[LAU] 5.1 to ambisonices (was: more than 4 channels for listening? Really?)
arnold at arnoldarts.de
Mon Mar 1 03:45:50 EST 2010
On Sunday 28 February 2010 21:03:48 Jörn Nettingsmeier wrote:
> On 02/28/2010 08:54 PM, Martin Leese wrote:
> > Arnold Krille<arnold at arnoldarts.de> wrote:
> > ...
> >> My rational is that most (all?) 5.1 mixes are: music and action stuff on
> >> L+R (front), dialogues on C, ambient and action effects on L+R(rear). So
> >> my idea is to split the rear channels into six or more channels
> >> distributed around the rear side of the circle (or even sphere) to have
> >> a more surrounding feeling and overcome the hole in the panning.
> > Multiple surround speakers fed with the same
> > signal is what you hear in a cinema.
> > Recreating that would therefore seem to be a
> > useful path to go down.
> why not just pan the 5 signals on an ambi rig?
That is the simpliest solution. But simply using the channels (after ambdec)
assumes your ambisonics speakers are at the positions 5.1 expects.
Doing so with panning-to-amb sets you free from that and even allows you to
rotate the set-up.
In that faust dsp file I also added a separate volume control for the center
channel, I found a 1:1 mix of that is to heavy.
The interesting part starts when you try to improve the sound within reason.
And thats where you use the rear-channels (which are normally used for ambient
in the movie) to create an ambient-half-sphere around you...
BTW: What to do with the LFE-channel? Add a lowpass and mix it directly on
'W'? Or ignore it altogether as all the other channels are full-range already?
> if you want to get rid of the hole in the panning (to the rear), you are
> prepared to cheat anyways. in that case, it might be interesting to try
> a sixth source in the rear, derived from summing SL and SR, maybe
> including a few allpass filters for decorrelation...
Hm, allpass-filters are something I haven't yet looked at at all. Will do so.
Currently I created six sources in the rear derived from the two rear-channels
of 5.1. Maybe I will also play with a bit of delay there to widen the space...
Next time I will post a compilable project instead of the single faust file.
Should make it easier for most people:-)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Size: 197 bytes
Desc: This is a digitally signed message part.
Url : http://lists.linuxaudio.org/pipermail/linux-audio-user/attachments/20100301/bda0f469/attachment.pgp
More information about the Linux-audio-user