[SDL] endianness in SDL_audio.c
frodet at nvg.ntnu.no
Thu Jan 27 15:29:17 PST 2005
> On Thu, 2005-01-27 at 16:37 +0100, Patrice Mandin wrote:
> The root of the problem is that developers use AUDIO_S16
> when they should be using AUDIO_S16SYS. I recently fixed
> Tux Paint, then patched up a wiki somewhere to warn about
> the issue. App developers are pretty much led into using
> AUDIO_S16 instead of the correct AUDIO_S16SYS.
I am aware of the problem with AUDIO_S16 vs. AUDIO_S16SYS. In my case I
have sound data produced from a separate library which delivers this in
LSB. SPARC expects MSB and the way I read the SDL code it has the
ability to convert between representations.
> As far as I can tell, one should never use AUDIO_S16 in
> app code. It should be removed from the headers I think.
I'm (pretty - I'll check again tomorrow) sure I tried both before and
after applying my fix and the result was the same. audio->spec.format =
AUDIO_S16MSB on SPARC.
I'm a novice with SDL audio so it could be that I'm missing something
tho'. If so, what?
^ Frode Tennebø | email: frode at tennebo.com | Frode at IRC ^
| with Standard.Disclaimer; use Standard.Disclaimer; |
More information about the SDL