Gepostet am: 08.02.2016 08:29:24
Some good news!
Over the weekend I managed to achieve bit-match for the new filter implementation. First simulation results already showed an improvement of sound quality of the 6581 filters. Maybe I will upload some simulation results soon, but it might be faster to get recordings from real hardware earlier because the simulation of 20 seconds of audio took about 5 hours...
Synthesis for the FPGA ran through with some timing problems: The resource utilization in the FPGA is critical in some places. All multiplier resources are used up and some additional multipliers had to be synthesized by normal logic which might be an explanation for the timing issues. But that's no show stopper so far.
With this achievement the project has taken the hurdle "RESID-alike filter implementation".
I will have to fix one or two minor issues and then I can do some recordings from the hardware.
Stay tuned for the first audio samples to appear in the next days...