I\'m interested to buy Gigastudio 96 or 160. But when I tried it at a local musicstore, I found that when using for ex. a Piano sample while holding down the sustain pedal, the same note (ex. C3) would incrase the used voices each time you use the same note. Is this normal ? If so, it\' useless in my setup as the piano sound would eat too much polyphony. Anyone ?
This is the way ALL synthesizers and samplers work.
In order to hear a sound you activate a \'voice\' circuit, which uses its own dedicated envelopes, filters, amplifiers, lfos etc.
In order to get discrete voices, a new voice circuit is triggered every time you hit a key (unless in solo mode - and that\'s another story).
This also applies for revoicing the same pitch repeatedly.
If you play a note and let it decay to 0, when you fire that key again it will not use more polyphony.
Layers which are mixed, crossfades, and overlapping notes all add up in terms of polyphony. That is why it is so GOOD that the Gigastudio has 160 voices. Most samplers max out at 128, many at only 64. Piano modules mostly don\'t offer more than 64.
On the other hand, If you have a 30 piece violin section, one key can use one note of polyphony while it sounds all thirty violins.
As far as the particular piano sound you\'re using, it may be possible to cut down the polyphony you\'re using by changing the nature of the patch. Maybe turning off release triggered samples, and checking for layered samples which are mixed rather than switched between.
[This message has been edited by Chadwick (edited 08-17-2000).]
>If you play a note and let it decay to 0, when you fire that key again it will not use more polyphony.<
I don\'t know really what patch it was, as it was in the music store, but I saw 2 voices of polyphony for one note strucked. But while holding the sustain pedal, strucking the same notes (pedal chords) brought me around 80 voices quickly and as the PC used was just a PIII450, ckracking ockured. Doing this on my Alesis QS6.1 doesn\'t give me this problem. I can play for minutes the same chord in pedal mode without reaching the polyphony limit. That\'s why I asked this question.
Also, what\'s the best CPU to use with Gigastudio to get up to 160 voices? I heard the latest 1G thunderbird are great. How many ram should one use ?
I know you don\'t easily hear polyphony max out on the Alesis, but I would bet money that if you could somehow tag the Alesis\' first piano voice with a \'ding!\', that eventually you would hear that ding again as you keep playing the same note with the sustain pedal down. What these machines usually do when they reach the maximum polyphony is recycle the oldest note. The trick is that if you have 64 notes built up, all playing a sustained C3, you don\'t notice that voice number 65 is actually voice number 1 stolen and replayed. One note out of 64 isn\'t that obvious. Polyphony max out and voice stealing becomes most obvious when the voice which is robbed just happened to be playing an essential drum loop sample, or was the lead voice in a part.
In Roland\'s JV 1080 they even let you reserve a minimum number of voices for a particular part. Sometimes if I\'m stuck with note stealing I\'ll prioritize my arrangement and reserve some notes for the most important parts. It\'s a balancing act because those notes stop being available to the other parts.