Originally Posted by shykedmi
Wow, you both were right, the Unison was on and I was not working in Manual Mode. Would you mind explaining briefly on what these two options are?
Manual mode - if it's the same on all other synths - does not use the preset memory of the synth but the knobs in the physical position they are in at the moment.
A preset may have the filter cutoff set at 10%. In reality, the knob is set to say, 90%. When you load up the preset, you won't notice this; until you turn the filter cutoff, then the value suddenly "jumps" to 90%. There are several solutions for that, but I don't know if the NL offers them. One option is to let the knob do nothing until it's near to the value in the memory; this means that on some synths you can simply turn the knob all the way down to 10% and hear no change, but once you cross the 10% and go up again, the value starts to follow the knob.
Manual mode is useful for creating sounds from scratch, but an initialized preset (usually a single oscillator playing, filter opened up/bypassed, envelopes set to zero attack/zero release) is more useful. If you don't have one, create one and save it in a remote location.
One other thing, when I'm trying to work in Manual Mode and then put the Unison option it makes the sound fatter. Why is that? It sounds like it's doubling it an octave down.
Unison triggers the same note several times.
Consider the Jupiter 6. It has 6 voices - each "voice" is a 2-oscillator monophonic (one note at a time) synthesizer with its own multimode filter and its own envelopes. When you play a C-major chord, the first voice's pitch is set to C, the second to E, the third to G. However; the separate synthesizers do not care what notes they're set to; it's just that on a conventional keyboard, playing C a second time means that you have to release the first C note you played. Otherwise, voices would pile up. With unison, one note tells all those 6 little monophonic synthesizers to play the same note. There are several unison modes too - with the Jupiter, playing two notes means each note gets 3 voices each. With three notes, each note gets 2 voices each.
The reason I use a JP6 is because if you open it up you can actually see the separate pieces of circuitry; on other synths, it's even more clear because they're completely separate circuit boards. On a Nord Lead, you have one big DSP chugging along so you don't see 16 (or 20) little chips, each responsible for one voice. The reason it can still create separate sounds is because they're separate in the memory of the DSP itself - and the polyphony limit is because the DSP in there has only so much power to generate so many sounds in a short enough time simultaneously.
The "fatness" you describe is a given with most analog synthesizers because those oscillators will always vary a little bit in pitch. On any digital synthesizer, pitch variations have to be explicitly added afterwards. This can be done with an extra parameter called "spread" (naming for this varies). This splits the voices in two groups - voices 1, 2 and 3 get tuned down say, -5, -10 and -15 cents, voices 4, 5, and 6 get tuned up +5, +10 and +15 cents. The amount of spread acts as a multiplier - so you can go from -1, -2 and -3 to -10, -20 and -30 in one smooth turn - if the synth actually offers that as an option.
Unison is not equal to guaranteed fatness. By adding several detuned copies of the notes, you're also blurring the initial attack. Also, the NL2X has a polyphony of 20 notes - 2-voice unison means you reduce that to 10 notes, 4 voice unison reduces it to 5, etc.