YAMAHA SY85
Overview
The second generation of SY
SY77/SY99's FM synth section is discarded in SY85, but the polyphony is doubled to 30. The Waveform memory consists of 6MB ROM and 512KB RAM.The sample RAM can be expanded up to 8.5MB.
SY85 Basics
SY85 has 3 Synth Modes - VOICE, PERFORMANCE and MULTI Modes.
The VOICE Mode has a simple architecture; 1OSC-1Filter-1Amp-1LFO-2FX. Layering is not available in this mode.
There are 5 Filter Modes - 4-pole resonant LPF, 4-pole HPF, 2-pole BPF, 2-pole BEF (Band Eliminate Filter), and 2-pole LPF.
In the PERFORMANCE Mode 4 Voices can be layered.
In the MULTI Mode SY85 becomes 16-part multi-timbral. A Voice or a Performance can be assigned to each part.
SY85 has two effect units which uses the same DSP as SY99 and SPX900.
To me, what characterizes SY85 is its waveforms. To put many waveforms in 6MB ROM, it seems YAMAHA chose to use very short loops. The raw ROM waveforms sounds like deeply processed with compressor. This gave unique sound character to piano, bass etc.
MIDI problems
SY85 is one of YAMAHA's earliest achievement of PCM synths, and it has several MIDI problems.
Filter controls
All SY series except the last one, TG300, suffers from insufficient filter control. Filter envelope and LFO/aftertouch are mutually exclusive. That is, when the filter envelope is used, LFO takes no effect. When the LFO is used, the filter envelope can't be used.
Voice/Performance
Usually PCM synths need layering to make new sounds, but SY85's Voice/Performance Mode complication makes editing complicated.
Deep editing is possible in the Voice Mode and not in the Performance Mode. And layering is possible in the Performance Mode, but not in the Voice Mode. This strict splitting architecture resembles that of KORG M3R.
One good feature is the Layer Voice Edit function. By getting into this menu in the Performance Mode, you can edit each layered Voice while staying in the Performance Mode.
Layer Voice Edit and System Exclusive
I tried to make an editor that uses this Layer Voice Edit function - that is, an editor that easily controls all 4 layered sounds deeply without the Voice/Performance Mode complication. But I found problems of SY85 MIDI System Exclusive.
The manual says that SY85 does not receive Voice SysEx parameter change in the Performance Mode. It seemed impossible to edit via MIDI the deep Voice parameters while layering the sounds.
But I found SY85 does send MIDI SysEx of Layer Voice Edit, though not documented in the manual. I implemented this SysEx in my editor, but it didn't work.
I finally asked YAMAHA about this problem, and their answer was: SY85 only happens to transmit Layer Voice Edit SysEx, but does not receive it.
If we consider SysEx as a language, it's the world's lonliest one - just flows out of the mouth, but can't be understood even by the narrator himself.
Voice SysEx
While making the SY85 editor, I found my SY85's SysEx doesn't work properly - the SysEx length of the Voice is one byte shorter than shown in the manual. I read the MIDI data and found one of the filter cutoff frequency scaling parameter is lost in my SY85. This often results in near muting because filter is closed at the higher keys.
Now I asked YAMAHA again. They said my SY85 may be broken, and it would take $100 or so to see if it can be fixed.
While I didn't feel like paying $100, from the answer I understood there seems to be bug-fixed version of SY85. It might be better to get the later version of SY85.
Next 
SY85
Deep Synthesis: Home
10. 15. 2005