carlsson wrote:Ultraplayer uses tight control of the 4 oscillators to provide 9 bits of frequency resolution and up to 6 bits of PWM. All you hear is the original VIC oscillators. No volume register modulations.
Hm, is this approach the same as TBC just implemented?

I know Stockholm is a small city but it seems quite remarkable if two people within a half year made pretty much the same thing.
I don't know him or what approach he uses and I don't think he knows my approach.
I did hear the mp3 he posted yesterday in the games forum so I decided it was time to take my player of the shelf and do a first release.
PWM is something I've been trying to do for quite a while using the same principle but didn't fully realize exactly how tight the control has to be.
I started implementing it mid December 2008 using new insights into the internals of the oscillators.
The new player and this music was completed by the end of January 2009/beginning of February 2009.
carlsson wrote:However I wonder how IRQ/CPU intensive the tight control is. For applications where the music is prioritized I am sure these new routines are superb but once you start to add advanced graphical effects requiring both CPU time and exact timing, the music player is preferred to take as few raster lines as possible.
In Ultraplayer the control is quite CPU intensitive, both NMI and code outside interrupt.
(enabling PWM for less voices makes it a bit less CPU intensitive)
The aim was maximum control and to make it fit on unexpanded VICs.