Dante: For some
time on PlanetZ there has been discussion regarding
the performance of the Solaris soft synth on the XITE-1. Some XITE-1 users
raised concerns regarding using the flagship on the XITE-1
because it didn't yield polyphony equal to what could be obtained on
a PCI card system.
John Bowen: While I do understand the frustration of
not being able to run my bigger plug-ins (Solaris,
Quantum Wave) with more polyphony on the XITE systems, the situation
is that these two plug-ins require more connections internally than
the XITE system can provide so there is no way to optimize them
further. The only thing I could do would be to reduce the
number of modulation options for each plug-in, so that they would
not be anywhere as flexible as they are now (which is a big part of
their appeal, it seems). I would have to make new versions,
severely reducing the number of connecting pathways internally for
them to give greater polyphony but then you wouldn't have all that
the synths are capable of.
Dante: From this I concluded
that if internal rework would reduce the synthesizers capabilities
it would not really be worth it, since the Zarg range already
covered sonic territory with less parts in other synths. I
asked myself whether reducing the complexity of the presets
themselves could achieve higher usability. I set about
checking out just what made many presets so complex, and discovered
that there was some redundancy in many of the presets. Certain
presets contained components that could be switched off completely
without adversely affecting the sound. I set about creating some test presets on my
XITE-1D and then asked if they could also be tested on a full XITE-1
and replies soon followed:
RP_001: I could go up to 8 or 9
voices and save and recall a project with the same amount of voices
without error or asking for optimizing the project. I also
succeeded switching through your other presets with 6 voices and in
some case 8 voices, which may take some time though so it may be
'saver' to reduce the amount of voices and then switching through
other presets. Tested on Solaris
4.1 which, I believe, has less parameters then the 5.0 version.
I will check out the 5.0 later this week.
here, I can load all your presets with 6 voices at 48Khz without any
problems. As usual, first Solaris
load has been done with 0 voices then it loads the presets fine and
the DSP allocation is 1 voice per DSP. I tried up to 8
voices with success, with the caveat that at the upper limits the
loading times become slow and sometimes it needs to be set back to 0
voices to solve issues.
Dante: Thanks for the
tests guys. So on the full XITE-1 we can get 6 voices without
issue and it's not until 7 or 8 that problems start to occur. It seemed that although the
caveat of preset surfing at low polyphony still applied, that the
general idea of optimizing factory or original presets to achieve
good polyphony on the XITE-1 was a reasonable one. So I set
about expanding and tuning my preset bank and ensuring it would run
with reasonable polyphony.
Some presets where derived
from those I had created for Solaris 'Core'
from factory banks. I scanned through all the factory sounds,
taking the ones I liked and stripped them down by turning
off less used features. The result is a bank of mostly pads
running at four voice polyphony with ASIO, mixer and FX loaded into
the 48KHz XITE-1D Scope project as well. In general, they run
at maybe one less voice polyphony than on the
Some hints for getting more mileage
out of Solaris v5 on the XITE-1
1) Use the small LED bar at top to switch things out
2) Program more conventional (bread and butter)
3) Avoid including external wavs or
files that don't exist causing errors.
4) The flanger with a fully right clockwise phase is
a great device for widening stereo pads, and inexpensive DSP wise.
5) If you really need more poly there's always the option of
reducing sample rate to 44KHz, which is still CD quality after all.
6) Add an
Optimus Prime (demo below).
In summary many of the factory presets seem centered around
exotic / soundtrack type sounds, with bread and butter type sounds
somewhat lacking. Also there seems many presets where some
components can be switched off without any detriment to the sound.
In reality how many of us would actually use those exotic
presets often? Sure they are great to play with whilst doing
that initial preset surf or to impress your friends, but day to day,
unless you're providing a backdrop to a rising paper mach'e model of
Stonehenge for the local band or ambient relaxation music for
anesthetized dentist patients, Buddhists or bored housewives then
you're more likely to be using more conventional sounds. This
is not to say there is not a place for complex sounds but for those
of us who are producing movie soundtracks it may be best to keep the
PCI cards for which the synth was designed.