On 01/22/2009 04:20 PM, James Sleeman wrote: > Hi John, great answer, thanks..
:-) >> We see that at the reference distance (r0), the signal is not >> attenuated at all. That's the defining property of the reference >> > So the reference distance is actually the distance from the microphone > to the sound emitting device when the sample was captured? Well, that would be true if your record and playback system and unity gain from end to end ... which is almost certainly not the case. There are too many places where it is possible (or even necessary) to mess with the volume knobs. There is a huge element of arbitrariness and artificiality in the whole exercise, because few gamers are going to turn up there computer audio systems loud enough to be faithful to the sound levels in a real airplane. All we can hope for in the cockpit is _balance_ i.e. a reasonable balance between the volume of engine noise, wind noise, gear motor noise, radio reception, et cetera. There is additional arbitrariness in the flyby view. Is the observer wearing ear projection? I hope so. What kind? I have no idea. > Here's what the docs (docs-mini/README.xmlsound) say, they don't quite > seem to match that. Or has all this just wooshed over my head and I > have to read your message again more carefully? I stand by what I wrote. Don't believe everything you read in the docs. > But given that only the person who captured the sound sample knows the > reference for real, isn't it at least ideal to specify that reference? Again, fiddling with the gain is tantamount to fiddling with the reference distance. If you obtain a calibrated recording, with calibrated distances and calibrated sound pressure levels, then you can try to do something _ab initio_. But FG is a long long way from that level of precision. My recommendation: Fiddle the volume so that it sounds about right at the point of closest approach, and scale everything else by 1/r^2. That would be a huge improvement over the existing code ... and going beyond that would yield diminishing returns. > Perhaps for our intents-and-purposes a reasonable guess could be made > based on the sound type, that is, it might be reasonable to assume that > if not specified the reference distance for an engine sound is probably > around 10 meters in most cases. See previous recommendation. > I can't imagine there would be a really big difference in saying the > reference was 10 meters or 20 meters. Actually that is a 6dB difference. Maybe not "really big" but certainly quite noticeable. > For things like flap, gear and > switches, the max-dist is probably more important than distance as these > things are really "internal" sounds, max-dist of a meter or two would > likely suffice. None of this "reference distance" stuff has any applicability inside the cockpit. Engine noise _fills_ the cockpit. Moving your head around inside the cockpit does not make much difference, and to the extent that it makes any difference at all you would need a treeeemendously complicated model. A simple "reference point" model is dead on arrival. Similarly I assume you (the pilot) are wearing headphones, so moving your head around in the cockpit won't have any effect on ATIS or other radio traffic. ------------------------------------------------------------------------------ This SF.net email is sponsored by: SourcForge Community SourceForge wants to tell your story. http://p.sf.net/sfu/sf-spreadtheword _______________________________________________ Flightgear-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/flightgear-devel
