tried mean/slope closed loop but floperoo

git-svn-id: https://svn.code.sf.net/p/freetel/code@3262 01035d8c-6547-0410-b346-abe4f91aad63

diff --git a/codec2-dev/octave/newamp.m b/codec2-dev/octave/newamp.m
index b1b8df4caa510ff0624e3f018c77f6dc2f3b25f5..7525817115307a5a67373847ac025e30ab2b7542 100644 (file)
--- a/codec2-dev/octave/newamp.m
+++ b/codec2-dev/octave/newamp.m
@@ -518,7 +518,7 @@ function [rate_K_surface rate_K_sample_freqs_kHz] = resample_const_rate_f(model,
 
     rate_L_sample_freqs_kHz = (1:L)*Wo*Fs/(2000*pi);
     
-    rate_K_surface(f,:) = interp1(rate_L_sample_freqs_kHz, AmdB, rate_K_sample_freqs_kHz, "spline", "extrap");
+    rate_K_surface(f,:) = interp1([0 rate_L_sample_freqs_kHz (Fs/2000)], [AmdB(1) AmdB AmdB(L)], rate_K_sample_freqs_kHz, "spline");
     %rate_K_surface(f,:)  = interp_para(rate_L_sample_freqs_kHz, AmdB, Fs/(2*1000), rate_K_sample_freqs_kHz);    
 
     %printf("%d\n", f);
@@ -714,6 +714,54 @@ function [Wo_ voicing_] = interp_Wo_v(Wo1, Wo2, voicing1, voicing2)
 endfunction
 
 
+function [diff_weighted weights error g slope min_ind] = search_vq_weighted(target, vq, weight_gain, fit_order)
+  [vq_rows vq_cols] = size(vq);
+
+  % find mse for each vector
+
+  error = g = slope = zeros(1, vq_rows);
+  diff =  weights = diff_weighted = zeros(vq_rows, vq_cols);
+
+  for i=1:vq_rows
+
+    if fit_order == 0
+
+      % work out gain for best match
+
+      g(i) = sum(target - vq(i,:))/vq_cols;
+      slope(i) = 1;
+    end
+
+    if fit_order == 1
+      % work out linear + gradient fit 
+      vi = vq(i,:);
+      A = [sum(vi) vq_cols; vi*vi' sum(vi)];
+      c = [sum(target) target*vi']';
+      b = inv(A)*c;
+      g(i) = b(2);
+      slope(i) = b(1);
+    end
+
+    % Find weighted difference.  This allocated more importance
+    % (error) to samples with higher energy, and stops really low
+    % level harmonics from having any impact.  Note addition in dB
+    % is multiplication in linear
+
+    diff(i,:) = target - slope(i)*vq(i,:) - g(i);
+    weights(i,:) = max(0.1, weight_gain .* (target + 20));
+ 
+    diff_weighted(i,:) = diff(i,:) .* weights(i,:);
+
+    % abs in dB is MSE in linear
+
+    error(i) = sum(abs(diff_weighted(i,:)));
+  end
+
+  [mn min_ind] = min(error)
+  
+endfunction
+
+
 % --------------------------------------------------------------------------------
 % Experimental functions used for masking, piecewise models, not part of newamp1
 % --------------------------------------------------------------------------------
@@ -813,7 +861,7 @@ function maskdB = determine_mask(masker_amps_dB, masker_freqs_kHz, mask_sample_f
 
     % calculate and plot masking curve
 
-    maskdB = zeros(1,length(mask_sample_freqs_kHz));
+    maskdB = -20*ones(1,length(mask_sample_freqs_kHz));
     for m=1:length(masker_freqs_kHz)
       maskdB = max(maskdB, schroeder(masker_freqs_kHz(m), mask_sample_freqs_kHz, bark_model) + masker_amps_dB(m)); 
       %maskdB = max(maskdB, parabolic_resonator(masker_freqs_kHz(m), mask_sample_freqs_kHz) + masker_amps_dB(m));