Polysix - MG Delay of Zero is not Zero

Earlier this week, when playing my Korg Polysix, I noticed that the LFO ("MG") was not having the effect that I expected.  I often like to set the MG to sweep the VCF very slowly, especially when playing the arpeggiator.  Based on my experience with the Korg Mono/Poly, I expect that the slow MG sweep would cause the arp notes to smoothly change their brightness from note to note.  Unfortunately, that's not what my Polysix is doing.  As you can hear in the soundcloud sample below, there is a clear transition at the start of each note where the VCF goes from its default value (as set by the VCF Cutoff knob) to the current value of the MG.  If you have a Polysix, does yours do this?


The settings for this sound are shown in the pictures below (MG Freq = 2, Delay = 0, Amount = 6, VCF Cutoff = 5, Resonance = 0, EG Intensity = 0, KBD Track = 0).

To show visually illustrate this unexpected audio behavior, check out the screen shots below.  These are screen shots of the audio in the sound cloud sample above.  Each segment shows four notes from the arpeggiation.  The top figure is when the MG is at the high end of its cycle, which means that the MG is opening the VCF beyond the setting from the VCF Cutoff knob.  As you can see in the screen shot, the note clearly starts at a lower VCF setting and then, after ~100 milliseconds, the filter opens up to the value defined by the MG.  This is with the MG Delay at zero!  It should not be like this.  The bottom figure shows the same thing, except where the MG is at the low end of its cycle where the MG is closing the VCF to a value lower than that set by the VCF cutoff knob.  Again, there is ~100 ms delay before it transitions to the MG's value.  In my mind, it should not be this way...the beginning of each note should be no different than the middle of the note.

Even with MG Delay set to Zero, There is Still ~100ms Before the MG Affects the Sound

So, I don't think that it is supposed to work this way.  Unfortunately, I don't know whether this is a new behavior of my Polysix (ie, it has become broken) or if it has always been this way (ie, it is a "feature" of the Polysix design).  If you have a Polysix, does it respond like this?

Update: I explored the MG circuit to find the cause of this behavior