I don't own Gigastudio yet, but am awaiting release of GS3 and reaction to it. I have sent (twice) some questions to Tascam seeking advice re specific functionality in GS3 but received no reply. My questions are not easy to explain -- maybe not to answer either. I would be very grateful for any feedback if anyone has ideas. Apologies in advance for the length of this first post...
I am doing a doctorate in composition, with emphasis on 'alternative tuning systems' (quarterones, sixthtones, just intonation etc). As I understand it all sampler architectures (including the present version of Giga) allow the fine-tuning of individual samples -- these can be mapped sample-by-sample to individual keyboard-notes (via a 'region' (GS), keygroup (Akai) or 'partial' (Roland).
In the early days (early to me) the Akai 1000 had only 100 keygroups, the Roland 750 had only 255 partials; with those machines, if you wanted to create a 'microtonal' simulation (say, a violin in quartertones), to achieve the ultimate sound quality each individual region/keygroup/partial had to be created (fine-tuned and mapped) for each note. A four-octave violin patch in quartertones (24 divisions to the octave) alone required 96 keygroups/partials: therefore having a whole string trio (say, in quartertones) simultaneously available for improvisation was impossible -- though it is if you have two keyboards available to copy (say) the violin patch, transpose it up/down by 50 cents, and play each patch on a separate keyboard etc.
At one stage Kurzweil introduced a feature which allowed a selected group of samples to be automatically 'optimally mapped' (avoiding chipmunkization) to a desired scale (quartertone, just intonation, you-ask-for-it). The latest Kurzweils seem not to offer this. For this feature to work reliably, the unit (or software) must determine the precise pitch of each selected sample (therefore require an accurate on-board pitch meter?). In any case, to re-programme every sample-to-region mapping of an orchestral set for a large number of *different* tuning systems (as my proposed research requires) is a task that would benefit from some automation!
I guess no-one on the list has seen GS3 yet, but I'm curious to know if anyone has any ideas about how to approach the task in GS2.X -- or whether there is any indication that GS3 might offer some appropriate and/or helpful functionality. It has been suggested to me that fine-tuning information could be sent from the sequencer, but it is important for me to be able to improvise with many *different* tuning systems (one at a time!) directly at the keyboard (ie., one note per scale step, no matter what the tuning), maintaining high fidelity to acoustic instruments. The tuning systems may be as complex as 41 divisions to the octave, and it is unclear to me how this could be achieved via the sequencer end.
Some years ago I proposed an 'ideal' tuning interface for sampler achitectures and sent it to a number of manufacturers, but I'm aware there's very little commercial gain likely from implementing it. For your info, this is pasted below in simplified form.
Click on button called 'Tuning' - up comes a dialog box, which says:
'New' or 'Load' or 'Edit'.
'New' creates a new mapping of samples-to-region in terms of tuning;
'Load' loads a predefined tuning configuration; 'Edit' edits one.
Suppose we choose 'New' - up comes a dynamic dialog box which offers:
----'Global' or 'Instrument' or 'Instrument-Element' or 'MIDI Channel'.
As you click different options, radio buttons/check boxes become
available or not, eg.:
---- Define tonic keynote on the MIDI keyboard (usually 'C', but possibly 'G' for a violin, etc) in terms of key and pitch (conventional =
name plus or minus cents, and in Hz)
---- Choose Equal-Tempered or Non-Equal-Tempered;
Depending on what you choose, dialogue box changes etc:
---- Define terms: Cents or Hz or Ratios;---- Choose cyclic interval (1200cents (1 octave));
---- Choose cyclic interval (1200cents (1 octave as default, but it
could be otherwise)
---- If Equal-Temperament - choose number of scale steps;
---- If not-Equal-Temperament - choose number of scale steps and exact
sequence of intervals;
---- Assign sampler sounds - user chooses from a list of samples.
For this to work properly each sample contains the true original sampled
pitch (showing cent deviations from equal temperament) as part of the
sample information, obtained by incorporating an absolute pitch meter as
part of the software. Like many DSP functions this could be implemented
non-real-time. Setting pitch offsets would be incredibly simplified with
this feature onboard. Of course, this would only be expected to work
with solo instruments, not choruses or vaguely pitched instruments like
most percussion. The software would then assign samples to midi-note numbers accordingly, using the nearest pitch (avoid chipmunkization). The user is then offered the opportunity to test and override sample mapping if necessary. Templates can be saved and recalled.
All of the Native Instruments soft-synths and samplers provide microtonality in both "wild" and traditional/cultural models. Many of the models are not perfect, since their architecture is always octave-based. But you can work around it if you're aware of the implications (which is sounds as if you're more than versed).
This didn't make it into Giga 3. It was one of my big requests, but unfortunately, one that couldn't be implemented within the time frame.
However, I think the architecture for it may be already implemented in iMIDI, and the only thing required would be the rulesets to make it go. I will inquire and get back to you on that.
It would make sense that iMIDI would be able to do that. It would just be a table of notes in to notes out with tuning offsets.
So Bruce, does iMIDI include an editor of sorts, is it a declarative scripting language, or something else? This potentially ranks as one of the top GS3 features for me. If it's powerful to the point of being dangerous, I'm in.
and many thanks Bruce and Jon for your replies. Sorry not to acknowledge them before -- I'm travelling and have no laptop...
An all too brief search for iMidi seemed to indicate this is Mac only? If its not too much trouble could either of you point me to a page which exlains iMidi? I'm PC-based, generally, but wouldn't be allergic to a G5 if it solved the problem!
I'll look into the Native Instruments thing - thanks.
>>the rulesets to make it go. I will inquire and get back to you on that
I'm extremely interested to know what these rulesets might be and how flexible etc.
Very exciting to see GS3 materialising! Many thanks again.
The problem with 'alternative tunings' imo is that they are fixed -- I use flexible tuning heavily with GS 2.5, by means of specific midi pitch wheel commands and a standard 2 semitone pitch wheel response (all instruments configured this way in Gigastudio). A recent example is the 'Romance' on www.soundclick.com/guglielmo (the music page) -- some people hear it as 'microtones', but it's my intention that it be simply expressive tuning and pure consonances.