Distracted
Joined: 20/01/08
Posts: 84
Loc: London England
|
Binaural Recording Live vs IR Post Production
#1001873 - 05/08/12 02:01 PM
|
|
|
|
Hi
I have read and listened to alot of recordings using various techniques. The
overall consensus seems to be that IRs do not seem to convey the same sense of space as a
spaced pair of mics using the variety of possible techniques. Does anyone know if this is
down to IRs representing a snapshot of what is going on or is there something more
fundamental which is going on?. It seemed to me like the IRs fail to represent something
which our brains use to finesse the localising other than delays and filtering, perhaps
some kind of modulation that the reflected signals apply to the resultant L&R
signals!
Any links that might help would be gratefully appreciated!
Kenny
|
Hugh Robjohns
SOS Technical Editor
Joined: 25/07/03
Posts: 18403
Loc: Worcestershire
|
Re: Binaural Recording Live vs IR Post Production
[Re: Distracted]
#1001908 - 05/08/12 10:14 PM
|
|
|
|
Depends on how you capture the impulse responses. I've heard very convincing stereo
imaging from carefully captured stereo IRs.
H
--------------------
Technical Editor, Sound On Sound
|
Distracted
Joined: 20/01/08
Posts: 84
Loc: London England
|
Re: Binaural Recording Live vs IR Post Production
[Re: Hugh Robjohns]
#1001920 - 06/08/12 06:28 AM
|
|
|
Hi Hugh do you mean using a sine sweep as opposed to a spike, i had wondered if
the sine sweep might better represent the imaging. Now you are going to tell me the
inverse is true!  Kenny
|
Distracted
Joined: 20/01/08
Posts: 84
Loc: London England
|
Re: Binaural Recording Live vs IR Post Production
[Re: Distracted]
#1001921 - 06/08/12 06:44 AM
|
|
|
|
Hi Hugh
Were the IRs you heard from commercial software or were they recorded
by an individual?
Thanks Kenny
|
Hugh Robjohns
SOS Technical Editor
Joined: 25/07/03
Posts: 18403
Loc: Worcestershire
|
Re: Binaural Recording Live vs IR Post Production
[Re: Distracted]
#1001943 - 06/08/12 09:38 AM
|
|
|
Quote Distracted:
do you mean
using a sine sweep as opposed to a spike
No -- an impulse response is just that. It doesn't matter how you
acquire it... and what you're describing is the means of acquiring an impulse response
from a real space.
The 'spike' or 'click' is the obvious way, since that
delivers the impulse response directly. However, the click should be infinitely fast,
which is tricky since it contains no energy, and if it isn't infinitely brief there will
be frequency response anomalies.
Consequently, frequency sweeps or 'chirps'
are used more often because they put a usable amount of energy into the acoustic space --
and the chirps usually repeated 8 or 16 times and averaged to reduce the ambeint recording
noise, before being transformed from the frequency domain into the time domain.
The room response can only ever be as good as the impulse response acquired, and if you
are looking for a stereo representation the impulse responses have to be acquired using
the same microphone technique placed in the same position as the stereo image you are
trying to create.
Few commercial IRs (if any?) wil be acquired with a binaural
technique.
You can't 'process' an IR -- it is what it is.
Hugh
--------------------
Technical Editor, Sound On Sound
|
Hugh Robjohns
SOS Technical Editor
Joined: 25/07/03
Posts: 18403
Loc: Worcestershire
|
Re: Binaural Recording Live vs IR Post Production
[Re: Distracted]
#1001944 - 06/08/12 09:43 AM
|
|
|
Quote Distracted:
Were the IRs
you heard from commercial software or were they recorded by an individual?
Both -- but for specific puposes. The
commercial ones were never released commercially, but were intended for sophisticated
surround headphone monitoring, recreating the 'sound' of 5.1 monitoring set ups in
established control rooms. I've heard some very convincing binaural IRs reatead by a
friend of mine a few years ago when he was experimenting with the system using an
expensive Yamaha hardware IR reverb processor.
Hugh
--------------------
Technical Editor, Sound On Sound
|
Distracted
Joined: 20/01/08
Posts: 84
Loc: London England
|
Re: Binaural Recording Live vs IR Post Production
[Re: Hugh Robjohns]
#1002027 - 06/08/12 06:38 PM
|
|
|
As ever i am impressed by your broad base of knowledge. I have read reviews of the
headphone systems and also played with the IRCAM IRs IRCAM
IRs download page I also experimented with the HEAR plugin. Although they all seem to handle extreme L and R images well there always feels like
anything with any distance near the centre just feels almost internal (like a traditional
stereo image), it always makes me feel that something about IRs removes part of the
process our brain uses to localise the sound. This is bourne out by recording the source
directly with the same rig used to capture the IR, removing the need for the processing.
|
Anonymous
Unregistered
|
Re: Binaural Recording Live vs IR Post Production
[Re: Distracted]
#1002173 - 07/08/12 01:01 PM
|
|
|
Hello, probably not much use here, but I stumbled on this NASA
document which might be of interest to engineers. I'm not sure if it's aimed at
providing astronauts with a more natural auditory experience or what. One of the related
subjects is 'quantum mechanics'.
'ow queer, oi said ti mesewf.
|
Anonymous
Unregistered
|
Re: Binaural Recording Live vs IR Post Production
[Re: Distracted]
#1002179 - 07/08/12 01:30 PM
|
|
|
Quote Distracted:
...it always
makes me feel that something about IRs removes part of the process our brain uses to
localise the sound.
Have
you compared actual live binaural recordings with equivalent IRs (i.e. captured in the
same situation)?
If not, perhaps the 'removed process' our brain uses to
localise sound, involves vision, touch, or imagined vision via knowledge of recorded
environment.
I mean, binaural recordings are more convincing when you're sat
in the same (or similar) space they were actually recorded. For example, the famous
binaural 'virtual haircut' is much more convincing when you're sat infront of a wall (like
a barber's shop). You can feel air/heat as well as other reflexions if you're stood near
of a wall that will better match the auditory information in the recording. It's pretty
amazing how all the senses are so finely integrated.
Also, knowledge has a
bigger influence on our perception than people realise. There has been research done (by
Stanford Graduate School of Business and the California Institute of Technology) into the
influence of wine labels, showing that people prefer mediocre wine with a high price tag
than better wine with a cheap price tag. It's not so much that they're blagging, but that
they actually taste the wine differently because of what they (think they) know about
it.
“Our senses don't deceive us: our judgment does.” ~Goethe
|
Distracted
Joined: 20/01/08
Posts: 84
Loc: London England
|
Re: Binaural Recording Live vs IR Post Production
[Re: ]
#1002229 - 07/08/12 07:01 PM
|
|
|
|
Hi Josif
i believe you posted about binaural recording late last year and a few
days ago i was attempting to draw your attention to the Roland CS 10EM mic/headphone combo
but i digress.
I recorded with microphones mounted in my ears and did both IR
and live recording. The IRs sound like they can convey a sausage shape of response whereas
the live recording is more spherical ( I would say like a horizontal rugby ball if you
will. I believe that we do use visual cues and familiarity to enhance the processing but i
can only describe what i hear. I have never heard a recording either IR or live which
localises sound in the way it occurs when presented directly in front of me (at any
distance). I accept that i cannot explain why this is but i would bet on my life that we
are missing something simple but fundamental and we haven't quite cracked binaural
recording yet!
Kenny
|
Hugh Robjohns
SOS Technical Editor
Joined: 25/07/03
Posts: 18403
Loc: Worcestershire
|
Re: Binaural Recording Live vs IR Post Production
[Re: Distracted]
#1002247 - 07/08/12 08:30 PM
|
|
|
|
Sight and realtime knowledge of what is going on round you has a huge influence on your
perception of sound, and that is missing from all recordings.
As for the IRs,
I think you might be misunderstanding what an IR does or can do.
An IR is a
record, if you like, of a specific signal path. In your case its the path from the test
speaker, across the room and it the binaural mics, including sound reflected around the
room.
However, it doesn't provide any information of how the signal path
changes for sources in different places in the room, so its never going to provide
accurate stereophony, only an impression of the room sound from a single point source.
H
--------------------
Technical Editor, Sound On Sound
|
Anonymous
Unregistered
|
Re: Binaural Recording Live vs IR Post Production
[Re: Hugh Robjohns]
#1002256 - 07/08/12 09:32 PM
|
|
|
Quote Hugh Robjohns:
Sight and
realtime knowledge of what is going on round you has a huge influence on your perception
of sound, and that is missing from all recordings.
I think though, with binaural IRs and recordings, the
comparative realism of the stereo image makes us notice what is not accurate, whereas with
standard stereo we're more used to directions being quite different from what we'd
naturally expect to hear.
|
Distracted
Joined: 20/01/08
Posts: 84
Loc: London England
|
Re: Binaural Recording Live vs IR Post Production
[Re: Hugh Robjohns]
#1002295 - 08/08/12 07:24 AM
|
|
|
|
Hi Hugh
i am probably not explaining myself very well i accept, so thanks for
your patient explanations. Referencing your reply i do feel i understand the basic premise
of what an IR is representing. I am trying to express something that feels missing so it
does feel like i am chasing my tail somewhat!
Kenny
|
Previous
Topics
Next
Threaded
Edit
Reply
Quote
