[openal] Attenuation Roll-Off problems (too silent)

Roland Plüss roland at rptd.ch
Thu May 19 21:54:57 EDT 2022


Figured something similar. But the problem exists also for more sane 
situation. For example if I put a source where an actor is and he is 
speaking with volume 1 (normal conversation loudness) then you can go 6 
meters away and here nothing at all. I can though hear very well 
somebody speaking across a room. So is the roll-off of 1 not correct?

Am 20.05.22 um 00:34 schrieb Chris Robinson:
> On Thursday, May 19, 2022 6:22:57 AM PDT Roland Plüss wrote:
>> I'm still trying to figure out why the attenuation in OpenAL is not
>> working properly.
>>
>> According to the documentation a roll-off of 1 is the default or
>> "natural". If I use this though and I have let's say a sound source a
>> couple of meters above ground (think of an airplane or similar) with
>> gain set to 1 then it is pretty much inaudible a couple of meters below
>> on the ground. I tried setting the gain like to 10 but this does not
>> help much. I ended up using a roll-off of 0.05 to make the sound audible
>> on the ground.
> Airplanes are very VERY loud when their engines are running. A gain of 10 is
> only +20dB, whereas I'd start thinking in terms of +60dB to +100dB (gains of
> 1,000 to 100,000) or higher[1]. OpenAL will clamp a source's gain to
> AL_MAX_GAIN (default 1) after applying distance attenuation to the source
> gain, so you don't need to worry about blowing out your users' ears if they
> manage to get close to such a sound source, but these are the kinds of levels
> needed for very distant sounds to remain audible.
>
> [1] https://aviation.stackexchange.com/a/50507
>
> It's also important to set the source's AL_REFERENCE_DISTANCE property
> properly, along with the source's AL_GAIN. The reference distance specifies
> the distance where the source is as loud as specified by the gain (which is
> relative to the actual sound in the buffer). How to set the properties in that
> case will depend on the actual sound recording and the distance it's meant to
> simulate at full volume.
>
>
> _______________________________________________
> openal mailing list
> openal at openal.org
> http://openal.org/mailman/listinfo/openal


More information about the openal mailing list