diff --git a/speech_data.py b/speech_data.py
index 23d741f..c7356b6 100644
--- a/speech_data.py
+++ b/speech_data.py
@@ -83,11 +83,10 @@ def create_spectrogram_data(audio_group='audio'):
         , quoting=csv.QUOTE_NONE)
     # audio_samples = audio_samples.loc[audio_samples['word'] ==
     #                                   'sunflowers'].reset_index(drop=True)
-    file_names = audio_samples.loc[:, 'file'].apply(lambda x: 'outputs/' + audio_group + '/' + x)
-    audio_samples['file_exists'] = apply_by_multiprocessing(file_names,os.path.exists)
-    audio_samples = audio_samples[audio_samples['file_exists'] == False]
-    audio_samples['spectrogram'] = apply_by_multiprocessing(file_names,generate_aiff_spectrogram)#.apply(
-            #generate_aiff_spectrogram)
+    audio_samples['file_paths'] = audio_samples.loc[:, 'file'].apply(lambda x: 'outputs/' + audio_group + '/' + x)
+    audio_samples['file_exists'] = apply_by_multiprocessing(audio_samples['file_paths'], os.path.exists)
+    audio_samples = audio_samples[audio_samples['file_exists'] == True].reset_index()
+    audio_samples['spectrogram'] = apply_by_multiprocessing(audio_samples['file_paths'],generate_aiff_spectrogram)#.apply(
     audio_samples['window_count'] = audio_samples.loc[:,'spectrogram'].apply(lambda x: x.shape[0])
     audio_samples.to_pickle('outputs/{}-spectrogram.pkl'.format(audio_group))