From a7256287f599dbaaffb107df0b8b4ea4a2b1f2b1 Mon Sep 17 00:00:00 2001
From: jules on spawn <spawn@cortex>
Date: Mon, 9 Apr 2018 00:13:27 +0200
Subject: dataset script

---
 datasets/dataset.pl | 67 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 datasets/spread.sh  |  8 +++++++
 2 files changed, 75 insertions(+)
 create mode 100755 datasets/dataset.pl

(limited to 'datasets')

diff --git a/datasets/dataset.pl b/datasets/dataset.pl
new file mode 100755
index 0000000..4e3b9a9
--- /dev/null
+++ b/datasets/dataset.pl
@@ -0,0 +1,67 @@
+#!/usr/bin/perl
+
+use strict;
+use Getopt::Std;
+
+our $opt_c;
+getopts('c');
+
+my $fmt = <<FMT;
+Input File     : '01_snapping_clean.wav'
+Channels       : 1
+Sample Rate    : 44100
+Precision      : 16-bit
+Duration       : 00:03:09.35 = 8350438 samples = 14201.4 CDDA sectors
+File Size      : 16.7M
+Bit Rate       : 706k
+Sample Encoding: 16-bit Signed Integer PCM
+FMT
+
+sub process($) {
+  my $filename = shift or die "Usage: $0 [...filenames]\n";
+  my ($name, $ext) = split(/\./, $filename, 2);
+
+  my $soxi = `soxi $filename`;
+  my @lines = split("\n", $soxi);
+
+  print $soxi;
+
+  my $seconds;
+  for my $line (@lines) {
+    if ($line =~ /Duration       : (\d\d):(\d\d):(\d\d)\./) {
+      my $h = $1;
+      my $m = $2;
+      my $s = $3;
+      $seconds = (((($h * 60) + $m) * 60) + $s) + 0;
+    }
+  }
+
+  my $scale = sprintf("%.09f", 5e-7 * $seconds);
+
+  print "Seconds: $seconds\n";
+  print "Scale factor: $scale\n";
+  print "\n";
+
+  my $a_tmp = "a_" . $filename;
+  my $b_tmp = "b_" . $filename;
+  if ($opt_c) {
+    print "Normalizing...";
+    system("sox", "-v", 0.945, $filename, $a_tmp);
+  } else {
+    $a_tmp = $filename;
+  }
+  system("./spread.sh", $a_tmp, $b_tmp, 0.999, $scale, 1.001);
+  system("./split44k.sh", $b_tmp, 8, $name);
+  if ($opt_c) {
+    system("/bin/rm", $a_tmp);
+  }
+  system("/bin/rm", $b_tmp);
+}
+
+foreach my $file (@ARGV) {;
+  if ( -e $file ) {
+    process($file);
+  }
+}
+
+
diff --git a/datasets/spread.sh b/datasets/spread.sh
index bec1da3..60c3551 100755
--- a/datasets/spread.sh
+++ b/datasets/spread.sh
@@ -1,3 +1,4 @@
+#!/bin/bash
 
 if [ "$#" -ne 5 ]; then
     echo "Usage: $0 <in.wav> <out.wav> <rate_min> <rate_step> <rate_max>"
@@ -11,15 +12,22 @@ STEP=$4
 MAX=$5
 
 ITER=0
+
 while true; do
   if (( $(echo "$RATE > $MAX" | bc -l) )); then
     break
   fi
   let ITER+=1
   RATE=`echo "$RATE+$STEP" | bc`
+  if ((ITER % 20 != 0)); then
+    echo "${ITER}... ${RATE}"
+    sleep 1
+  fi
   sox $FN_IN "tmp_$ITER.wav" speed $RATE
 done
 
+echo "made $ITER copies"
+
 sox tmp_* $FN_OUT
 rm tmp_*
 
-- 
cgit v1.2.3-70-g09d2