From 59513fc306421994710a6d226bdfea26b325ba41 Mon Sep 17 00:00:00 2001 From: ha7ilm Date: Tue, 28 Feb 2017 17:03:21 +0100 Subject: [PATCH] Fixed even more docs --- README.md | 4 ++++ libcsdr.c | 2 ++ 2 files changed, 6 insertions(+) diff --git a/README.md b/README.md index 117aea6..dbc3f45 100644 --- a/README.md +++ b/README.md @@ -329,6 +329,10 @@ It performs a bandpass FIR filter on complex samples, using FFT and the overlap- Parameters are described under `firdes_bandpass_c` and `firdes_lowpass_f`. + old_fractional_decimator_ff [num_poly_points [transition_bw [window]]] + +This is the deprecated, old version of `fractional_decimator_ff` (only uses linear interpolation, its filter cuts at 59% of the passband). + agc_ff [hang_time [reference [attack_rate [decay_rate [max_gain [attack_wait [filter_alpha]]]]]]] It is an automatic gain control function. diff --git a/libcsdr.c b/libcsdr.c index ff7415c..2d512b2 100644 --- a/libcsdr.c +++ b/libcsdr.c @@ -767,6 +767,7 @@ void fractional_decimator_ff(float* input, float* output, int input_size, fracti } /* + * Some notes to myself on the circular buffer I wanted to implement here: int last_input_samplewhere_shouldbe = (index_high-1)+xifirst; int last_input_offset = last_input_samplewhere_shouldbe - d->last_input_samplewhere; if(last_input_offset < num_poly_points) @@ -781,6 +782,7 @@ void fractional_decimator_ff(float* input, float* output, int input_size, fracti } d->last_input_samplewhere = d->las } + However, I think I should just rather do a continuous big buffer. */ void apply_fir_fft_cc(FFT_PLAN_T* plan, FFT_PLAN_T* plan_inverse, complexf* taps_fft, complexf* last_overlap, int overlap_size)