Re: spatial & more
Posted: Thu Oct 24, 2013 10:49 am
As I said I did some experiments with in-ear binaural mics, and:
1) the spectral difference between my recordings and the "matchbox reference" is, that my recording use wide range of spectra to cover vertical axis of location, while matchobx seem to use only narrow and defined range of high spectra.
2) because my recordings were made with my ears - both recordings (my and matchbox) - sound in the same way to me in terms of spatialization and location (to other ears - some folks hear vertical location, some folks hear that sound becomes distant), althogh different aspects (band width) of sound seem to create that effect.
Just out of curiosity - I wonder if there is a way to research that part "as is". Reverse engineering.
So how plugins like Q-clone (from Waves) work? I suspect, that the part with "filtering" profiles - could be done this way? Taking short portions of sound, and averaging the spectral curve over time?
Generally I suspect, that positional aspects of location are rather present in sounds in motion, than statiic, so even adding a filter with morphed characteristic could be interesting here. Note, that I'm not necessarily focused on replicating human head (spatial dimensions 1:1). I'm rather interested in "exposing" certain aspects of spatialization and location, beyond typical boundaries, defined by replication.
1) the spectral difference between my recordings and the "matchbox reference" is, that my recording use wide range of spectra to cover vertical axis of location, while matchobx seem to use only narrow and defined range of high spectra.
2) because my recordings were made with my ears - both recordings (my and matchbox) - sound in the same way to me in terms of spatialization and location (to other ears - some folks hear vertical location, some folks hear that sound becomes distant), althogh different aspects (band width) of sound seem to create that effect.
Just out of curiosity - I wonder if there is a way to research that part "as is". Reverse engineering.
So how plugins like Q-clone (from Waves) work? I suspect, that the part with "filtering" profiles - could be done this way? Taking short portions of sound, and averaging the spectral curve over time?
Generally I suspect, that positional aspects of location are rather present in sounds in motion, than statiic, so even adding a filter with morphed characteristic could be interesting here. Note, that I'm not necessarily focused on replicating human head (spatial dimensions 1:1). I'm rather interested in "exposing" certain aspects of spatialization and location, beyond typical boundaries, defined by replication.
