10 Feb
2009
10 Feb
'09
6:17 p.m.
Hello,
I am helping a friend to mix an orchestral piece using Ardour. To add depth to
the different instruments, we have created three buses, each with a TAP
Reverberator with the same settings (Hall HD type) excepted the dry/wet ratio.
The different instruments are panned according to a common orchestral layout.
This works, but I feel that both the depth and the stereo positioning could be
improved. Indeed, the Hall HD reverb settings has fixed early reflections
patterns while the different instruments should have different ones according to
their physical positions.
I was wondering if any of you knows about some ways to improve this. As a
solution, I am thinking about writing a software to compute the early
reflections patterns of an instrument given its 2D position in the orchestra.
In this case, do you have suggestions of how many early reflections I should
use (left/right walls + back for left/right walls?). Does the allpass filter
settings depend on the position of the instrument/walls?
My feeling is that the reverb GUI should include stereo panning as well in
order to launch rays from the correct 2D positions.
What do you think about such idea? Would someone be interested to contribute
to such project?
Thank you, kind regards,
Stéphane
--
http://stephane.magnenat.net
10 Feb
10 Feb
6:37 p.m.
On Tue, Feb 10, 2009 at 11:17:49PM +0100, Stéphane Magnenat wrote:
I am helping a friend to mix an orchestral piece using
Ardour. To add depth to
the different instruments, we have created three buses, each with a TAP
Reverberator with the same settings (Hall HD type) excepted the dry/wet ratio.
The different instruments are panned according to a common orchestral layout.
If all three have the same settings you can have *exactly*
the same effect using just one Reverberator. Set it to 100%
wet. On each track use a post-fader aux send to the
reverberator, and mix its output into the master bus.
This works, but I feel that both the depth and the
stereo positioning could be
improved. Indeed, the Hall HD reverb settings has fixed early reflections
patterns while the different instruments should have different ones according to
their physical positions.
I was wondering if any of you knows about some ways to improve this. As a
solution, I am thinking about writing a software to compute the early
reflections patterns of an instrument given its 2D position in the orchestra.
In this case, do you have suggestions of how many early reflections I should
use (left/right walls + back for left/right walls?). Does the allpass filter
settings depend on the position of the instrument/walls?
This is absolutely right, but not easy to achieve unless you
have dedicated software to do it. Even in that case, using
just three different patterns of early reflection (left,center,
right) produces very good results.
If you don't have dedicated early reflections for each
distance, the parameter to play with (assuming the routing
as desribed above, with the reverb in 100% wet mode), is
a delay in the aux send to the reverb. For close sources
you use a larger delay and lower send gain, for far sources
the delay is shorter and reverb level is higher. Ardour
doesn't allow to do this in a simple way (there is delay
control on the aux sends, nor any way to insert one).
Ciao,
--
FA
Laboratorio di Acustica ed Elettroacustica
Parma, Italia
O tu, che porte, correndo si ?
E guerra e morte !
6:39 p.m.
Hello!
I'm no real expert in the field, but I did a few experiments in 3D sound,
using headphones and reverbaration. If you're not afriad of text-applications,
you could use csound. Or possible clm(I don't know if they have the same
capabilities, but I'd strongly guess so).
Csound has an opcode to design a simple four wall, floor, ceiling room, in
which you can position sounds (in real 3D). The realism of the sound of course
depends on your output setup (headphones, 8 speaker cube). But with my last
piece, I found, that the csound way gave a nice realastic ambience to my
piece.
If you're not so keen on 3D sound and 3D positioning, you could of course
look for a nice IR (impulse response) and use jconv or its graphical mate.
These natural reverbs sound very nice and clean. They don't have the grainy
quality.
If you're interested in the results of csound's spat3d opcode you might look
here:
http://juliencoder.de/songs/a_little_moaning.ogg
(first 3 minutes, big room, with 9 placed sources)
and
http://juliencoder.de/songs/a_life_in_keys.ogg
(small band in a smaller room, nice ambience
I can't verify the URLs right now, the server seems to have a small
downtime. Sorry for the inconvenience.
Kindest regards
Julien
--------
Music was my first love and it will be my last (John Miles)
======== FIND MY WEB-PROJECT AT: ========
http://ltsb.sourceforge.net
the Linux TextBased Studio guide
======= AND MY PERSONAL PAGES AT: =======
http://www.juliencoder.de
8:22 p.m.
Back on Tuesday 10 February 2009 04:17:49 pm Stéphane Magnenat was saying stuff
like:
Hello,
I am helping a friend to mix an orchestral piece using Ardour. To add depth
to the different instruments, we have created three buses, each with a TAP
Reverberator with the same settings (Hall HD type) excepted the dry/wet
ratio. The different instruments are panned according to a common
orchestral layout.
This works, but I feel that both the depth and the stereo positioning could
be improved. Indeed, the Hall HD reverb settings has fixed early
reflections patterns while the different instruments should have different
ones according to their physical positions.
There is another program by the same author of the tap plugins called reverbed
that lets you edit just about every aspect of the reflections.
It's been suffering from bit-rot and may be hard to compile, but it works
pretty well.
-Reuben
11 Feb
11 Feb
3:02 a.m.
Reuben wrote:
There is another program by the same author of the tap
plugins called reverbed
that lets you edit just about every aspect of the reflections.
It's been suffering from bit-rot and may be hard to compile, but it works
pretty well.
Reverbed compiled no worries and works fine on my Fedora 10 system.
3:47 a.m.
There is another program by the same author of the tap
plugins called
reverbed that lets you edit just about every aspect of the reflections.
I've already played with it. It is very nice indeed, but it requires
handcrafting of all parameters. For instance, one has to specify each early
reflection by hand. It would be much more convenient to be able to specify
the size of the room, the position and size of the instrument, the position
of the listener, and some parameters for the walls ; and let the program
compute the corresponding early reflections patterns. It is probably not that
difficult to do, as it only requires basic 2D geometry. The main issues that
I see are:
- How many early reflections to take into account?
- What is the relation between the room dimensions and the allpass impulses?
- How to make nice use of stereo input; for instance, should the left wall
early reflection use the left channel source positioned at the most-left
point of the instrument area, and the corresponding layout for the right one?
If these questions find answers easily, building an easy-to-use reverb editor
for ladspa is probably only a matter of dumb coding, by extracting the TAP
reverberator source code and fusing it to a GUI application for instrument
and listener positioning.
What do you think?
Have a nice day,
Stéphane
--
http://stephane.magnenat.net
8:51 a.m.
On Wednesday 11 February 2009 02:47:31 Stéphane Magnenat wrote:
This is my dream born of ignorance...
Take a multitrack recording in something like ardour.
NAME each track.
Feed all tracks to a plugin that has a graphical representation of a sound
stage.
The NAMES (with icons) from the adrour tracks show up from left to right on
the stage.
You can grab the icons and drag them about the stage in realtime and hear the
changes in the output of the plugin.
Thoughts?
drew
There is
another program by the same author of the tap plugins called
reverbed that lets you edit just about every aspect of the reflections.
I've already played with it. It is very nice indeed, but it requires
handcrafting of all parameters. For instance, one has to specify each early
reflection by hand. It would be much more convenient to be able to specify
the size of the room, the position and size of the instrument, the position
of the listener, and some parameters for the walls ; and let the program
compute the corresponding early reflections patterns. It is probably not
that difficult to do, as it only requires basic 2D geometry. 
9:12 a.m.
drew Roberts wrote:
This is my dream born of ignorance...
Take a multitrack recording in something like ardour.
NAME each track.
Feed all tracks to a plugin that has a graphical representation of a sound
stage.
The NAMES (with icons) from the adrour tracks show up from left to right on
the stage.
You can grab the icons and drag them about the stage in realtime and hear the
changes in the output of the plugin.
Thoughts?
I like your dream, Drew. :)
Perhaps an extension of this software might be a starting point :
http://sourceforge.net/projects/meloncillo/
Be sure to look at the screenshots.
There are numerous VST plugins that provide neat graphic controls for
multichannel sound placement. Alas, I nothing about their realtime
utility nor how they perform under WINE. Check out these plugins :
http://acousmodules.free.fr/
HTH,
dp
10:25 a.m.
On Wednesday 11 February 2009 08:12:04 Dave Phillips wrote:
drew Roberts wrote:
Thanks Dave, not sure it helps in the big picture, it may, but it reminded me
of one point I left out. (I am not much of a programmer these days and likely
never was although I still have some stuff in daily use that was written back
in the 80s.)
Put and object on the front and centre of the sound stage representing the
listener.
Allow moving the the listener object back and sideways into the "audience" or
anywhere on the soundstage as well.
Alternately, allow the "listener" to be any one of the players/tracks.
This is my dream born of ignorance...
Take a multitrack recording in something like ardour.
NAME each track.
Feed all tracks to a plugin that has a graphical representation of a
sound stage.
The NAMES (with icons) from the adrour tracks show up from left to right
on the stage.
You can grab the icons and drag them about the stage in realtime and hear
the changes in the output of the plugin.
Thoughts?
I like your dream, Drew. :)
Perhaps an extension of this software might be a starting point :
http://sourceforge.net/projects/meloncillo/
Be sure to look at the screenshots.
There are numerous VST plugins that provide neat graphic controls for
multichannel sound placement. Alas, I nothing about their realtime
utility nor how they perform under WINE. Check out these plugins :
http://acousmodules.free.fr/
HTH,
dp
drew
9:53 a.m.
Am Mittwoch, den 11.02.2009, 07:51 -0500 schrieb drew Roberts:
On Wednesday 11 February 2009 02:47:31 Stéphane
Magnenat wrote:
This is my dream born of ignorance...
Take a multitrack recording in something like ardour.
NAME each track.
Feed all tracks to a plugin that has a graphical representation of a sound
stage.
The NAMES (with icons) from the adrour tracks show up from left to right on
the stage.
You can grab the icons and drag them about the stage in realtime and hear the
changes in the output of the plugin.
Thoughts?
drew
That reminds me on glmix, there you have 8 Speakers (8 jack_inputs) in a
3D room and you can move them around in the room.
http://devel.tlrmx.org/audio/
regards hermann
There is
another program by the same author of the tap plugins called
reverbed that lets you edit just about every aspect of the reflections.
I've already played with it. It is very nice indeed, but it requires
handcrafting of all parameters. For instance, one has to specify each early
reflection by hand. It would be much more convenient to be able to specify
the size of the room, the position and size of the instrument, the position
of the listener, and some parameters for the walls ; and let the program
compute the corresponding early reflections patterns. It is probably not
that difficult to do, as it only requires basic 2D geometry.
10:17 a.m.
hermann meyer wrote:
Am Mittwoch, den 11.02.2009, 07:51 -0500 schrieb drew
Roberts:
That's a neat app. I built and opened it without troubles on my JAD 1.0
and 64 Studio 2.1 boxes, haven't tested it though. Looks cool, thanks
for the tip. :)
Best,
dp
This is my dream born of ignorance...
Take a multitrack recording in something like ardour.
NAME each track.
Feed all tracks to a plugin that has a graphical representation of a sound
stage.
The NAMES (with icons) from the adrour tracks show up from left to right on
the stage.
You can grab the icons and drag them about the stage in realtime and hear the
changes in the output of the plugin.
Thoughts?
That reminds me on glmix, there you have 8 Speakers (8 jack_inputs) in a
3D room and you can move them around in the room.
http://devel.tlrmx.org/audio/
5:30 p.m.
On Wed, Feb 11, 2009 at 08:47:31AM +0100, Stéphane Magnenat wrote:
... It would be much more convenient to be able to
specify
the size of the room, the position and size of the instrument, the position
of the listener, and some parameters for the walls ; and let the program
compute the corresponding early reflections patterns. It is probably not that
difficult to do, as it only requires basic 2D geometry.
It's not so simple as it looks. The best spaces (for
orchestral music) do indeed have a simple 'shoebox'
shape rather than an 'auditorium' one that widens
towards the back, has a sloped ceiling, etc. But
they also have quite complex walls, with lots of
decoration, balconies, curved surfaces, etc.
Reflections from such structures are not so simple,
they are almost never specular (as in a straight
mirror) but include a lot of (frequency dependent)
scattering as well. Also 2D geometry is a bit too
simple - ceiling reflections are important as well.
The main issues that I see are:
- How many early reflections to take into account?
Unless you have a very weird shape (e.g. very long
and narrow), the first 4 or 5, or up to about 80ms.
But see above - what is heard (and measured) as a
single reflection can already be quite a complex
affair.
- What is the relation between the room dimensions
and the allpass impulses?
None AFAIK. A reflection in a real room is almost
never allpass - some bandpass would be a better
approximation. The allpass filters are used to
add artificial 'complexity' to the all too simple
models used by most reverberators, they don't really
correspond to anything physical. They reduce obvious
comb filter effects that could result from too simple
reflection models (e.g. just delay), and also improve
the reverb tail by 'randomising' it.
Allpass filters could be used to synthesise the
complexity of any single reflection, but not in
the way it's done in most reverberators.
- How to make nice use of stereo input; for instance,
should the left wall
early reflection use the left channel source positioned at the most-left
point of the instrument area, and the corresponding layout for the right one?
You normally wouldn't use stereo input to such a
spatialiser, but position each source independently.
Ciao,
--
FA
Laboratorio di Acustica ed Elettroacustica
Parma, Italia
O tu, che porte, correndo si ?
E guerra e morte !
12 Feb
12 Feb
6:06 a.m.
It's not so simple as it looks. The best spaces
(for
orchestral music) do indeed have a simple 'shoebox'
shape rather than an 'auditorium' one that widens
towards the back, has a sloped ceiling, etc. But
they also have quite complex walls, with lots of
decoration, balconies, curved surfaces, etc.
Reflections from such structures are not so simple,
they are almost never specular (as in a straight
mirror) but include a lot of (frequency dependent)
scattering as well. Also 2D geometry is a bit too
simple - ceiling reflections are important as well.
Of course, I'm fully agree that the model we speak about is an
over-simplification of the reality. However, should we do simplification,
shouldn't we try to have the most correct over-simplified model? This is a
real question, I don't know whether the simplification is just too important
or if by computing the "correct" early reflections on the sides, the ceiling,
the back of the stage, and the back of the room leads to a better space
feeling.
I was considering ceiling as well when speaking about 2D geometry, but if the
ceiling is not flat you are right, it becomes 3D ; however if we do a single
reflection this is still easy. I am not suggesting to build a general purpose
ray-tracer ;-)
Unless you have a very weird shape (e.g. very long
and narrow), the first 4 or 5, or up to about 80ms.
But see above - what is heard (and measured) as a
single reflection can already be quite a complex
affair.
Do you have any experience about whether a simplified but realistic model is
worth its development time?
None AFAIK. A reflection in a real room is almost
never allpass - some bandpass would be a better
approximation. The allpass filters are used to
add artificial 'complexity' to the all too simple
models used by most reverberators, they don't really
correspond to anything physical. They reduce obvious
comb filter effects that could result from too simple
reflection models (e.g. just delay), and also improve
the reverb tail by 'randomising' it.
Allpass filters could be used to synthesise the
complexity of any single reflection, but not in
the way it's done in most reverberators.
Ok, so should we use an allpass filter, we could as well tune it by ear?
You normally wouldn't use stereo input to such a
spatialiser, but position each source independently.
So if one has a stereo input of a violin ensemble that spans 3 meters in
width, one should use two independent sources to compute the reflections?
What about depth, should one use two sources, one in front and one in back,
and compute early reflections for both?
Thank you, have a nice day,
Stéphane
--
http://stephane.magnenat.net
11 Feb
11 Feb
10:46 p.m.
Back on Wednesday 11 February 2009 01:47:31 am Stéphane Magnenat was saying
stuff like:
If you want to model the the ambient reflections with that type of physical
accuracy, you would probably save yourself a lot of trouble by finding an
actual performance venue that you want to model after, and get several stereo
impulse response samples from the desired listening location, with the sources
coming from the different locations on stage that you are wanting to use as
instrument sources.
-Reuben
There is
another program by the same author of the tap plugins called
reverbed that lets you edit just about every aspect of the reflections.
I've already played with it. It is very nice indeed, but it requires
handcrafting of all parameters. For instance, one has to specify each early
reflection by hand. It would be much more convenient to be able to specify
the size of the room, the position and size of the instrument, the position
of the listener, and some parameters for the walls ; and let the program
compute the corresponding early reflections patterns. It is probably not
that difficult to do, as it only requires basic 2D geometry. The main
issues that I see are:
- How many early reflections to take into account?
- What is the relation between the room dimensions and the allpass
impulses? - How to make nice use of stereo input; for instance, should the
left wall early reflection use the left channel source positioned at the
most-left point of the instrument area, and the corresponding layout for
the right one? If these questions find answers easily, building an
easy-to-use reverb editor for ladspa is probably only a matter of dumb
coding, by extracting the TAP reverberator source code and fusing it to a
GUI application for instrument and listener positioning.
12 Feb
12 Feb
6:06 a.m.
If you want to model the the ambient reflections with
that type of physical
accuracy, you would probably save yourself a lot of trouble by finding an
actual performance venue that you want to model after, and get several
stereo impulse response samples from the desired listening location, with
the sources coming from the different locations on stage that you are
wanting to use as instrument sources.
Yes, but that requires specific hardware, where a more mathematical model does
not. Speaking about impulse response modernization, is there any open source
reverb tools that use them?
Thank you, have a nice day
Stéphane
--
http://stephane.magnenat.net
7:29 p.m.
This story was shown on TV here last night
http://www.abc.net.au/catalyst/stories/2488923.htm
It is a detailed look at the new recital auditorium in Melbourne, Australia.
The designer walks us through and explains what he has done to deal with high, mid and low
frequency reflections etc.
About 5-10 minutes duration.
Norv
Make Yahoo!7 your homepage and win a trip to the Quiksilver Pro. Find out more
10:18 p.m.
That's amazing. I live in Melbourne too, must go see a concert there.
Norval Watson wrote:
This story was shown on TV here last night
http://www.abc.net.au/catalyst/stories/2488923.htm
It is a detailed look at the new recital auditorium in Melbourne, Australia.
The designer walks us through and explains what he has done to deal with high, mid and
low frequency reflections etc.
About 5-10 minutes duration.
Norv
13 Feb
13 Feb
12:17 a.m.
My parents live there and love chamber music so I might get to see/hear it with them one
day.
That's amazing. I live in Melbourne too, must go
see a concert there.
Norval Watson wrote:
>http://www.abc.net.au/catalyst/stories/2488923.htm
Make Yahoo!7 your homepage and win a trip to the Quiksilver Pro. Find out more
5838
days inactive
5840
days old
linux-audio-user@lists.linuxaudio.org
17 comments
10 participants
participants (10)
-
Dave Phillips -
drew Roberts -
Fons Adriaensen -
hermann meyer -
Julien Claassen -
M Watts -
Norval Watson -
Reuben -
Roger E -
Stéphane Magnenat