Yamaha P-90 is this exact combination for me. Cannot agree more, that's why as I had Yamaha P-90 in addition to Motif ES7 (soon to be Motif XF7).

Dmitry...audio demos do NOT (I repeat) do NOT convey
the products playability/overall quality and since all of these players seemingly to me lack the connection from piece to the pianist's
interpretation audio demos severely lack the definitive quality
playing it yourself brings to the table.

VSL is bar none the best samplers have to offer.
I have one of the most discernible ears I've ever known
especially when it comes to piano emulators, VST's and patches
that have ever come out...being a pianist myself first before
playing artificial piano replicas...

AK's has much character at a price of $99 -which makes
it the best bargain priced piano VSTi ever, I own and endorse it
thoroughly. It sits/cuts through a mix, has dynamics
and was recorded dry, which many piano VSTi's avoided
(namely EWQL) If we are talking about piano VSTi's
versus (insert ANY hardware synth) there is ZERO competition.

Compressed and tiny sample data cannot compete with
24bit uncompressed quality samples with sympathetic resonance
damper pedal nuances and overall tonality, depth of character
and superior dynamics...period.


When I mentioned acoustic guitars I was talking about
the programmer behind the Pettinhouse acoustic
guitar sample library, which IS the best there is.
Before striking judgment why not listen and before being so
quick to reply?

We aren't talking about real live instruments
which we can all agree are superior in every aspect compared
to artificial copies.

I don't know what type of monitors you run...
My M-audio EX66's (coupled to AK's) don't have a left nor a right.
I have a huge sound stage that captures a real piano's
sound with staggering accuracy I've never heard before
with any other piano library either played/listened to using them.

[DmitryKo]

Dmitry...audio demos do NOT (I repeat) do NOT convey the products playability/overall quality and since all of these layers seemingly to me lack the connection from piece to the pianist's interpretation

If I do not like the sound in the first place, how can I connect to it when actually playing it? IMHO it' rather the other way round - audio demos may sound good, but it will not let you connect through the real keyboard; exactly the case with Pianoteq and GigaPiano for me.

 

VSL is bar none the best samplers have to offer.

Can not say anything, their site is broken as of now.

 

If we are talking about piano VSTi's versus (insert ANY hardware synth) there is ZERO competition. Compressed and tiny sample data cannot compete with 24bit uncompressed quality samples with sympathetic resonance damper pedal nuances and overall tonality, depth of character and superior dynamics...period.

Yes, that's what they have been saying about GigaPiano like some 12 years ago

 

I don't know what type of monitors you run...

I have KRK RockIt RP6 monitors and Sennheiser HD448 headphones.

[DmitryKo]

I believe the Imperfect Samples Fazioli requires 50 Gig, maybe you could get an octave of it into a Motif XF, assuming you could convert it.

50 Gigabytes of "imperfectly sampled" piano? What a bull{censored}. Thank you, but if I wanted to make a record like this, I would just put a cheap dynamic microphone into my upgright piano, lots of so-called "imperfection" guaranteed (not to mention all the tuning and mechanical issues). I went digital to get a sound that can only be reproduced by using fine crafted instruments, large soundrooms and lots of expensive recording gear. Call it "sterile" of you want.

 

Beyond that you'll be making a compromise if you're trying to play piano live.

Just like Elton John who still uses... what, Rolang MKS-20 piano module?

 

processing a well crafted 64 or 128 MEG piano library within the PC3K8's superior synthesis environment (which has already done amazing things with that small 1990 piano sample)

I'm not a synth geek and I fail to hear these amazing things you are talking about. Then again, a well-crafted piano sample should sound good on any instrument, hardware or software.

[kilon]

I'm not a synth geek and I fail to hear these amazing things you are talking about. Then again, a well-crafted piano sample should sound good on any instrument, hardware or software.

 

 

Ok time for a little test, I downloaded a sample of an accoustic piano from the net , and used Alchemy in sampler mode where it just plays back the sound with no alteration or influence. And then used Alchemy in additive mode where it additive resynthesize the sample and made a synthesizer emulation using only additive and with no effort whatsoever. In short that "1000 oscilator " thing I was talking about earlier , eventhough this emulation is ahrdly anything beyond 500 , Alchemy can only reach up to 600 but this may be alot less than that.

 

That means I did not edited the sound , in any way this is just the additive resynthesis of the sample so its the absolutely worst that Alchemy can do to emulate the sample. No effort from me, Alchemy did all the work.

 

By the way sorry for the diffirence in volumes I did this in absolute hurry and did not care to fix the volumes. I spend only 5 minutes on this, I am sure that if spend more I could make it alot closer , but it already shows that even at its worst a synthesizer is more than capable to create emulation of accoustic sounds and with additive resynthesis is dead easy.

 

I used only one note because , making more notes would need a multisample , no time for that. But the same principle apiles for multismaples so it would not make any diffirence (Alchemy can resynthesize multisamples too) .

 

Can you spot which one is the real piano and which one is the synthesizer ?

[Dr.Innovation]

I have KRK RockIt RP6 monitors and Sennheiser HD448 headphones.

 

 

This might be an issue.

You won't get an accurate image with headphones and

2nd RP6's are extremely colored and don't reproduce music

with accuracy I would think a discernible listener/ear would need

as a requisite to comment on high resolution sample playback.

 

 

 

In closing Dmitry you must know that all controllers

are not equal. Velocities are inconsistent on most.

Velocity curves across the board are incongruent...

 

^^^This can lead to a premature conclusion that a piano

library lacks this, or doesn't convey that...

 

We can easily agree that some controllers actually OMIT

certain numerically assigned velocity figures (some don't reach 127)

while others can skip 96 altogether.

 

There was a thread I participated in

(at another forum) which I can no longer find, however in

subsequent forums (relative to piano VSTi consumers and their choice

in controller) every user almost reported having issues which ranged

from going from pianissimo to fortissimo without any connectivity

from these two types of playing...to not being loud enough, or

being too loud.

 

Yamaha controllers were found to be a common derivative of

the issues aforementioned and although it is no where near

a conclusive stereotype of these anomalies that plagued

Yamaha controllers they did outnumber every other controller listed.

 

I have a Korg Triton Extreme 88, and through

light tweaking between the velocity curves on the AK's interface

and my own Triton's edit parameters I was able to find the perfect

balance....Though YMMV.

[DmitryKo]

 

this emulation is ahrdly anything beyond 500 , Alchemy can only reach up to 600 but this may be alot less than that. Can you spot which one is the real piano and which one is the synthesizer ?

Interesting. No, of course I can't, this is indistinguishable (putting aside the quality of the sample).

 

My point is, I did not doubt that extracting 200-500-1000 harmonics with Fourier transform will work, if the synth engine supports that much oscillators. Actually I believe this is quite close to Yamaha's "Spectral Component Modelling" which is used on the latest CP-1/5/50 series, even though it probably doesn't use hundreds of pure sine waves, most likely a dozen "harmonic layers" which dissect the sound into complex spectral components. The same approach is probably used for their "FM emulation" in CP-1; I also heard an opinion that they use a reincarnation of FDSP technology found in the EX-5.

 

 

The practical question was if typical 2-4-6-8 oscillators or non-typical 32 oscillators are enough for a faithful representation or acoustic piano using pure sine waves. I was told that V.A.S.T. synthesis does not need quality samples, or does not need samples at all, as it can enhance samples of real-world instruments using its enormous processing power alone. However I still doubt that even 32 harmonics (pure sine waves) are enough for a convincing representation using only the virtual modelling approach, especially if you consider practical polyphony limits of PC3X. And since what I hear in PC3X piano demos is not as top-notch, IHMO there is no practical substitution for quality samples, whether they are samples of whole notes or some spectral components artificially constructed in software.