我想出了这个解决方案。更改
freq为
newfreq，然后更改阶段，如下所示：
newphase = 2*np.pi*t*(freq-newfreq)+phase
import pyaudioimport numpy as npfrom time import timeCHANNELS = 2RATE = 44100TT = time()freq = 100newfreq = 100phase = 0def callback(in_data, frame_count, time_info, status): global TT,phase,freq,newfreq if newfreq != freq: phase = 2*np.pi*TT*(freq-newfreq)+phase freq=newfreq left = (np.sin(phase+2*np.pi*freq*(TT+np.arange(frame_count)/float(RATE)))) data = np.zeros((left.shape[0]*2,),np.float32) data[::2] = left data[1::2] = left TT+=frame_count/float(RATE) return (data, pyaudio.paContinue)p = pyaudio.PyAudio()stream = p.open(format=pyaudio.paFloat32, channels=CHANNELS, rate=RATE, output=True, stream_callback=callback)stream.start_stream()start = time()try: while 1: now = time() if now-start>1/24.: newfreq=200+np.sin(2*np.pi*1/20.*now)*100 #update the frequency This will depend on y on the future print newfreq start=nowfinally: stream.stop_stream() stream.close() p.terminate()
