• Register
  • Help
Results 1 to 4 of 4

Topic: New demos of Galaxy steinway with Xavier B. Soluce

  1. #1

    New demos of Galaxy steinway with Xavier B. Soluce

    Olivier FRAPPIER special version of Galaxy\'s STEINWAY 5.1 :
    05 22 2004 (May 2004) Version 2.01 beta RC1 -
    available soon for download if Xavier is agree (Because He give me a lot of inspiration/ideas...)

    1 ) pedal effect :
    Real pedal down behaviour - more sound/harmonics when the pedal is down - less when up)
    (M.Post idea)

    2 ) String release

    3 ) natural sound - with less lowpass filter or bypassed

    4) symphatic resonance of the non-struck singing strings.wav created with all harmonics of existing Galaxy steinway - programmed in one group for pedal down
    (Olivier\'s invention)

    5) real string behaviour with pedal down (Xavier Bidault idea)

    6) saving polyphony about 40% or 60% (Xavier Bidault idea)

    (http://jazzphoton.free.fr/ -> Xavier bidault examples)

    Programing -> O. FRAPPIER (i apply the Xavier\'s rule [img]/ubbthreads/images/graemlins/blush.gif[/img]ne key,one group)



    Midi of demo 4 :
    midi file of the directory demo of the galaxy Steinway5.1

    midi of demo 3 :

    midi of demo2 :


    midi of demo 1 :

    What Xavier wrote :

    I wanted to recreate a real string behaviour with pedal down. When you play one note on a piano, you play only a (one, a couple, or three) string. And when you use sustain pedal while playing again this note, the previous one “disappears” to let the new one sounding. This is not really a problem when the second note is louder than the first one, but it is when the second is softer! On each actual patch, the first loud note keeps on sounding and expressivity is very poor. So i made a modification to gain this realism while playing.
    And this improvement has the good idea to save polyphony... no need to have 256 voices to avoid loosing notes in rich playing (for example : a bass note keeping on ringing although the polyphony limit should have cut it a long time ago).
    I improved this for each note, and managed not to destroy the sympathetic resonance but keeping the reinforcement effect, and still with realtime sustain.

    What M. Post Wrote :
    (Michiel Post,
    Creative Director PMI
    www.postpiano.com )

    Repedalling is an issue so often overlooked.
    on a real grand piano you can press the sustain pedal at any moment and the strings will start to generate extra harmonics caused by the singing of the non-truck but resonant strings. This occurs also when a note is already playing and usually the pedal is pressed AFTER the initial note start. In the traditional samplers (like GigaStudio2 and every other application for that matter) you need to press the sustain pedal BEFORE you play a note in order to get the resonant notes to play. This is because the sustain pedal is a simple on-off switch that operates the sustain dimension. If the pedal is up you select sample \"NOTEXpedal UP\" to play, if the sustain pedal is down you hear sample \"NOTEXpedal DOWN\".
    This is insane for a real pianist. It’s like starting a vibrato before you start singing.
    What you need to hear is that when you press the pedal there is resonance coming in. If you press the pedal while you have notes already sounding this also has to occur. In the conventional way pianos are programmed nothing happens except that the notes sounding will be sustained. But no extra resonance will start.
    So re-pedalling lets you switch between the sustaining and non-sustaining samples in real-time. It gives a smooth crossfade between the two samples on the exact moment you press the sustain pedal, just like in a real grand piano. The difference with the conventional sampled piano behavior is amazing: you get more expression, more realisme and an overall more natural sound. I have created a special page about the effect of re-pedalling on my website:
    The page has screendumps and sound clips that illustrate the effect of re-pedalling.

    There are several ways to achieve this effect. The first way is to layer the sustain sample under the non-sustained sample. When you operate the pedal the sustain layer is allowed to fade in. Simple, but it doubles the polyphony load for the sample engine, since you always play two stereo samples for each note (wheter the sustain sample is audible or not). The second way is to add the resonance with convolution. Thist is where GigaPulse comes in! GigaPulse can capture the body resonance impulse. This is an abstract term; it is a calibrated recording of the body resonance of the piano that was recorded. It will respond to the notes that are fed into the convolustion engine by adding the corresponding body resonance for that note. The body resonance IR did not capture any sound from the strings; it is the isolated body resonance sound, induced by special techniques. Upon playback of a sample you set up GigaPulse to add the body resonance when the sustain pedal is used. Also simple but very realistic given that the body resonance impulse is a perfect representation of that piano body.

    BTW for the new pianos I also recorded all sustain pedal down velocity layers with sustain pedal down for the traditional ON-OFF switched methods of sustain pedal behaviour (AND for non convolutionized sample platforms).


  2. #2
    Senior Member
    Join Date
    Aug 2002
    Castle Rock, Colorado

    Re: New demos of Galaxy steinway with Xavier B. Soluce

    This is really good work Oliver. I have Galaxy Steinway and would love to have these updates. Very forward advancing stuff - I loved the piano before the modifications but I think now when and if these updates can come into being that the Galaxy will most likely be my first call piano.


  3. #3

    Re: New demos of Galaxy steinway with Xavier B. So

    Hi Olivier,
    just a question regarding the 5.1, do you (or others) have any benchmarks when you enable full 5.1 playback ?

    If my math is not flawed assuming the 24bit WAV files contain 6 channels (5+1) playing let\'s say 80 voices would place an unsustainable load on the harddisk.
    80 voice equals to about 63MB/sec which cannot be achieved by a single drive (usually only a sustained 20-25MB/sec can be achieved because seeking through many active streams lowers performance, if we were playing a single stream then probably up to 40MB/sec could be achieved but not in a high polyphony softsampler enviroment).

    I guess for satisfactory 5.1 playback you need a RAID disk of at least 2 10.000 rpm disks.


  4. #4

    Re: New demos of Galaxy steinway with Xavier B. Soluce

    Sure! I agree.
    It\'s a very new way to program patch and i would like this to be used on every sampled pianos, and specially on all these great ones coming soon.
    Vive kontakt 1.5.2!!!

    But i really wonder if kompakt can support all the kontakt tweaks (for example release in the ADSR controled by velocity with unusual response active, or other parameter controled by inverted key position, or controled by another parameter with response in V form...)


Go Back to forum


Posting Permissions

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