VGM Driver Update
Moderator: Mask of Destiny
Sigflup, I committed an update to the VGM driver code that should solve the empty PCM data block problem. You should be able to pull this and sync it with your changes.
Yeah, the problem is normally when specifying the length of a PCM data block, the you add the length to vgm address *before* the data length portion. Essentially this requires reading in this length and then subtracting 4 bytes.
In the case of a PCM data block of length 0, subtracting 4 bytes is incorrect, since that is essentially trying to get the system to process opcode 0, which doesn't exist.
Yeah, the problem is normally when specifying the length of a PCM data block, the you add the length to vgm address *before* the data length portion. Essentially this requires reading in this length and then subtracting 4 bytes.
In the case of a PCM data block of length 0, subtracting 4 bytes is incorrect, since that is essentially trying to get the system to process opcode 0, which doesn't exist.
Updated vgm_drv: https://dl.dropboxusercontent.com/u/101302896/vgm_drv.h
This should fix certain VGM files that have empty PCM blocks and probably some other problems as well.
This should fix certain VGM files that have empty PCM blocks and probably some other problems as well.
The average 68k memory access for my musicdisk is about 8.8Kb/s. That equates to at most 1.6K cycles lost on the 68k's clock every NTSC frame. Which isn't that bad, I guess. About how many instructions would you as assume 1.6k cycles is?
personal stuff....
I'm noticing skipping on my program and I think it's because my main loop gets extended beyond a frame by the z80. I could be wrong on this.
personal stuff....
I'm noticing skipping on my program and I think it's because my main loop gets extended beyond a frame by the z80. I could be wrong on this.