Pan controls for individual oscillators?

Please? I'm having trouble finding a use for a global pan control. The ability, however, to pan individual oscillators would be a great help in creating some huge stereo patches. I know you can spread unison voices across the stereo field under the Voicing tab, but extending the synth's felixibility with true stereo assignment of each Osc would be great.

 

Another vote from me!
or pan controls for individual filters.

regards
Thomas

 

Indeed it would be great to have pan knobs on all oscillators, also with control slots.

 

as above. and i don't know why there are volume sliders for the filters when there is a crossfade to mix them up. in my opinion they can be replaced by pan controls.

 

It would be nice to keep what is there for the filters (sliders and so on).
Maybe just add a row of 3 or so horizontal sliders to the bottom of the window, with drop down menus for assigning them to either the osc/filter/insert/fx.

 

+1
Pan for individual oscillators (and noise generator) and for both filters. Like Fabfilter Twin.

 

ditto all the above

 

i'd really like to be able to pan each oscillator individually too. but depending on what NI's archetecture is like, offering this might increase the cpu load.

i suspect that right now the filters and everything else up to the main effects get to run in mono, since a stereo signal path isnt currently required. to offer panning of the oscillators, everything after them would have to be run in stereo...including the filters and the insert effects. this would likely increase the cpu load. on the flip side, it would also allow the insert effects to have all the same options as the main effects.

NI is already receiving some criticizm of the current cpu load. if my guesses about the signal path and the possible downside of adding oscillator panning are correct, then they may have to find ways of reducing the current load first before going stereo. otherwise the cpu load could get mighty high.

 

yes, definitely desireable.

I do not see the CPU problem - I often only need one voice but calculate 4 all the time. So it would be great to have more flexibility - as duo/quad-mode, stereo with two voices, etc.

if you can switch between uni/duo/quad you can freely choose how to manage your CPU consumption. A 4 output version would also be fine (one output per calculated voice)

best

 

My guess is the the filters and insert FX would have to be calculated seperately for left and right, what could almost double the CPU load for some of the sounds.
I would prefere an improved unisono mode with a modulateable spread parameter on the most important parameters, like Oscillator volume, pitch, table position, all filter parameter, what should do exactly what you need and a bit more...

 

That'd be my guess as well.

ew

 

but there are always four voices calculated - so it would not necessarily be just one calculation more....

 

Well, imagine the signal part, everything after the oscillators would have to be stereo and thus calculated twice once for left and once for the right channel...

 

Yes, but what he is saying is that if you normally play one or two voices only then there is wasted capacity since Massive always plays voices in groups of 4. Presumably the reason is that where available it is using SSE instructions and calculating 4 voices in parallel with each instruction instead of 4 samples worth of one voice.

So in a case where you are only playing one or two voices, you do not lose anything by having to calculate twice as much, because 50-75% of the CPU usage is going towards nothing anyway.

It would be nice as an option but who knows, it might not be technically feasible or impractical for some other reason.

 

well actually i would prefer a 3rd filter instead. so that you could assign the 3rd Osc to that filter and have your panning for each Osc that way, this would open more sound shaping possibility's
and would only require 1 additional filter to be added. witch can be turned off if your not using it. as for Gui implementation that could be don bye replacing filter 2 with filter 3 bye a small switch in the filter 2 window.
As mentioned above adding Panning for the Osc's would require the whole chain from the Osc's (the lowest part actually) and up to be calculated twice (one for each channel *left and right*) this means 2 additional filters with no separate controls would have to be added, and having panning on the Osc's would not open op allot of new sound shaping possibility's, so you will be utilizing allot of CPU usage for a very little-useful feature.

 

well, surely not for 1 dimensional sound shaping ..... but there is more for sound - more than 1 dimension. And it depends if it costs something - for me it would often just mean that I can actually use what gets processed anyways - as always 4 voices get processed.

nothing against another filter, but as you cannot even pan the filters this seems not at all relied to this topic .....

best

 

It could be implemented as separate Amp Sections. One each for Mix 1 and Mix 2, and possibly one more for bypass.

Shouldn't mess up the architecture too much, but it might be trouble fitting the controls though.

 

Ok, fair enough that oscs are probably best left non-pannable for performance and simplicity reasons. But the two filters need to be pannable - this is a basic feature for a semi-modular of Massive's complexity.