zipper noise and stairstepping?

[summers]

Will this be fixed in v2?
As this puts me off getting the v synth.

[nsheldon]

No, as this "zipper noise" effect is caused by insufficient controller resolution. This is not correctable by software (it's a hardware limitation).

Regards,

Nathan Sheldon
http://www.nathansheldon.con/

[ArghBlarg_]

I know Access (makers of the Virus line) were always touting that they had a controller smoothing algorithm on their boards -- it's possible in theory to do some kind of velocity curve to make things seem a bit smoother... but without infinite-spin knobs one could get out of sync with their real physical location pretty easily if the algorithm were buggy. I wish more boards used those infinite-spin knobs, really eliminates the resolution problem nicely.

Error 144 - Reboot? Y/N

[SoftRoom_]

Zipper noise is not a factor of controller resolution in this case - the knobs are transmitting sysex, not MIDI CCs. Nor are continuous encoders necessary for smoothness. Yes, Access devised a form of "adaptive smoothing" which, I'm guessing, is a series of lag processors that are applied to smooth incoming 7 bit MIDI controllers to the internal resolution (is it 24 bit, I forget?).
Anyway, I think you'll find the V-Synth's existing TB-303 filter is way, way smoother than the original filters so Roland *could* introduce some form of smoothing for the others too. I can't say whether they have in version 2; we'll have to wait and see.

Paul

[Watermelon]

I also noticed that when you use a CPU-hungry patch, zipper noise tends to be higher (if zipper noise refers to the steps one sometimes hear when rotating a front panel knob). Am I right? If so, it looks as if priority was given to the sound quality, not the real-time tweakability. This means that, if CPU has a hard time generating a sound under a given polyphony, it may start scanning the knobs position in slower way, hence some more abrupt sonic changes as the knob moves further between two scans. Makes sense to me.

[esden9]

In software, Roland could use a lag function to smooth it out without the need for endless encoder type knobs or encountering out of sync issues. Would be cool to be able to tweak the lag time too...incorporating a sort of velocity sensitivity would even allow for more dynamic changes while still smoothing out the small tweaks.

In this way, even MIDI CCs can be affected, while with a hardware solution this would not necessarily be true.

[anti_pulse]

Screw it- make that stair-stepping your trade mark sound.

Well I'll be your huckleberry....