Had to solve a little problem of HV leaking into the circuitry via the ferrite core (or more likely the metal clip that holds the two halves together) and my DIY primary winding. I was getting zapped touching the mosfet, the audio input plug and my power switch.
Apparently silcone sealant is not a good enough insulator against HV and I was getting internal arcing through the silcone and air pockets to the metal clip. I guess from there the insulation on the magent wire has no hope of stopping it.
In the end I slipped an irrigation riser tube over the solder joint on the tv flyback transformer and filled it with araldite and pooled araldite around the base of the riser tube. That seemed to stop the HV dead.
But anyway I have discovered that the duty cycle of the oscillator has an effect on the music volume coming through the HV arc. Suppose it stands to reason - with little 'room to move' the audio modulation has trouble transfering signficant power to the HV arc.
Any way here it is. The audio is now clearly audible above the crackling now: **broken link removed**
What about stereo plasma balls?