A SY99 voice consists of 1 or more (up to 4) voice element(s). A voice element can apply the Frequency Modulation (AFM) synthesis method or the sample playback (AWM) method. Cross-modulation, in which an AWM element can be used as an operator in an AFM element, is also supported.
This concept is very similar to that of the Synclavier II. In spite of that similarity there are huge differences in user interface between the SY99 and the Synclavier. Further the envelope generators and filters provide by the SY99 are significantly more advanced than those in the Synclavier.
Each single SY99 voice can use one of the following combinations of AFM and AWM synthesis elements:
|Voice Mode||AFM elements||AWM elements||Polyphony||AWM / AFM cross modulation|
A SY99 drum voice has no AWM or AFM elements, but instead consists of a different audio sample for each of the keys of the SY99 keyboard.
The scope of this section, working with samples, is limited to voice types that use one or more AWM synthesis element and to the drum voices. A drum voice has a different audio sample, if any, associated with each of the 76 keys of the SY99.
|MIDI SDS property||SY99 specific remarks|
|Multiple channels of audio (usually 2 for stereo).||
The SY99 supports only 1 channel of audio (e.g. mono recording) in
each of its samples.
It is possible to upload a multi-channel sample to the SY99, but then only the 1st audio channel in the sample will be stored.
Quality, e.g. the accuracy with which the audio is
recorded. Common accuracies are known as 8, 12, 16 or 24 bit
16 bit accuracy is CD / DVD audio quality.
24 bit accuracy is modern recording standard.
Note: Samples with a higher bit resolution require, obviously, more memory storage.
The SY99 supports accuracies from 8 to 16 bit.
From experience I can tell that samples with accuracies of 8, 12 and 16 bits will be uploaded to, and used by, the SY99 successfully.
Be not surprised when samples with other bit resolutions give unpredictable results.
Sample frequency, the number of measurements per second of
the original audio signal.
A sample frequency of 44.1 kHz is CD / DVD audio quality.
Sample frequencies of 48 kHz or even 96 kHz are not uncommon these days.
Note: Samples with a higher sample frequency require, obviously, more memory storage.
The SY99 supports sample frequencies up to 48 kHz, which was
exceptional for its time (1990).
From experience I can tell that samples with frequencies of 22.05 kHz and 44.1 kHz will be uploaded to and used by the SY99 successfully.
Loop start: is fixed to the beginning of the audio recording
Loop end: is fixed to the end of the audio recording
Following loop types are supported:
Overview of SY99 sample and MDR RAM sizes
SY99 with full memory extension
The picture above shows a SY99 with full memory extension: 2.5 Mbyte RAM in addition to the internal 512 kbyte RAM. The rightmost slot (slot #1) holds an original Yamaha SY99 memory extension module. The leftmost 4 slots (slots #2-#5) hold sector101 SYMB05 modules. Each of these modules, whether Yamaha or Sector101, offers 512 kbyte battery backed-up RAM.
An absolute maximum of 99 user samples can be loaded into the SY99 RAM. The actual number of samples that can be loaded may be smaller, depending on:
The most reliable and easiest way to upload user samples into the SY99 is as follows:
Each AWM element in a SY99 voice has exactly 1 "AWM Waveform" An AWM Waveform is not identical to an audio sample. Instead it is a collection of audio samples. To be able to use a sample in a AWM element, two assignments must be made:
The SY99 has 64 user editable AWM Waveforms in addition to its preset AWM Waveform and the AWM Waveforms on a waveform card.
Initially the process of assigning samples to these AWM Waveforms is rather confusing and tricky. Main cause for this is that this process is a system-wide task while the edit window for the assignment can be accessed only via the AWM edit section of any individual voice ...
This becomes somewhat more understandable when applying "reverse thinking". Suppose you want to edit a voice or create a new voice and use one of the loaded user samples in it. Then the following would be natural steps to take:
Now comes the ultimate tricky part, in the rules that apply to memory management of the user samples:
Each sample in an AWM Waveform must be assigned to a specific keyboard
region. A special rule applies to this assignment:
The high key of a sample in a Waveform is always one semitone below the low key of the next sample in the AWM Waveform. It is not possible to overlap samples in an AWM Waveform or to leave gaps between them.
What is somewhat confusing is that the on-board editor of the SY99 does not enforce this rule. It allows you to place any samples in the AWM Waveform in any region of the keyboard resulting in undefined behavior of the writable sections (if any) of both the waveform memory and the user sample memory.
Note that it is not mandatory to assign a sample to every area of the keyboard, as long as the above rule is satisfied. The AWM element will not produce sound for keyboard areas that do not have a sample in the AWM Waveform assigned to it.