4 files changed, 41 insertions(+), 4 deletions(-)

M REFACTOR
M libpippi/src/mir.c
M pippi/buffers.pyx
M tests/test_newbuffer.py

@@ 56,8 56,8 @@ Just some refactoring notes...
 - [x] speed
 - [x] vspeed
 - [x] taper
-- [ ] toenv
-- [ ] towavetable
+- [x] toenv
+- [x] towavetable
 - [x] trim
 - [x] write
 

@@ 224,7 224,7 @@ lpenvelopefollower_t * envelopefollower_create(lpfloat_t interval) {
 
 void envelopefollower_process(lpenvelopefollower_t * env, lpfloat_t input) {
     env->phase += 1;
-    env->last = lpfmax(env->last, lpfabs(input));
+    env->last = (lpfloat_t)fmax(env->last, (lpfloat_t)fabs(input));
     if(env->phase >= env->interval) {
         env->phase -= env->interval;        
         env->value = env->last;

@@ 906,7 906,7 @@ cdef class SoundBuffer:
         return SoundBuffer.fromlpbuffer(out)
 
     def vspeed(SoundBuffer self, object speed, str interpolation=None):
-        warnings.warn('DEPRECATED: SoundBuffer.vspeed() is deprecated. Please just use SoundBuffer.speed()', DeprecationWarning)
+        warnings.warn('SoundBuffer.vspeed() is deprecated. Please use SoundBuffer.speed()', DeprecationWarning)
         return self.speed(speed, interpolation)
 
     def taper(self, double start, double end=-1):


@@ 923,6 923,39 @@ cdef class SoundBuffer:
         LPBuffer.taper(out, <size_t>(start * self.samplerate), <size_t>(end * self.samplerate))
         return SoundBuffer.fromlpbuffer(out)
 
+    def toenv(SoundBuffer self, double window=0.01):
+        cdef lpbuffer_t * out
+        cdef size_t blocksize = <size_t>int(window * self.samplerate)
+        cdef size_t length = <size_t>int(len(self) / float(blocksize))
+        cdef size_t i, b
+        cdef lpfloat_t sample
+        cdef int processing = 1
+
+        out = LPBuffer.create(length, 1, self.samplerate)
+
+        i = 0
+        while True:
+            # find the peak in the block
+            sample = 0
+            for b in range(blocksize):
+                sample = max(sample, abs(sum(self[i*blocksize+b])))
+            sample /= self.channels
+
+            if i*blocksize+blocksize >= len(self)-blocksize:
+                break
+
+            out.data[i] = sample
+            i += 1
+
+        return SoundBuffer.fromlpbuffer(out)
+
+    def towavetable(SoundBuffer self, *args, **kwargs):
+        warnings.warn('SoundBuffer.towavetable() is deprecated. Use SoundBuffers as wavetables directly. Please use SoundBuffer.remix(1) to mix down to a single channel.', DeprecationWarning)
+        if self.channels == 1:
+            return self
+        cdef lpbuffer_t * out = LPBuffer.remix(self.buffer, 1)
+        return SoundBuffer.fromlpbuffer(out)
+
     def trim(SoundBuffer self, bint start=False, bint end=True, double threshold=0, int window=4):
         """ Trim silence below a given threshold from the end (and/or start) of the buffer
         """

@@ 544,6 544,10 @@ class TestNewBuffer(TestCase):
         out = sound.speed(speed)
         out.write('tests/renders/newbuffer_vspeed_5_50.wav')
 
+    def test_toenv(self):
+        snd = SoundBuffer(filename='tests/sounds/linux.wav')
+        env = snd.toenv()
+        env.write('tests/renders/newbuffer_toenv.wav')
 
 """
     def test_stack_soundbuffer(self):