STFT --> inverse_STFT not reconstructing the original signal when using no window

[giamic]

Tensorflow is not reconstructing the original signal when applying the STFT followed by the inverse STFT when using no window. The problems arise when the frames of the STFT overlap: It seems like every frame contributes with a weight of 1 regardless of the number of overlapping frames N = frame_size / frame_step. As a result, the central part of the signal is N times larger than the original. Here is a minimal code to reproduce the error:

import matplotlib.pyplot as plt
import numpy as np
import tensorflow as tf

size = 2048
frame_length = 512
frame_step = 128
waveform = np.sin(np.arange(size) * 1 / 100)

stft = tf.signal.stft(waveform, frame_length, frame_step, window_fn=None)
inverse_stft = tf.signal.inverse_stft(stft, frame_length, frame_step, window_fn=None)

plt.plot(waveform)
plt.plot(inverse_stft)
plt.show()
plt.clf()

Broken stft --> istft applied to sinusoidal signal:

Notice that I'm using no window. If I put the Hann window, the central part works well but the borders are smoothly going to zero, a related but surprisingly different bug that is already documented here: #36616. The implementation of scipy works well under all circumstances.

I'm using Tensorflow 2.6.0 installed with tensorflow-macos on an Apple M1 MacBookPro.

[mohantym]

Hi @sanatmpa1! Could you please look at this issue. It's replicating in Tf 2.5 ,2.6 and 2.7 . Attaching Gist for reference.