The tdfb_controls.m4 is changed for better 15 degree control enum
spacing for beam direction. The file is with this change dedicated
for line arrays with physical limitation to -90 to +90 degrees that
is just achievable with 16 max. enum values.
The new tdfb_controls360.m4 uses the previous 30 degree spacing for
360 degrees arrays such as circular. The pipeline macros for 360
degrees are added to pipelines macros. The difference is only include
of 360 degrees controls version.
The topology development CMakeList.txt is changed to build topologies
with TDFB with single dual mono -90 to +90 degree angle steerable
beam. The stereo dual beam setup files are preserved and updated but
currently used only for testbench test for more complex configuration.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This patch contains improvements to TDFB setup scripts and examples
those are related to sound direction estimation. The changes are add
of 15 degree angle spacing to enum control, add of rotated noise sound
create for sound direction testing in simulated acoustics.
The example line, circular, and xyz array scripts are updated for single
dual mono beam that is controllable in 15 degree or 30 degree steps. All
the two beam design examples are moved into example_two_beams.m.
The patch also contains noise field covariance matrix diagonal load
increase to -40 dB. I noticed uneven frequency response in some
designs due to numerical instability. The increased diagonal load makes
the automatic design procedure more robust at minimal hit in noise
suppression performance.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
Added singular value decomposition algorithm
for diagonal matrix calculation. Added filters for side
lobe artifacts.
Signed-off-by: ShriramShastry <malladi.sastry@intel.com>
Topologies 1.0 location has been changed. This update allows the
tool to create the files into correct path.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
The delay-and-sum beamformer has highest white noise gain (WNG)
performance. It's good id suppressing microphone self noise. The
beam is steered with pure delay adjust in the all-pass filter
bank. Default type is still super directive. Use DSB or SDB
as beamformer type to select.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
Matlab does not support isfile(), so another command is used to
check if a file exists. The new way works both in Octave and Matlab.
The bf_defaults.m is fixed to properly initialize empty strings
to cell arrays. It's needed with minimal command like usage
as shown in SOF Docs examples.
In bf_design.m the num_filters determination is moved after array
helpers because other than line and circle do not use mic_n parameter.
The bug resulted to zero num_filters.
In bf_merge.m the unused bf3 and bf4 are removed.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This patch essentially adds support for steer_az/steer_el to be
vector instead of scalar. The beam off preset is added by default
automatically for N-channel in N-channel out configurations. Microphone
(x,y,z) coordinates are passed in to blob for direction of arrival
reporting.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This patch fixes the incorrect delaying of simulation input signal to
microphones. Instead of delaying the old code advanced the input the
more the larger the input is. E.g. It caused the tests for line arrays
to show a mirrored beam pattern.
The num_filters is replaced by mic_n (microphones count). It is the
correct count to use here though usually filters count is the same.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
The plots help to understand beamformer channel filters characteristic.
Frequency response is usually all-pass type and group delays match
the delays need to time-align the waveforms from steer angle.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This patch improves the beamformer performance (vs. tap count) by
shifting the extract window location for IFFT output by mean of
all impulse responses center. Previously the first filter was centered
and other filters were more towards start or end of impulse response.
The change helps to cope with smaller filter length with larger size
arrays with steer angles those spread the impulse response further
away time-wise from each other.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
The bf_defaults() now sets the default directory locations for
for bf_export() so it does not need to be set in every script.
Also the defaults for input channels and output mixing are set to
that in typical design they do not need to be set.
The default distance for look direction is set to 5.0m so it is better
guaranteed in far field for even relatively large arrays. The far field
assumes planar sound waves. The design equations do not fully support
near field with spherical waves.
The FFT length is changed to 1024 to enable design of longer filter bank
filters. There's no compromise in performance for for e.g. length 64.
The incorrect use of mic_n (number of microphones) is replaced by
num_filters. It's possible define beamformer that ignores some mic
channels so these are not the same.
The automatic systematic design file names are now created for all array
types, not just line.
The script is made more flexible to run without need to export data and
clutter just to quickly interactively check from plots what kind of
beam pattern and other characteristic some array, e.g. circular with
some radius and mic count gives.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This patch adds the missing XYZ 3D array helper and does minor fixes
for two others. The L-shape microphones orientation is made similar as
in rectangular array by inverting the y-axis. The missing centering
of rectangular array is added.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
This patch adds the tool for creating beamformer configurations. The
microphone array geometry and beam angle (azimuth, elevation) need
to be specified. See the example scripts and sample array helper
functions.
The FIR blob quantize function needed a minor change to prevent strip
of trailing zero coefficients. The beamformer filter bank needs to use
equal length filters.
Signed-off-by: Seppo Ingalsuo <seppo.ingalsuo@linux.intel.com>
Seppo Ingalsuo