UDG in VDC and VIC-II for the C128 target

[Fabrizio]

@Stefano, I think we have a problem with testing the SID as there is not one but many and two main types: old and new.
I am not sure what is found on the C128.
The old one should produce a click sound by setting volume to max and 0.
Maybe the new one does not produce a click sound by setting the volume to max and then to 0. I am not sure about this but I know for sure that some effects are only possible on the old one.

[stefano]

It is just my guess, but I think I heard the click while testing the last fix, it happens when opening the sound device.
I knew already there are differences in the SID versions, affecting also historical programs.
We could be lucky, but only the reports from the c128 owners can confirm it.

[Fabrizio]

@Stefano, C128 has correct sounds but now we have a problem with the Spectrum and maybe other targets:
On the Spectrum the bit bang sounds produce a change in the border color.

[stefano]

ok, I'll try to reproduce the problem.
Is it happening on any bit_fx() effect?

[Fabrizio]

I can see the problem on the fixed sounds but not on the ones that were already ok.
I observe the problem on the spectrum target.

[Fabrizio]

@Stefano, you can reproduce the problem by using the same test code I already posted:

Code: Select all

#include <sound.h>
#include <stdio.h>
int main(void) {
    while(1) {
        bit_fx(7);
        getchar();
        
        bit_fx(1);
        getchar();
        
        bit_fx3(0);
        getchar();
    }
    return 0;
}

The border color is set by the previously broken sound effects:
bit_fx(7) and bit_fx(1)
while bit_fx3(0) seems to reset it to its normal color (which may not be necessarily the correct behavior).

[Fabrizio]

We also have an older problem with startup. It takes about 20 seconds to start the games and it does not depend on VDC vs VIC-IIe.
You can see the problem with:
https://github.com/Fabrizio-Caruso/CROS ... _80col.d64
What is the loader doing? Is it relocating the code? If this is done by the loader, it may be done in MOS6502 Assembly in 2mhz mode, which is roughly 2.5 faster than the Z80@2mhz (effective).
If this has to be done in Z80 it could still benefit marginally by disabling the Vic-IIe.

[stefano]

sound fixed, while I was on it I fixed also the EACA EG2000 conflict between sound routines and the VDC. I also tuned the bit_play routine to the EG2000 speed, to make it coherent with the PSG frequencies and have an average speed between the earlier and later TRS80 models.

[stefano]

before I forget it, probably the text6 mode is still inheriting the 80x50 range limitations from the existing libraries, the Make process requires yet another small correction.

[stefano]

> @Timmy, I just finished a new "UDG" mid resolution mode (80x75), I'm adding it soon (possibly moving some more code in a general text6 section).

It turned out as a buggy/messy pull, which I fixed only recently !
gfx128udg.lib should be in working order, now.

[Timmy]

> @Timmy, I just finished a new "UDG" mid resolution mode (80x75), I'm adding it soon (possibly moving some more code in a general text6 section).

Are you addressing this to me? I have no plans for doing C128 targets for the moment?

[stefano]

Oops, I don't remember why I mentioned you at the time Probably it was just a random answer to your "like".