• Register
  • Help
Page 1 of 5 1234 ... LastLast
Results 1 to 10 of 50

Topic: What makes a Grand Piano library Great ? many things. :)

  1. #1

    What makes a Grand Piano library Great ? many things. :)

    What makes a great grand piano sample library ?
    A lot more that the actual number of recorded velocities goes into making a great Grand Piano sample library, Kip\'s Bosendorfer and our Steinway Upright are a great example of this with only 4 velocities pedal up and pedal down, both of these sound wonderful and can hold their own against any other sample set even if they have twice or 4 times more velocities, you can also add Michiel Posts 2 great Steinway and Bosendorfer piano libraries to that list.

    You see first off there\'s the actual piano being recorded, if a Grand piano sounds good, it should sound good totally unprocessed to begin with, our Yamaha C7 and Steinway Upright come absolutly unEQ\'d and without any processing whatsoever except normalizing to ensure equal levels across the entire velocity range, if you got a good base you can then EQ that piano to your own preference you aren\'t stuck with a developer\'s idea of what it should sound like, you can EQ it to your own preference , it\'s comforting when you know you\'re working with the raw and untouched sampled material, but some other might prefer to have a premade sound and that\'s fine too [img]images/icons/smile.gif[/img] .

    Next comes the actual recording equipment, tuning of the piano and repairs, the recording space used and level care put into the actual recording by the actual person who does the recording, again if the samples are well recorded ,minimal and I mean really minimum audio processing is needed to achieve sonic excellence.

    These are crucial to the end results, but again \"crap in crap out\" as they say, you need a good piano to start with. Next comes the actual editing of the material and the choice of using loops or not, we do not beleive in using loops but if you want more velocities you are going to have to use loops. Our Yamaha C7 has 6 velocities pedal up and Pedal down and each recorded perspective is 4 Gigs in size, this is about the limitation to date for a Grand Piano instrument size, our next project will be 8 velocities pedal up and pedal down unlooped this will probably bring each instrument\'s size up to the 6 gigs , we also beleive purchasers should be informed before hand that they are buying looped material that\'s why if we ever use loops in any of our future libraries, the costumer will be informed of this fact before making their purchase. Is there a big difference ? not if the loops are done well enough, but you should know this before purchasing the library.

    Does 6 recorded velocities give you better dynamics than 4 ? yes, does 16 velocities give you more dynamics than 8 or 6 ? yes, but a well recorded 6 velocity instrument can sound as delightful and offer great dynamics without having all those velocities as well, also remember than using more velocity samples means a bigger cpu hit since the computer is loading more samples in it\'s buffer, but that doesn\'t determine everything you should be thinking about before buying a new Grand Piano Sample library, a lot of other factors determine what\'s a good value and that starts with what kind of piano sound and brand you\'re looking for, do you want to be able to process the piano to suit your needs and what does the library offer for the price it\'s being sold at.

    Our Yamaha C7 library offers you 3 wonderful 4 Gig sized instruments that could easily be sold seperatly at the price some other libraries are selling just one perspective which is usually close miked or a mix of ambient and close miked samples.

    With our Yamaha C7 library,

    You get close miked : 2 Microphones positioned inside the Grand piano very near the soundboard placed strategically in the same way most of the Top recording studios place their mics to record Pop and Rock recordings with this piano using the best microphones and the best preamps in the industry.

    Ambient : We take it a step further, you also get a Ambient instrument which are microphones recorded at around 6 to 10 feet from the actual instrument on the side where the audience would usually listen to the piano but much closer.

    Player Perspective : Even more, you get a third instrument sampled with a Stereo microphone place at the position where the head of a player would be when playing the piano, this is wonderful to create spacial effect for film use, it also sounds wonderful just playing it in front of the speakers, you can really fell like you are there.

    All this for $149.95 during February, we beleive this is a great value for the price.

    Again, 16 velocities may offer you better dynamics but there\'s a lot more to choosing a Grand Piano sample library than the number of velocities. Also, if you strip away 4 or 8 velocities from a Piano instrument to see the difference with more or less velocities, the results will be much more dramatic with a instrument designed to be played with 16 velocities since it was created and edited to be played with 16 velocity samples.

    If you do the same test with a piano designed to playback 6 or 4 velocities you might be very impressed with how much dynamic response you actually get from the instrument since it was designed to respond to fewer velocities.

    These are all personal observations of mine on the subject of \"What makes a great Grand Piano Library\" and should be read as such, other opinions will differ, in the end I guess we choose Grand Piano sample Libraries for what we think they will bring us and in the hope they will inspire us to create wonderful music, so there is no \"best\" or \"best ever\", only what you perceive to be a Grand Piano library that will help you in your creative endeavours.

    Keep making great music [img]images/icons/smile.gif[/img] !


  2. #2

    Re: What makes a Grand Piano library Great ? many things. :)

    Do people in general think the huge sizes of the pianos are a problem? Probably a stupid question in this forum but I have never used anything above 10-20 MB myself. If I was looking for a piano, I\'d love to have 64 or so velocity levels, 16bit samples and 5 octaves only. (4 and a half would actually be fine to). That to save space. *Wasting* space could be done for the velocity.

    With 10sec sustain, that would be about
    3 386 880 000 bytes

    Is this a common wish or am I just strange?

  3. #3

    Re: What makes a Grand Piano library Great ? many things. :)

    Good question Hakan !

    I get what you\'re saying, maybe that would be a good option. The thing is many developers try and offer the full ranged instrument sampled as deeply across the entire keyboard, but it would be interesting nevertheless to see more velocities in a shorter span and hear the results, good idea. [img]images/icons/smile.gif[/img]

    But as computers catch up, we\'ll be able to have more and more velocities of samples unlooped, we\'re getting there but slowly.


  4. #4

    Re: What makes a Grand Piano library Great ? many things. :)

    If the filters are used right I think 1 layer can sound better than 4. Of course I prefer 16 layers but when I don\'t have the power to load those huge pianos I most of the time pick a piano with 1 layer only. I really hate when you can hear the velocity splits when playing...


  5. #5

    Re: What makes a Grand Piano library Great ? many things. :)

    Hi Tobias, I was just about to write that in response to Richard\'s post but you beat me to it. [img]images/icons/smile.gif[/img]

    You can get a piano with only 8 velocity layers to sound just as interesting dynamics wise by using filtered layers apropriatly, it won\'t be as smooth as 16 but will be dynamically very interesting, so to me the lower velocity sampled grand piano libraries we have today are still pertinent in today\'s market.

    Very interesting points guys.


  6. #6

    Re: What makes a Grand Piano library Great ? many things. :)

    It\'s a big subject Another thing that is annoying is when the samples aren\'t cut tight. Seems like some developers tend to not care about those ms of silence before the actual sound sometimes.. I mean, I got enough of latency in my systems.. [img]images/icons/wink.gif[/img]


  7. #7

    Re: What makes a Grand Piano library Great ? many things. :)

    Yes Tob, this is what took me the longest to edit with the Yamaha, I made sure every single sample start time was perfect, this is very importand when designing Grand Piano libraries.


  8. #8

    Re: What makes a Grand Piano library Great ? many things. :)

    I\'d say a good sampled piano should not let hear the velocity splits, as a developer would record more than the needed layers and choose 4 (or more) out of the pool that fit best. With 16 velocities and a few more recorded this should not be a difficult task, and I\'d say if 16 velocities are well done it\'s not possible to hear because of the greater overall \'variety\' (subtle nuances). It was best if a developer used a machine striking the keys (I saw this once on the internet), that gives you equal power for all keys and if the real piano is in a good condition also constant levels. Then it should be even easier than if you record it manually. Power issues should be taken care of with a lite edition of the patch, that the detailed one can be rendered later.


    PS: I\'m wondering which piano libraries are pre-EQed or looped and not telling the customer?

  9. #9
    Senior Member Bruce A. Richardson's Avatar
    Join Date
    Sep 1999
    Dallas, Texas

    Re: What makes a Grand Piano library Great ? many things. :)

    I\'d say that the 16-layer design is a pretty good thing, though. It\'s definitely significant. You can get the same playable dynamic range with three layers, heck, even one. But what\'s missing is the emergence of partials that define the dynamic at a given level.

    If you take the most extreme example, a one-layer piano, the dynamics are being completely generated by the velocity response of the sampler engine. In essence, barring any filtering, the volume is just getting turned up or down depending on how hard you hit. Since the overtone content of each \"hit\" will be the same, with only the volume altered, this gives you the infamous \"piano moving front to back\" imaging scenario.

    The more layers you add, the less this scenario plays out, because the aural clues about playing force start to resemble the behavior of dynamics more than the behavior of placement. As overtones come in our out, our brain is fooled into believing that string really is being struck at different velocities.

    If you want to idealize the process, the formula is simple enough. How many dynamic layers per note are required to model the behavior of the perceivable overtone series in the instrument\'s performable dynamic range?

    Scarbee pretty much nailed a Rhodes at 12. His Wurlie is 16, and it\'s also nailed. The acoustic pianos are another story. The stiffness of the upper register has played out in favor of fewer dynamic layers...eight seems to cover that fairly well, and you don\'t get too many perceivable imaging artifacts with even four, as long as you\'re not putting it out there bare.

    However, I think the proof of concept offered by Worra\'s White Grand is pretty unassailable as it applies to an ideal design. Sixteen seems to be the magic number. The lower ranges, in particular those below C3, are pretty drastically improved in smoothness over pianos with fewer layers. It is noticeably smoother than an eight layer bottom end. And it makes a lot of sense. Because the lower register is \"looser,\" and becuase the number of perceivable harmonics is greater when we start with a lower fundamental, it explains the sensation of more control, and the end result of a lot more stable imaging.

    You can compromise by filtering to a degree, so that the number of steps is virtualized to somewhat mock the brightness or dullness of timbre within a given dynamic layer. Michiel did this with a couple of his pianos. But it\'s not quite the same as literally hearing the partials speak out. There are still some skips in overtone emergence perceivable, despite the virtual timbre adjustment. One would almost have to design a resonance filter, rather than a lowpass, to do this type of emulation. It would be quite consumingly difficult, a note-by-note thing. I am pretty sure people have thought of this before me, and just opted not to drive themselves insane trying to implement it. The lowpass method does the trick just fine in all but the most nitpicking, exposed situation.

    Don\'t get me wrong...I don\'t have any particular dog in the fight as far as marketing a specific instrument goes. But I would say that the advantages of more layers, as a design parameter, outweigh the negative aspects of looping. Funny, too, because I wouldn\'t have said that unless it had been proven to me.

    I also think a good many of the fewer layered pianos sound excellent, and I still reach for them as often as not, depending on the sound needed. There is definitely a palpable advantage to the 16-layer design, though. It seems to be the \"break point\" for getting stable \"distance\" imaging without filtering, most noticeably in the low register. The playing experience is also quite noticably more controllable. You don\'t get surprised, where I have found myself, especially on 3-4 layer instruments, needing to go back into a track and \"un-surprise\" a given note that leaped timbrally within a line due to a velocity switch.

    Interesting topic.

  10. #10

    Re: What makes a Grand Piano library Great ? many things. :)

    I\'m curious as to how you sample designers think when designing the piano library. Do you come up with your velocity layers with the idea of a somewhat evenly distributed pp, p, mp, mf, f, ff, or do you design based on what velocity range most people would play in? For example, I know I can hit a level 127 frequently when I get loud, but it is actually quite difficult to \"play\" a level of 1. I would think it would be best to put a greater weight of the velocity layers to the upper range. I guess in some regard, that would give the appearance of more layers as they would be tighter in the average playing range. How do you usually approach it?


Go Back to forum
Page 1 of 5 1234 ... LastLast

Tags for this Thread


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts