KVR Audio

jbuonacc · Post by **jbuonacc** » Thu Nov 26, 2015 9:45 pm

yep, four voice polyphony, so any new voices cut off existing ones.

Mutant wrote:... My demo version timed out long time ago and i am too lazy to install it again (if it is even possible).

unfortunately, no, which sucks. unless they've changed it, even a fresh install will "know" that it's been installed before and go right into the noise bursts (or whatever demo limitation). not sure if there's some way around this or not.

never got a chance to really test out the 101, was pretty busy the couple weeks after i installed the demo.

raymondwave · Post by **raymondwave** » Thu Nov 26, 2015 10:05 pm

But shouldn't 4 voice poly play two overlapping notes even if the keys are not pressed simultaneously?

I just tapped a C2 note with long release and then C5 and the C2 was cut.

Mutant · Post by **Mutant** » Thu Nov 26, 2015 11:20 pm

aciddose · Post by **aciddose** » Fri Nov 27, 2015 3:03 am

It depends upon the logic of the voice allocator.

If you were playing a single voice and released it, then played another and the first voice was stolen it means it does not use "oldest" voice stealing.

It seems to count from the first voice upward checking for a free voice, and immediately checks released voices rather than checking free voices.

A free voice is one which has been released for greater than the release time of the amp envelope. In my synthesizers I have an extra stage after release, "off". When the envelope enters that stage the voice is marked as free/off.

If this isn't tracked it means the only thing left to track is age and gate (on/off status). We can pretty safely assume they don't keep track of age since if they did measuring (gate == off AND age > released at + release time) would be trivial. So all they have is gate.

A "round-robin" voice stealing algorithm doesn't always start from the first voice. Rather it starts from the last voice and increments that voice each time. This will in many cases match the oldest voice stealing algorithm without needing to keep track of the age.

Starting from the first voice and stealing any released voice would be called "first released" or similar. Likely the dumbest possible algorithm other than always stealing the same voice.

A voice allocator is actually quite complicated. There are many parameters involved.

... it could also just be a bug though. For example if the oldest algorithm is used but the age of inactive voices is never incremented it would give every voice age = 0, except for the voice that was played first. In that case it would become the oldest voice and the buggy voice-allocator would steal it as the oldest.

raymondwave · Post by **raymondwave** » Fri Nov 27, 2015 8:04 am

That's a thorough answer, thanks.

Anyone care to try if their copy acts the same? I'm on a Mac and Studio One v3. VST crashes so I used AU.

I own their SH-2 and PROMARS plugins and love them despite their shortcomings. The circular knobs are horror, luckily PROMARS and SH-2 have more sliders than knobs, but System-1 has a lot more knobs, which makes it more cumbersome to use. Also the arpeggiators haven't worked properly on any of these, the first note is either left out or played twice (depending on version, I guess they've tried to fix it).

I like the sound of this, it's quite Roland-like in a good way, but the four voice polyphony - the way they have it - is a bit funky. It's ok for stabs, but for any pads I like to make it's pretty useless.

10 minutes for demoing and having to restart daw everytime is a major pain in the a**.

Mutant · Post by **Mutant** » Fri Nov 27, 2015 1:44 pm

BTW for anyone who did not read the whole thread:
https://app.box.com/s/tte3zbpw0r3h2u9c0cm768rpyfg13g3v
https://app.box.com/s/2r1zvc3xma7ur2x7ovwi6htl5z8pjsur
One of them is System-1, the other is a freeware VSTi (which is way more complex and offers tons of additional sound design possibilities).
That OFC doesn't say that System-1 is bad, only that the current state of free virtual synths is kinda great.

aciddose · Post by **aciddose** » Sat Nov 28, 2015 5:19 am

Although the free synth was released a decade ago in that form

raymondwave · Post by **raymondwave** » Sat Nov 28, 2015 11:22 am

I take it that no one is actually using this? Very interested in the poly behaviour if it's a bug or a feature.

FWIW, I'd take a simple synth with good oscs+filters and envelopes any day over a mind-boggling set of features. PROMARS and SH-2 are good examples. Both are relatively restricted when it comes to features, but PROMARS can create the wierdest noises and beautiful arps. SH-2 on the otherhand has amazing raw sound. Doesn't really need much more than say LFO modulated PWM, filter + res.

crystalmsc · Post by **crystalmsc** » Fri Jun 23, 2017 7:47 am

It's been a while with several updates, I wonder how's the CPU usage on most system now. On my computer the System-8 is way more effective, around 35% top. But the System-1 easily goes up to 50 - 60% and start crackling. Is it just me, or the plugin is still heavy on CPU? Roland confirmed that it supposed to be much lower than the System-8. Also I can't freeze track on Live 9.7.1, Win 10, always crashed when finished.

moskaluha · Post by **moskaluha** » Thu Feb 07, 2019 12:37 pm

roland system 1 not save plugins downloading in synth plyse HELP

Roland - SYSTEM-1 Software Synthesizer