Re: [Flightgear-devel] FGCOM gets a discussion page on the wiki

[Willie F. <wi...@tu...>]

On Monday 26 November 2007 09:53:16 Holger Wirtz wrote:

> Ok, but I see many things that should be solved at the same time. We
> should create a list og topics to be done and we should try to give them
> priorities and perhaps name who try to solve these topics. But where to
> place this list? FG-Wiki? My own Wiki?

I suggest FG-wiki with a link to your own wiki for now.


> I see the following realism problems:
>
> 1.) absence of random white noise
> 2.) the voice channels should be limited with a high cut (at 4 kHz) (or an
> band pass between 300 Hz and 4 KHz)
> 3.) as in real life there should be a mechanism for realising
> crosstalking (e.g. the sender with the maximum of output pushes other
> senders in the background)
> 4.) real com radios have an automatic noise limiter (squelch) which
> makes a little noise after releasing the PTT key.
> 5.) pilots have engine sounds in the background

I think we need to go for _some_ more realism but not at the expense of too 
much extra complication which may gain us very little in the end.

With headphones at both ends, Jester and I did a good test with his latest 
patched code. I need to get a better headset mic-- For now I only have 
headphones and an old omnidirectional mic -- not realistic at all. I think we 
will see very good results with Jesters latest patch and properly positioned 
mics and headsets
>
> Here are my ideas:
> 1.) A daemon in the background ca radomly place short samples of white
> noise and/or athmospherical noise an used channels. The problem is that
> crosstalking cannot be recognized from such a daemon (see 4).
> 2.) Perhaps an EQ in the sound chain. The best place would be iaxclient.
> But also ALSA would be working - but this is not real protable.
> 3.) That's a real problem. If this should work something like a complete
> new conference module for asterisk must be developed __AND___ a
> mechanism (schedular) for the voice clients (perhaps a simple FIFO). Not
> as easy as it sounds...
Lets leave that to one side for the moment - sounds overly complex . I should 
be careful what I wish for.......

> 4.) Why not sending a simple short noise sample after muting the mic?
> This could be placed inside iaxclient.
> 5.) Again: mixing engine sound inside the mic stream... perhaps also at
> iaxclient?
>
> What I see: Everything is more a sound/VoIP problem rather than a FG
> problem...
Indeed

Im with Jester on this  - lets get good clean sound and then add a (possibly 
optional) realism layer on top which would include a bandpass filter, white 
noise and PTT clicks etc.
 Engine noise is more complex. To be right it would have to be different for 
each type. The background cockpit noise in a Pitts is not what we'd expect in 
the 777. My vote is to give this a low priority for now.
> > Please note -- Im used to hearing ATC chatter in the UK --probably a LOT
> > different in the rest of Europe and the US. I don't know how the ATC
> > chatter we have sounds to the rest of you guys both in terms of content
> > and quality.
>
> I tried to follow some web ATC streams and I udnerstand nearly nothing -
> especially the pilots are very difficult to understand.
Its gibberish until you understand a little about what information is being 
tranferred and the idioms involved.

"Downwind for full stop, Golf Oscar Mike"  means little in isolation..
in context, Ive just told EGPK tower that Im about 1.5km to the SW of the end 
of runway 31 on an approx heading of 130 deg at 1000feet and I want to land 
and taxi back to the clubhouse, if thats OK with him...

Its actually fairly easy once you know some background  :-)

 How did you feel the first time you saw a >10 line shell script?

Scared me silly..