http://www.purgatorycreek.com/

Listen to the piano sample on this site done by

William Coakley 'Perfect Piano 'Series which are 16MB samples or his 'Ultimate piano series which are 32MB in size , then listen to the Roland RD700sx 'Ultimate grand' File size unknown but you can bet its more than 16MB or 32mb . Then listen to
Sweetwater's Ultimate Grand Piano "Yamaha" which is just 16mb in size. Finally listen to
Native Instruments Kontakt 3.0 "Concert Grand" which is a 1.96GB sampled concert grand piano. Then post up your own opinions about the differences in in sound and texture and whether the argument about large VST sample size hold as much water as you would imagine it should.

My favourite Piano VST http://www.truepianos.com (Professionals would probably disagree as Ivory http://www.synthogy.com/products.html is the standard Piano used in virtually all studio and film production)
Also take a look at the fully modelled piano (No samples at all) being developed by http://www.pianoteq.com/index and for a fully modelled Guitar (As well as others) http://www.sonic-core.net/en/products/soniccore.html (And yes, the scope sound engine and DSP card is the one used in Wersi OAS 7)
As I have said before, 10 years from now all sounds will be modelled, with samples (Of any size) being consigned to the history books.
In fact 10 years from now, people will asking how on earth we ever considered today’s sounds to sound anything like the real instruments.
Roll on the future

Bill

yes i hear you Bill but the point i was making is that the Kontakt is many many times larger than the 16mb or 32mb sampled pianos. Does the gap in the performance and playability of the sample reflect the huge difference in Sample Size ?????

My ears tell me the answer is no. So why would anyone make the ability to take huge samples within an instrument the main selling point of the instrument ?????

They wouldn’t, however it is nice to have the option without requiring a degree in studio engineering or computer science to use them.

NI (Makers of Kontakt) has never been held in high regards with their Acoustic Piano VST, however if you listen to The Grand by Steinberg, or the studio standard Ivory, you will most definitely hear the difference that sample size makes.

Bill

Online demos are NOT a reasonable representation of what the sound quality truly is unless of course you are auditioning a full bit WAV or AIFF file. Unfortunately most demos are low grade MP3's which can mask much of the overtones, resonance, and subtle articulations in a sound that can separate it from the pack.

VST capable instruments such as Wersi can offer a sound palette that no embedded instrument can. There is a great advantage to using a Wersi, Mediaststation, or Neko but it seems many people on this forum choose to disregard these offerings altogether having never tried one of these instruments in person.

I make the offer for anyone who wants to make the trek to come by and audition my Wersi Abacus Duo Pro in person and base your opinions on fact rather than conjecture. I'll bet few if any would walk away with a low opinion of the Wersi after using it first hand. I'm not far from Sacramento, CA for anyone who wants to come by.

how does that make any sense ensnare you ? Yamaha sells the Tyros based on amazing online demos,surprisingly they sound good enough to make buyers mouths water, you posted a link to various VST products using online demo's.(They sounded great by the way just as the T3 demos sound) what do you think the purpose of an online demo is for if both proprietory and VST sounds are used in both????

If the vst suffers in terms of audio quality because of the limitations in MP3 recordings, then so do the Yamaha , korg etc demos. they also sound better in real life as most instruments do.

I am not saying that there would not be better sound quality in real life of any keyboard compared to a recoring. I am saying however that the diference in the fidelity of the sound is not like night and day. A poorly programmed arranger using VST's does not make the keyboard better than well balanced proprietory sounds from yamaha and korg and that can be readily distinguished by MP3 recordings which i believe is precisely why no matter how much it is requested nobody, not the manufacturers and not the enthusiasts of either wersi or mediastion will provide demos of their arranger instruments using the overly touted VST capability. I am not knocking you ensnareyou or abacus and this is not a pissing contest as some would argue. I am simply pointing to the real limitations of claiming that simply because an arranger product can use a VST that it will automatically sound better than a 'closed 'system. Its a great theory but unfortunately no theory is worth a hoot if the reality does not support it.

If a 16 meg piano sounds ok to you, use it. if a 16 gig piano sounds good to you, use it. Bigger isn't always better, but bigger is often better. Good programming is also important. Poor programming with huge amounts of samples probably won't be as good as great programming with smaller sample size.

But, a piano played softly sounds much different than a piano played loudly, and it is more than just something that can be emulated with filters. So using 1 layer and filtering it for dynamics doesn't sound as real as multiple velocity layers. Unless of course the velocity layers are poorly programmed and there are severe jumps from one layer to the next which makes playability difficult.

Finally, if large numbers of people here are duped into believing that Chet Atkins and Tommy Emmanuel are great keyboard emulations of guitar, who really cares what those same people think about the sounds of various pianos. Decide it for yourself.

Spalding ..I got the impression..you were trying to knock the Roland RD700sx piano sample ..more than judging the value of VST's ..

BTW : you would lose that bet about the "larger size" of the RD700sx piano sample..

Total wave memory:

128 M bytes (16-bit linear equivalent)

I also agree that size does/can matter..as FAEbGBD..touched on..the layers of sampled piano is what gives the overall sample the dynamics..this will use more memory [Size)..Sure you can get a great sound with 16 -32 megs..as that Italian grand sample posted..BTW: many consider that Italian brand the best in the world..

For the most part the 3 strike Yamaha, or 4 layered Roland piano will sound good and do the job...but to state that a larger sample would not do a far better job is just silly... 12 or more layers of velocity switched piano samples will show the dynamics at different volumes and attack(velocity) dynamics too..that using a filter by itself will not do..

I still like the RD700 pianos..they are clean and very playable..not as bright and buzzy as some of the others you suggested..

I am using my G70 (the FantomX 'Ultimate Grand' at P.Creek, if you want to listen) in studios that have Ivory on hand.

No-one has ever said 'We HAVE to replace this with Ivory".

We even have a 1928 Steinway. Most of the time, we hook up the K2500X's 88 wood to the G70. Record it dry, and run it through a nice convolution reverb... Perfect!

File size does NOT matter... What matters is the quality of the recordings in the first place, how good the initial piano is (some dogs out there... beware!), how well it was tuned and regulated, how consistent the playing, how well miked it was, how accurate the phase between mikes (there's the main thing, IMO ).

1GB of out of phase, hollow sounding piano samples doesn't come CLOSE to even equaling 64MB of exquisitely recorded, perfectly matched samples.

Once upon a time, sampling a grand piano took unbelievable skill. Memory wasn't sufficient to allow a brute force approach, so what little you DID have to work with needed to be pristine, as perfect as it could be, to overcome the flaws that looping and short samples would invariably add. But nowadays, with GIGA sized libraries popping up all the time, many are content to just sample every note, as many velocities as the can, no loops, multiple mike positions, pedal-up, pedal down, soft pedal, you name it. But what they usually forget is, how does it SOUND?

I've been a Purgatory Creek addict since it started. And, as more and more huge libraries got added, I started to ask myself 'why do they sound so bad?'. Nine times out of ten, what I usually come away with is a sense of 'out of phase-ness', a lack of 'body' and warmth, the inability to be BOTH warm and detailed at low to mid velocities, and bright and clear at high velocities. A few manage it, most do not...

Personally, I am MORE than happy to use a small size piano sample set, with no pedal down layers, with no soft pedal samples, with no different mike positions... I don't need a computer, it has basically zero latency, it comes to the gig with me with no extra hardware, and it keeps some of the pickiest ears in the business very happy.

What more do you need?

Hi Diki

All valid points; and this is where sound modelling comes into its own, as the sound is generated like a real piano produces sound, therefore microphones are no longer required.
Where not there yet, but with the accelerated pace it wont take many years before we are.
Roll on the future

Bill

Well, crystal balls might tell us the future, but for now....

The Pianoteq demo doesn't impress me as much as even Ivory, or in fairness, the good Roland's.

BTW, I'm curious, but, for right now, what modeled sounds can you say are currently the best available? Virtual analog... still not as good as the real thing (IMO). Virtual Hammond... B4 is great, but still not up to the standards of a XK-3c, which uses samples as well as modeling. Virtual Rhodes? Lounge Lizard is pretty good, but I've heard sampled Rhodes that spank it.

For a technology to be 'the future', surely SOMETHING now ought to live up to this hype, or it is just another 'also ran'.

Does anyone remember when FM was going to be 'the future'? Does anyone remember when additive was going to be 'the future'? Does anyone remember when wavetables were going to be 'the future'? Until ONE modeled sound comes out that is head and shoulders above any other form of synthesis, it is merely just another possibility, and by no means the de facto path we will all tread.

Just ONE... Anyone? Anyone...? Bueller....? Bueller......?

Hi Fran. No i wasnt knocking the Roland i threw it in as a comparrison simply because of its reputation for great sounding pianos.

I dont think you'll find that i said a larger sample would not do a better job.

i said 'Then post up your own opinions about the differences in sound and texture and whether the argument about large VST sample size hold as much water as you would imagine it should.' Well do they ????

I also said 'yes i hear you Bill but the point i was making is that the Kontakt is many many times larger than the 16mb or 32mb sampled pianos. Does the gap in the performance and playability of the sample reflect the huge difference in Sample Size ?????' Well does it ?

Well the world we live in is analogue, all sounds produced by real instruments are analogue, and sound modelling (Although virtual) generates sound in the analogue way. (Have you noticed with all the retro keyboards about, how many are trying to emulate analogue instruments?
Its like saying CD is better the Vinyl, it sounds great, but when you compare it to the real world, the vinyl recordings leave the CD way behind. (Again analogue wins out over digital)

B4II: I know a few people who have a genuine B3 (Some even have a collection) and all use B4II as well, and the comments have always been as close as it gets to the real thing. (Preferred over the latest Hammonds)

XK3: All the latest Hammonds use virtual tonewheels (They mimic the analogue sound production of the mechanical tonewheels) with all sounds modified the analogue way, (Just like the original) and as far as I am aware there are no samples used.
Interesting to note also is that they have unlimited polyphony, which can’t be done with samples as the cost would be prohibitive.

Pianoteq: If you had a listen to it 12 months ago, you would not recognise it now, as the transformation is massive, which is why I say give it a few years (And not that many) it will leave sample players for dead.
The future is most definitely back to analogue, just in a virtual environment. (However as you say, there is always the possibility of other technologies being developed)

Bill