Re: [Stella-main] Sound problems with stella.x11 & .sdl, stella 1.3

[Stephen A. <ste...@ro...>]

On March 21, 2004 11:28 am, Thomas Boutell wrote:
> > Yes, the problem has always been with the sound synchronization,
> > not the actual generation.  In Stella 1.2, the sound code was
> > external (stella-sound).  This made sound sync impossible...
>
> I can see that, but it's not clear to me why I get only intermittent
> sound -- I'd expect just to have some noticeable jitter between the
> timing of the game and the timing of the sounds. Hmm.

Well, the timing would be causing that as well.  If the emulated sound 
chip is updated more often than the operating system can play the sound 
(ie, more sound bytes are created than can be played in a certain time 
frame), the result may be intermittent sound.  It would be that way 
because a sound that was meant to be played at a specfic time could be 
overwritten by no sound, and it would result in sound dying out at 
times.  It still comes back to the fact that sounds aren't being played 
when they should be; hence a timing issue.

> I'm reading straight from the mouse devices (which isn't scary at
> all, as it turns out). I've added it successfully to both the X11 and
> the SDL main.cxx files. SDL and X are not configured to see the
> trackballs.

I recently sent a message to the SDL mailing list about multiple USB 
mice (which is what a trackball essesntially is), and SDL has no direct 
support for it.  So that bit of code will have to be platform-specific 
(in another class I guess).

> Under win32 you'd have to use DirectInput (for older versions of
> Windows) and, since multiple-mouse support has been removed (!) from
> DirectInput, the new "raw input" API (for windows xp and later). The
> latter has already been done for you:
>
> http://link.mywwwserver.com/~jstookey/arcade/cpnmouse/
>
> Also note the notes about the existing support for this stuff
> in Mame Analog+.

I'll take a look at this, but for now, my main concentration is Linux, 
and more importantly, sound.

> USB trackballs do work *very* well as paddles; all I had to do was
> borrow your existing mouse-as-a-paddle code. Players need about two
> seconds to figure out that only the X axis matters, and then they're
> off to the races.

I really have to get some hardware like this to test it on.

> Consolidating on SDL makes sense, I would encourage you to retire
> the unnecessary extra workload of maintaining other GUI back ends
> if it makes it more pleasant for you to improve Stella. (:

Yes, since there is only one (semi) active maintainer, that's what I've 
had to do.  And it doesn't really make sense to have separate codebases 
anymore.  It did when Stella was first started, since SDL didn't exist 
at that point.

> Personally, though, I get the wrong colors with the SDL Stella
> front end. I noticed something similar when I tried one of my own
> SDL games without X, but with Stella I have the problem even when
> Stella.SDL is running *under* X. That's pretty bizarre. But since
> SDL is so portable I should probably blame the SDL implementation
> I'm linking against, not Stella, although I can't figure out what
> it might be doing; I'll try another video mode or something.

Are you saying that the colours aren't what you think they should be 
(but are the same in both the X11 and SDL versions), or that the SDL 
version actually produces different colours than the X11 version?  If 
it's the former, then I plan to add (at some point) configurable 
palettes, since almost everyone disagrees with how a real Atari 2600 
looked.  If it's the latter, then that should be impossible, and I'd 
like to see a screenshot.

Steve