error in applying trained mlp classifier model to new audio chunk

[DolphinRidz]

file = 'chunkchannel_1.wav'
ans =[]
new_feature= extract_feature(file, mfcc=True, chroma=True, mel=True)
ans.append(new_feature)
ans = np.array(ans)
print(np.shape(ans))
Emotion_Voice_Detection_Model.predict([ans])

I have trained my model using the RAVDESS dataset for emotion recognition. I want to apply the same model to audio chunks that I have downloaded from youtube and converted to .wav format.
It's giving error as
Found array with dim 3. Estimator expected <= 2.