Code cleanup, minor bugs fixed

-> Fix related to audio offset
-> Temp moviepy audio file now cleaned up if left over

.gitignore

@@ -1,4 +1,8 @@
 *.pyc
+.idea
+
+mugen/testing.py
+
 output/
 segments/
 rejected_segments/

README.md

@@ -37,18 +37,18 @@ Below, an installation walthrough is provided for Mac OS X to give you a better
 
 ## Installation Walkthrough (Mac OS X)
 
-**1 - [Install Miniconda 3.5](http://conda.pydata.org/miniconda.html) (A Python virtual environment and package manager)**
+**1 - [Install Homebrew](http://brew.sh/) (General purpose package manager for mac)**
 
-**2 - [Install Homebrew](http://brew.sh/) (General purpose package manager for mac)**
+**2 - [Install Miniconda 3.5](http://conda.pydata.org/miniconda.html) (A Python virtual environment and package manager)**
 
-**3 - Ensure Miniconda & Homebrew are at top of path by appending below text to `~/.bash_profile`**
+**3 - Ensure Homebrew & Miniconda are at top of path by appending below text to `~/.bash_profile`**
 
 ```
-#Homebrew
-export PATH=/usr/local/bin:/usr/local/sbin:$PATH
-
 #Miniconda
 export PATH="~/miniconda/bin:$PATH"
+
+#Homebrew
+export PATH=/usr/local/bin:/usr/local/sbin:$PATH
 ```
 
 **4 - [Install tesseract](https://github.com/tesseract-ocr/tesseract) via Homebrew**

mugen/audio/audio.py

@@ -1,4 +1,3 @@
-import sys
 import shutil
 import pprint
 import logging

mugen/audio/utility.py

@@ -9,5 +9,5 @@ def load_audio(audio_file):
         print("Error reading audio file '{}'. Cannot continue. Error: {}".format(audio_file, e))
         sys.exit(1)
     else:
-    	audio = loader()
-    	return audio
+        audio = loader()
+        return audio

mugen/make_music_video.py

@@ -175,7 +175,7 @@ def parse_args(args):
     common_parser.add_argument('-db', '--debug', dest='debug', action='store_true', default=False, help='Pass in this argument to print useful debug info and save all rejected segments.')
 
     # Video Common Parameters
-    video_parser.add_argument('-o', '--output-name', dest='output_name', help='The name for the music video. Otherwise will output music_video_0' + s.OUTPUT_EXTENSION + ', music_video_1' + s.OUTPUT_EXTENSION + ', etc...')
+    video_parser.add_argument('-o', '--output-name', dest='output_name', help='The name for the music video. Otherwise will output ' + s.MUSIC_VIDEO_NAME_DEFAULT + '_0' + s.VIDEO_OUTPUT_EXTENSION + ', ' + s.MUSIC_VIDEO_NAME_DEFAULT + '_1' + s.VIDEO_OUTPUT_EXTENSION + ', etc...')
     video_parser.add_argument('-crf', '--crf', dest='crf', type=int, default=s.music_video_crf, help='The crf quality value for the music video. Takes an integer from 0 (lossless) to 51 (lossy). Defaults to 18.')
     video_parser.add_argument('-vd', '--video-dimensions', dest='video_dimensions', type=int, nargs=2, help='Pass in this argument to manually set the pixel dimensions for the music video, width and height. All video segments will be resized (cropped and/or scaled) appropriately to match these dimensions. Otherwise, the best dimensions for the music video are calculated automatically. Takes width then height integer values separated by spaces e.g., 1920 1080')
     video_parser.add_argument('-pvd', '--preserve-video-dimensions', dest='preserve_video_dimensions', action='store_true', default=False, help='Pass in this argument to preserve the various screen dimensions of the videos, and not perform any resizing.')

mugen/settings.py

@@ -9,9 +9,9 @@
 VERSION = "1.0"
 HELP = " Please review supported inputs and values on the help menu via --help"
 
-FILE_TYPE_AUDIO = "audio"
+FILE_TYPE_AUDIO = 'audio'
 FILE_TYPE_VIDEO = 'video'
-FILE_TYPE_SPEC = "spec"
+FILE_TYPE_SPEC = 'spec'
 
 ### VIDEO ###
 WIDESCREEN_ASPECT_RATIO = 16/float(9)
@@ -34,21 +34,22 @@
 AUDIO_TRACK_BEAT_LOCATIONS = 'beat_locations'
 SUBS_TRACK_SEGMENT_NUMBERS = 'segment_numbers'
 SUBS_TRACK_SEGMENT_DURATIONS = 'segment_durations'
-SUBS_TRACK_SPEC ='spec'
+SUBS_TRACK_SPEC = 'spec'
 
 ### PATHS ###
 TEMP_PATH_BASE = 'temp/'
+TEMP_MOVIEPY_AUDIOFILE = TEMP_PATH_BASE + 'moviepy_temp_audio.mp3'
 
 OUTPUT_PATH_BASE = 'output/'
-OUTPUT_NAME_DEFAULT = 'music_video'
-OUTPUT_EXTENSION = '.mkv'
+MUSIC_VIDEO_NAME_DEFAULT = 'music_video'
+VIDEO_OUTPUT_EXTENSION = '.mkv'
 SUBTITLES_EXTENSION = '.srt'
 SPEC_EXTENSION = '.json'
 ESSENTIA_ONSETS_AUDIO_EXTENSION = '.wav'
 
 SEGMENTS_PATH_BASE = OUTPUT_PATH_BASE + 'segments/'
 RS_PATH_BASE = OUTPUT_PATH_BASE + 'rejected_segments/'
-RS_PATH_REPEAT = RS_PATH_BASE + RS_TYPE_REPEAT +'/'
+RS_PATH_REPEAT = RS_PATH_BASE + RS_TYPE_REPEAT + '/'
 RS_PATH_SCENE_CHANGE = RS_PATH_BASE + RS_TYPE_SCENE_CHANGE + '/'
 RS_PATH_TEXT_DETECTED = RS_PATH_BASE + RS_TYPE_TEXT_DETECTED + '/'
-RS_PATH_SOLID_COLOR = RS_PATH_BASE + RS_TYPE_SOLID_COLOR +'/'
+RS_PATH_SOLID_COLOR = RS_PATH_BASE + RS_TYPE_SOLID_COLOR + '/'

mugen/utility.py

@@ -6,7 +6,6 @@
 import Tkinter as tk
 import tkFileDialog
 import subprocess as sp
-from fractions import Fraction
 from collections import OrderedDict
 
 # Project modules
@@ -21,7 +20,7 @@ def get_music_video_name(output_name, is_regenerated):
         count = 0
         while True:
             music_video_name = 'regenerated_' if is_regenerated else ""
-            music_video_name += s.OUTPUT_NAME_DEFAULT + "_%s" % count
+            music_video_name += s.MUSIC_VIDEO_NAME_DEFAULT + "_%s" % count
             if not os.path.exists(get_output_path(music_video_name)):
                 break
             else:
@@ -107,10 +106,10 @@ def get_files(file_type, *sources):
             files.extend([file.encode('utf-8') for file in source])
 
         if len(files) == 0:
-            print("No {} files were selected.".format(file_type, source))
+            print("No {} files were selected.".format(file_type))
             sys.exit(1)
 
-    logging.debug("{}_source {}".format(file_type, source))
+    logging.debug("{}_source {}".format(file_type, sources))
     for file in files:
         logging.debug("{}_file: {}".format(file_type, file))
     return files
@@ -166,7 +165,7 @@ def get_segments_dir(music_video_name):
     return s.SEGMENTS_PATH_BASE + music_video_name + '/'
 
 def get_output_path(music_video_name):
-    return s.OUTPUT_PATH_BASE + music_video_name + s.OUTPUT_EXTENSION
+    return s.OUTPUT_PATH_BASE + music_video_name + s.VIDEO_OUTPUT_EXTENSION
 
 def get_spec_path(music_video_name):
     return s.OUTPUT_PATH_BASE + music_video_name + '_spec' + s.SPEC_EXTENSION
@@ -175,7 +174,7 @@ def get_audio_preview_path(audio_file):
     return s.OUTPUT_PATH_BASE + filename_from_path(audio_file) + "_marked_audio_preview" + s.ESSENTIA_ONSETS_AUDIO_EXTENSION
 
 def get_temp_output_path(music_video_name):
-    return s.TEMP_PATH_BASE + 'temp_' + music_video_name + s.OUTPUT_EXTENSION
+    return s.TEMP_PATH_BASE + 'temp_' + music_video_name + s.VIDEO_OUTPUT_EXTENSION
 
 def get_temp_subtitle_path(music_video_name, track_type):
     return s.TEMP_PATH_BASE + music_video_name + '_' + track_type + '_subs' + s.SUBTITLES_EXTENSION
@@ -184,7 +183,7 @@ def get_temp_audio_onsets_path(audio_file):
     return s.TEMP_PATH_BASE + filename_from_path(audio_file) + '_marked_audio' + s.ESSENTIA_ONSETS_AUDIO_EXTENSION
 
 def get_temp_audio_offset_path(audio_file):
-    return s.TEMP_PATH_BASE + music_video_name + '_offset_audio' + os.path.splitext(audio_file)[1]
+    return s.TEMP_PATH_BASE + filename_from_path(audio_file) + '_offset_audio' + os.path.splitext(audio_file)[1]
 
 def filename_from_path(path):
     """

mugen/video/io.py

@@ -20,7 +20,7 @@ def save_video_segments(video_segments):
 
     count = 0
     for video_segment in video_segments:
-        segment_path = segments_dir + "%s" % count + s.OUTPUT_EXTENSION
+        segment_path = segments_dir + "%s" % count + s.VIDEO_OUTPUT_EXTENSION
         video_segment.write_videofile(segment_path, fps=s.MOVIEPY_FPS, codec=s.MOVIEPY_CODEC, 
                                       ffmpeg_params=['-crf', s.music_video_crf])
         count += 1
@@ -44,22 +44,22 @@ def save_rejected_segments(rejected_segments):
         video_segment = rejected_segment['video_segment']
 
         if reject_type == s.RS_TYPE_REPEAT:
-            segment_path = s.RS_PATH_REPEAT + "%s" % rs_repeat_count + s.OUTPUT_EXTENSION
+            segment_path = s.RS_PATH_REPEAT + "%s" % rs_repeat_count + s.VIDEO_OUTPUT_EXTENSION
             video_segment.write_videofile(segment_path, fps=s.MOVIEPY_FPS, codec=s.MOVIEPY_CODEC, 
                                           ffmpeg_params=['-crf', s.music_video_crf])
             rs_repeat_count += 1
         elif reject_type == s.RS_TYPE_SCENE_CHANGE:
-            segment_path = s.RS_PATH_SCENE_CHANGE + "%s" % rs_scene_change_count + s.OUTPUT_EXTENSION
+            segment_path = s.RS_PATH_SCENE_CHANGE + "%s" % rs_scene_change_count + s.VIDEO_OUTPUT_EXTENSION
             video_segment.write_videofile(segment_path, fps=s.MOVIEPY_FPS, codec=s.MOVIEPY_CODEC, 
                                           ffmpeg_params=['-crf', s.music_video_crf])
             rs_scene_change_count += 1
         elif reject_type == s.RS_TYPE_TEXT_DETECTED:
-            segment_path =  s.RS_PATH_TEXT_DETECTED + "%s" % rs_text_detected_count + s.OUTPUT_EXTENSION
+            segment_path =  s.RS_PATH_TEXT_DETECTED + "%s" % rs_text_detected_count + s.VIDEO_OUTPUT_EXTENSION
             video_segment.write_videofile(segment_path, fps=s.MOVIEPY_FPS, codec=s.MOVIEPY_CODEC, 
                                           ffmpeg_params=['-crf', s.music_video_crf])
             rs_text_detected_count += 1
         else:
-            segment_path = s.RS_PATH_SOLID_COLOR + "%s" % rs_solid_color_count + s.OUTPUT_EXTENSION
+            segment_path = s.RS_PATH_SOLID_COLOR + "%s" % rs_solid_color_count + s.VIDEO_OUTPUT_EXTENSION
             video_segment.write_videofile(segment_path, fps=s.MOVIEPY_FPS, codec=s.MOVIEPY_CODEC, 
                                           ffmpeg_params=['-crf', s.music_video_crf])
             rs_solid_color_count += 1

mugen/video/sizing.py

@@ -29,7 +29,6 @@ def resize_video_segments(video_segments):
             # Crop top & bottom
             cropped_height = int(width/music_video_aspect_ratio)
             height_difference = height - cropped_height
-            cropped_dimensions = (width, cropped_height)
             video_segment = video_segment.crop(y1 = height_difference/2, y2 = height - height_difference/2)
 
         # Resize video if needed, to match music video dimensions

mugen/video/video.py

@@ -1,6 +1,4 @@
-import sys
 import random
-import logging
 import moviepy.editor as moviepy
 from tqdm import tqdm
 
@@ -22,7 +20,8 @@ def create_music_video(video_segments, audio_file, spec):
     music_video = moviepy.concatenate_videoclips(video_segments, method="compose")
     music_video = music_video.set_audio(audio)
     music_video.write_videofile(temp_output_path, fps=s.MOVIEPY_FPS, codec=s.MOVIEPY_CODEC, 
-                                audio_bitrate=s.MOVIEPY_AUDIO_BITRATE, ffmpeg_params=['-crf', s.music_video_crf])
+                                audio_bitrate=s.MOVIEPY_AUDIO_BITRATE, ffmpeg_params=['-crf', s.music_video_crf],
+                                temp_audiofile=s.TEMP_MOVIEPY_AUDIOFILE)
     v_io.add_auxiliary_tracks(temp_output_path, spec)
 
 def generate_video_segments(video_files, beat_interval_groups):