diff --git a/applications/SphericalHarmonics/ita_sph_mimo_error_simulation.m b/applications/SphericalHarmonics/ita_sph_mimo_error_simulation.m index 979af99c101407a58eda1f67c8fbc92d42f9f5c7..0af3103f5f82f33a06893a384973bd698ba32b03 100644 --- a/applications/SphericalHarmonics/ita_sph_mimo_error_simulation.m +++ b/applications/SphericalHarmonics/ita_sph_mimo_error_simulation.m @@ -41,6 +41,7 @@ sArgs = struct('pos1_source','struct',... 'samplingRate',ita_preferences('samplingRate'),... 'freqRange',[20 ita_preferences('samplingRate')/2],... 'samplingDisplacement',[],... + 'samplingDisplacementAbsolute',false,... 'nRuns',1,... 'SNR',60,... 'sma',true,... @@ -66,6 +67,7 @@ simSLA = sArgs.sla; SNR = sArgs.SNR; samplingDisplacement = sArgs.samplingDisplacement; +displacementType = sArgs.samplingDisplacementAbsolute; ao = ita_generate_impulse('fftDegree',sArgs.fftDegree,'samplingRate',sArgs.samplingRate); freqVec = ao.freqVector(ao.freq2index(sArgs.freqRange(1)):ao.freq2index(sArgs.freqRange(2))); @@ -269,19 +271,27 @@ for idxRun = 1:sArgs.nRuns if ~isempty(samplingDisplacement) && isempty(SNR) if simSMA - receiverSamplingErroneous = ita_sampling_displacement(receiverSampling,'relativeError',samplingDisplacement); + receiverSamplingErroneous = ita_sph_sampling_displacement(receiverSampling,samplingDisplacement,'absolute',displacementType); receiverYmismatch = ita_sph_base(receiverSamplingErroneous,receiverNmax); EreceiverMismatch = ita_sph_modal_strength(receiverSamplingErroneous,receiverNmax,kVec(idxFreq),'rigid'); - EreceiverMismatch = EreceiverMismatch.*receiverYmismatch - Mreceiver; + if receiverSamplingUniqueRad + EreceiverMismatch = receiverYmismatch*EreceiverMismatch - Mreceiver; + else + EreceiverMismatch = receiverYmismatch.*EreceiverMismatch - Mreceiver; + end else EreceiverMismatch = []; end if simSLA - sourceSamplingErroneous = ita_sampling_displacement(sourceSampling,'relativeError',samplingDisplacement); + sourceSamplingErroneous = ita_sph_sampling_displacement(sourceSampling,samplingDisplacement,'absolute',displacementType); sourceYmismatch = ita_sph_base(sourceSamplingErroneous,sourceNmax); sourceGmismatch = ita_sph_aperture_function_sla(sourceSamplingErroneous,sourceNmax,sourcerMem,'r',unique(sourceSamplingR)); EsourceMismatch = ita_sph_modal_strength(sourceSamplingErroneous,sourceNmax,kVec(idxFreq),'rigid','transducer','ls'); - EsourceMismatch = (EsourceMismatch.' .* (sourceGmismatch.'.*sourceYmismatch')) - Msource; + if numel(unique(sourceSampling.r)) == 1 + EsourceMismatch = (EsourceMismatch * (sourceGmismatch.'.*sourceYmismatch')) - Msource; + else + EsourceMismatch = (EsourceMismatch.' .* (sourceGmismatch.'.*sourceYmismatch')) - Msource; + end else EsourceMismatch = []; end diff --git a/applications/SphericalHarmonics/ita_sph_sampling/ita_sph_sampling_displacement.m b/applications/SphericalHarmonics/ita_sph_sampling/ita_sph_sampling_displacement.m index 8d393088f9cc89cda09a3b6e4f4b521023dc2794..1869b7ead14c5163f6845d5d6b435d14f31101ef 100644 --- a/applications/SphericalHarmonics/ita_sph_sampling/ita_sph_sampling_displacement.m +++ b/applications/SphericalHarmonics/ita_sph_sampling/ita_sph_sampling_displacement.m @@ -1,19 +1,24 @@ function varargout = ita_sph_sampling_displacement(varargin) %ITA_SPH_SAMPLING_DISPLACEMENT - displaced sampling positions % This function creates a sampling grid with an additive error in the sampling -% positions taken from a given sampling grid +% positions taken from a given sampling grid. The default option for the displacement +% is a relative displacement in percent/100. For a absolute displacement in meters +% choose the option 'absolute'. % % Syntax: -% samplingDisplaced = ita_sph_sampling_displacement(sampling, opts) +% samplingDisplaced = ita_sph_sampling_displacement(sampling, displacement, opts) % % Options (default): -% 'relativeError' ([0.1,0.1,0.1]) : relative displacement in percent/100 +% 'absolute' (false) : displacement is a absolute value +% 'weightPoles' (true): weigting for smaller errors towards the poles +% 'projectRad' (true) : project displaced sampling back onto the original radius +% % % Example: -% samplingDisplaced = ita_sph_sampling_displacement(sampling, [0.01,0.01,0.01]) +% samplingDisplaced = ita_sph_sampling_displacement(sampling, [5/10,5/10,5/10] * 1e-3, 'absolute') % % See also: -% ita_toolbox_gui, ita_read, ita_write, ita_generate +% ita_sph_sampling, ita_sph_mimo_error_simulation % % Reference page in Help browser % doc ita_sph_sampling_displacement @@ -27,39 +32,41 @@ function varargout = ita_sph_sampling_displacement(varargin) % Author: Marco Berzborn -- Email: marco.berzborn@akustik.rwth-aachen.de % Created: 29-Mar-2016 - -%% Initialization and Input Parsing - sArgs = struct('pos1_sampling','itaCoordinates',... - 'relativeError',[0.01 0.01 0.01]); -[sampling, sArgs] = ita_parse_arguments(sArgs,varargin); + 'pos2_error','double',... + 'absolute',false,... + 'weightPoles',true,... + 'projectRad',true); +[sampling, displacement, sArgs] = ita_parse_arguments(sArgs,varargin); -if isempty(sArgs.relativeError) - sArgs.relativeError = zeros(1,3); -elseif numel(sArgs.relativeError) == 1 - sArgs.relativeError = repmat(sArgs.relativeError,1,3); +if numel(displacement) == 1 + displacement = repmat(displacement,1,3); end -posError = randn(size(sampling.sph)); -posError = (bsxfun(@rdivide,posError,sqrt(sum(posError.^2,2))) .* repmat(sampling.sph(:,1),1,3)) * diag(sArgs.relativeError); +if ~sArgs.absolute + posError = randn(size(sampling.sph)); + posError = (bsxfun(@rdivide,posError,sqrt(sum(posError.^2,2))) .* repmat(sampling.sph(:,1),1,3)) * diag(displacement); +elseif sArgs.absolute + posError = randn(size(sampling.sph)); + posError = bsxfun(@rdivide,posError,sqrt(sum(posError.^2,2))) * diag(displacement); +end posError = itaCoordinates(posError,'cart'); -posError = posError.makeSph; -% consideration of the smaller deviations towards the poles -% Ref: Rafaely - 2005 - Analysis and Design of Spherical Microphone Arrays -posError.sph(:,2) = posError.sph(:,2)./sin(sampling.theta); - -posError = posError.makeCart; +if sArgs.weightPoles + % consideration of the smaller deviations towards the poles + % Ref: Rafaely - 2005 - Analysis and Design of Spherical Microphone Arrays + posError.sph(:,2) = posError.sph(:,2)./sin(sampling.theta); -% set all NaNs and Infs to zero as they are not wanted and lead to missing -% sampling points. -posError.cart(isnan(posError.cart)) = 0; -posError.cart(isinf(posError.cart)) = 0; + % set all NaNs and Infs to zero as they are not wanted and lead to missing + % sampling points. + posError.cart(isnan(posError.cart)) = 0; + posError.cart(isinf(posError.cart)) = 0; +end % new sampling coordinates with erroneous positions samplingError = itaCoordinates(sampling.cart + posError.cart,'cart'); -if sArgs.relativeError(:,1) == 0 +if sArgs.projectRad samplingError.sph(:,1) = sampling.sph(:,1); end