pygvideo

pygvideo

PyGVideo, video for Pygame. Using MoviePy video module to read and organize videos.

Description

PyGVideo or PyGameVideo is a Python library, particularly based on the Pygame library, for video playback or editing. You can process or edit videos and play them directly on a Pygame screen. With the MoviePy module or library, you can edit videos such as trimming, cropping, or adding effects available in MoviePy.

PyGVideo can play videos and sync audio playback. The supported formats by PyGVideo are video formats that contain audio, such as MP4, MOV, AVI, MKV, WMV, FLV, and WebM. Although MoviePy supports non-audio formats like GIF, PyGVideo currently does not support these. PyGVideo works only on Python versions >=3.10, Pygame >= 2.5.0, and MoviePy >= 1.0.3. Below is a simple usage examples:

import pygame
import pygvideo

pygame.init()
pygame.mixer.init()

running = True
video = pygvideo.Video('myvideo.mp4')
screen = pygame.display.set_mode((0, 0), pygame.FULLSCREEN)
clock = pygame.time.Clock()

video.set_size(screen.get_size())

video.preplay(-1)

while running:

    for event in pygame.event.get():

        if event.type == pygame.QUIT:
            running = False

        video.handle_event(event)

    video.draw_and_update(screen, (0, 0))

    pygame.display.flip()

    clock.tick(video.get_fps())

pygvideo.quit()
pygame.quit()

Another example:

import pygvideo

with pygvideo.Video('myvideo.mp4') as video:
    # add some effects

    video.preview()

    # if you want to save your edited video:
    # video.clip.write_videofile('output.mp4')

pygvideo.quit()

In fact, the MoviePy module has some fairly complex methods, and I still need to learn more about this module, so this is what I can provide for you so far :)

Installation

Installation is quite simple, you just need to use the pip method with the following command:

pip install pygvideo

Wait for the download process and for MoviePy (automatically) to be downloaded until it is complete.

Alternatively, you can also download this module from GitHub or through PyPi.

Getting started with PyGVideo

Here is a complete documentation explanation regarding PyGVideo:

Class Video

__init__

This functions similarly to VideoFileClip in MoviePy and also includes the necessary properties for Video. In this method, the following parameters are included:

• filename_or_clip: The video location or directly the VideoFileClip, CompositeVideoClip or ImageSequenceClip class. Ensure the video format is compatible and supported by Video.
• target_resolution: The target resolution. Similar to the resize method.
• logger: Logger type, consisting of:
  • 'bar': Displays a logger with a bar. Useful for tracking audio writing or caching.
  • 'verbose': Displays detailed information about events occurring. (I’m not sure if this works).
  • None: No logger is displayed.
• has_mask: Loads the video with alpha or transparency support. Only available for certain video formats such as WebM.
• load_audio_in_prepare: Creates or generates a temporary audio file when the prepare method is called. If set to False, the temporary audio will be loaded earlier. However, it is less recommended if you want to edit the video first before calling prepare.
• cache: When set to True, this automatically stores video frames in the cache or places them in temporary frames. Video will not need to retrieve frames from get_frame in VideoFileClip. This makes the video run more smoothly.

reinit

Reload the video or refresh the video. If for example you have quited or closed the video, you can call reinit to reinitialize it.

copy

Copies an instance of Video. All effect changes on the MoviePy clip will be copied.

get_original_clip

Retrieves the original clip instance.

get_clip

Retrieves the clip instance.

get_filename

Retrieves the video filename path. This will return None if the clip is CompositeVideoClip or ImageSequenceClip.

get_temp_audio

Retrieves the temporary audio filename path.

get_total_cache_frame

Retrieves the total number of frames that have been stored in the cache.

get_original_size

Retrieves the original size of the video clip in raw form, without any clip modifications.

get_clip_size

Retrieves the original size of the video clip.

get_size

Retrieves the current video size.

get_file_size

Retrieves the video file size (uses os.path.getsize to get the file size, so there may be slight differences, meaning this is just an estimated file size).

get_original_width

Retrieves the original width of the video clip in raw form, without any clip modifications.

get_clip_width

Retrieves the original width of the video clip.

get_width

Retrieves the current video width.

get_original_height

Retrieves the original height of the video clip in raw form, without any clip modifications.

get_clip_height

Retrieves the original height of the video clip.

get_height

Retrieves the current video height.

get_loops

Retrieves the number of loops played by the video. The loop count will reset to 0 when prepare is called.

get_pos

Retrieves the current position of the video while it's playing. Returns a floats or a value in milliseconds, or one of the following codes:

• -1: Video is not ready. prepare has not been called.
• -2: Video has not started playing. play has not been called.

get_alpha

Retrieves the alpha or transparency of the video.

get_duration

Retrieves the duration of the video.

get_start

Retrieves the start time of the video.

get_end

Retrieves the end time of the video.

get_total_frame

Retrieves the total number of video frames. Used this code: int(clip.duration * clip.fps).

get_fps

Retrieves the frames per second (fps) of the video.

get_volume

Retrieves the volume of the video.

get_frame_index

Retrieves the current frame index (while the video is playing).

get_frame

Retrieves a frame at a specific time index. The parameters are as follows:

• index_time: The time index of the frame. If you want to get the frame using a regular index, use the code x * (1 / video.get_fps()) or x * (1 / video.clip.fps).
• get_original: To retrieve the raw frame from the clip or not.

get_frame_array

Similar to the get_frame method but returns the frame as an array using numpy.

iter_chunk_cache_frame

Loads the cache in the form of a generator function, allowing you to directly retrieve the frame surface and the ongoing index. This is suitable for debugging or as part of your project. Here's how to use it:

1. First, create an instance of the generator:

func = video.iter_chunk_cache_frame()

1. Create a loop for the generator:

for frame_surf, index, ran in func:
    # your code

The generator returns yield values as follows:

• frame_surf: The cached frame that has been saved. If the frame is completely black or blank, it means there was an error when retrieving the video frame.
• index: The current cache index. (It will return a value of -1, indicating that caching is complete).
• ran: The total cache range at that moment, or you can get this through get_total_frame.

The generator also captures messages from the generator's send method, which, when called, will stop the generator process and display a message on the console:

Video - Done with the generator stopped. Reason: {MESSAGE FROM SEND PARAMETER}

When you call the send function, you should also close it with stop to properly terminate the generator. Here's an example usage:

func.send('Memory is full.')
func.close()

This will display a message on the console:

Video - Done with the generator stopped. Reason: Memory is full.

is_cache_full

Indicates whether the cache memory is full.

is_ready

Indicates whether the video is ready or prepare has been called and is ready to play.

is_pause

Indicates whether the video is paused.

is_play

Indicates whether the video is currently playing.

is_mute

Indicates whether the video is muted.

is_quit

Indicates whether the video has exited, or quit / close has been called.

is_close

Same as the is_quit method.

draw_and_update

Updates the video while simultaneously drawing the displayed frame. This method returns the current frame surface. Here's an example usage:

frame = video.draw_and_update(SCREEN, (0, 0))

This method has several parameters:

• screen_surface: The surface on which the frame will be drawn. This is optional.
• pos: The position where the frame will be drawn.

If you need to modify the video frame before it is finally drawn to the main surface, you can simply omit the parameters and store the return value of this method as follows:

frame = video.draw_and_update()
# do something with the frame surface

This method is called when the video is ready and playing.

FYI, the frame obtained is not a raw frame.

preview

Displays a preview of the video. Equivalent to the code: video.clip.preview(*args, **kwargs).

You can change the type of preview function in the _type_ parameter with 2 categories, namely:

• clip: from clip.preview.
• ipython-display: from clip.ipython_display
• video-preview: from video_preview.

prepare

Prepares the video and audio. This method loads the temporary audio __temp__.mp3 / __temp_X__.mp3 and then loads the audio into pygame.mixer.music. It also checks whether other Video class instances are active/ready, and if not, raises a pygame.error. exception. This method is called after all video editing or configuration is completed so that it only needs to be played with play.

release

Releases temporary audio resources, allowing other Video class instances to call prepare again.

play

Plays the video and audio. It has the following parameters:

• loops: Determines how many times the video will repeat. If set to -1 or a negative number, it will loop indefinitely.
• start: Specifies the starting point for playback.

This method cannot be called before prepare is called because the audio must be ready.

preplay

prepare and play video audio at once. The rest of the parameters are the parameters of play.

stop

Stops the video and audio.

restop

Stops the video and released.

pause

Pauses the video and audio. The difference from the stop method is that you can still call the draw_and_update method, and the video won't reset to 0 when you call unpause.

unpause

Unpauses the video and audio.

toggle_pause

Pauses and unpauses video.

mute

Mutes the audio. If you call the get_volume method, it will return 0.

unmute

Unmutes the audio.

toggle_mute

Mutes and unmutes video.

jump

Skips the video to a specific ratio between 0 and 1. The ratio parameter determines the video’s skip position. For example, if you want to go to the middle of the video, you can set the parameter as ratio=0.5 or ratio=1/2.

In addition to calling this method, you can also use the xor operator with the ^. syntax. For example:

# regular call
video.jump(0.5)
# calling with xor
video ^ 0.5

next

Skips the video forward by a specified time interval. The distance parameter determines the time in seconds to skip the video.

This method can also be called using the rshift operator with the >>. syntax. For example:

# regular call
video.next(5)
# calling with rshift
video >> 5

previous

This method is almost the same as the next method, except that it rewinds the video instead of skipping forward.

Similar to the next method, you can call this with the lshift operator using the << syntax. For example:

# regular call
video.previous(5)
# calling with lshift
video << 5

seek

This method combines the functionality of previous and next into a single function. To skip forward in the video, pass a positive value to the distance parameter. To rewind the video, pass a negative value.

create_cache_frame

Creates a cache of frames. The difference between this and iter_chunk_cache_frame is that this method is not a generator. You can set the maximum number of frames to cache by passing the max_frame parameter as an integer or None if you want to cache all frames.

clear_cache_frame

Deletes or clears the cache of frames. This method is called when you edit the video with custom_effect or other Video methods.

reset

Resets the video clip's effects back to its original state. You can call this method using the invert operator with the ~ syntax. For example:

# regular call
video.reset()
# calling with invert
~video

custom_effect

Applies or customizes an fx effect from MoviePy or the clip’s methods. There is an important parameter:

• _func_: The fx function or method name as a string.

The remaining parameters are the arguments or keyword arguments for the fx function.

For xample:

# set rotation to 180 degrees with clip.rotate(180)
video.custom_effect('rotate', 180)

# method from fx
import moviepy.video.fx.all as vfx
video.custom_effect(vfx.rotate, 180)

You can directly edit the video using video.clip, but it is strongly discouraged.

invert_colors

Inverts the video’s colors, making them negative.

grayscale

Converts the video to grayscale or black and white.

split

Splits the RGB color channels of a Video instance into three new Video instances (the original Video is not modified).

This method accepts initialization parameters for the three resulting videos simultaneously.

This method also includes console logging.

Note: This method is inspired by cv2.split!

crop

Crops the video using pygame.Rect (version 1.2.0 and above you can use tuple or list type with pygame.Rect parameter content). The rect parameter determines the position and size of the cropped area.

Like before, this method can be called using the modulus operator with the % syntax. For example:

# regular call
video.crop(pygame.Rect(0, 0, 100, 100))
# or
video.crop((0, 0, 100, 100))

# calling with modulus
video % pygame.Rect(0, 0, 100, 100)
# or
video % (0, 0, 100, 100)

rotate

Rotates the video frame. The rotate parameter specifies the degree of rotation.

Like before, this method can be called using the matmul operator with the @ syntax. For example:

# regular call
video.rotate(180)
# calling with matmul
video @ 180

resize

Resizes the video in the clip. The scale_or_size parameter can take different types:

• Integer type for scaling the video size. For example, 0.5 makes the video half its original size.
• List or tuple type for specific dimensions. For example, [100, 100] or (100, 100) resizes the video to 100x100.

Like before, this method can be called using the mul or truediv operators with the * and / syntax. For example:

# regular call
video.resize(2)
# calling with mul
video * 2

# regular call
video.resize(1/2)
# or
video.resize(0.5)
# calling with truediv
video / 2

# both lists and tuples work similarly
# they yield the same output
video * (100, 100)
video / (100, 100)

mirror

Mirrors the video frame. The axis parameter determines the mirror axis: 'x' for horizontal and 'y' for vertical.

As before, you can call this method using the or operator with the | syntax. For

# regular call
video.mirror('x')
# calling with or
video | 'x'

fade

Applies an intro or outro by fading to or from black. The parameters for this method are:

• type: The type of fade, either 'in' for an intro or 'out' for an outro.
• duration: The duration of the fade.

cut

Cuts the video's duration. The parameters for this method are:

• start: The starting point of the cut, in seconds.
• end: The ending point of the cut, in seconds.

reverse

Reverses the video playback. This method has the following parameters:

• step_sub: Specifies the step size for cutting the video if an issue occurs during the reversal process. (vfx.time_mirror issue). The default is 0.01
• max_retries: The maximum number of retries for video slicing.

The method returns either None or one of the following codes:

• -1: The retry limit was exceeded.
• -2: The video duration is less than 0 seconds.
• None: The video was successfully reversed.

concatenate_clip

concatenate the video itself with other videos in 1 video at once. The parameters for this method are:

• clip_or_clips: The clip or clips to be concatenate.

The remaining parameters are the arguments or keyword arguments for the concatenate_videoclips function.

add_volume

Increases the video's volume. The parameters for this method are:

• add: The amount to increase the volume.
• max_volume: The maximum volume increase allowed. The default is 1 (recommended).
• set: Allows volume adjustment even when the audio is muted. (It will not cause an exception, but no volume change will occur when called).

In versions below 1.0.1 you can use the add operator with this + syntax as a method of add_volume. In versions 1.1.0 and above this feature is replaced with concatenate video.

sub_volume

This method is similar to add_volume, but it is used to decrease the video's volume. The changes are as follows:

• The add parameter is renamed to sub.
• The max_volume parameter is renamed to min_volume, and its default is 0 (recommended).
• No operator is provided for this method.

set_alpha

Sets the alpha or transparency for the frame surface. The value parameter defines the alpha level, ranging from 0 (fully transparent) to 255 (fully opaque).

set_size

Adjusts the size of the video frame surface. Unlike the resize method, this one only performs a scaling transformation on the surface. The size parameter specifies the desired video size. Set it to None if you want to reset the size.

set_audio

Replaces or assigns audio to the video. The provided audio must be a valid AudioFileClip or CompositeAudioClip. If the audio duration is shorter than the video's duration, silent audio will be appended to fill the remaining duration.

set_speed

Sets the speed of the video clip. The speed parameter adjusts the video playback speed.

As before, you can call this method using the pow or floordiv operators with the ** and // syntax, respectively. For example:

# regular call
video.set_speed(2)
# calling with pow
video ** 2

# regular call
video.set_speed(1/2)
# or
video.set_speed(0.5)
# calling with floordiv
video // 2

set_fps

Sets the FPS (frames per second) of the video clip. The fps parameter defines the desired FPS. This method may reduce the number of frames in videos with many frames, which can use up a lot of RAM, especially if caching is enabled. It is recommended to set the FPS between 24 and 30 FPS.

set_volume

Sets the volume of the video. The parameters are:

• volume: The value for volume adjustment, ranging from 0 to 1.
• set: Allows volume adjustment even when the audio is muted. (It will not cause an exception, but no volume change will occur when called).

set_pos

Changes the position of the currently playing video in seconds. The pos parameter sets the position in seconds for the video to resume. This will raise an exception if the value exceeds the video duration.

handle_event

Handles events within the event loop in Pygame, serving as the default controller for the video. This method has the following parameters:

• event: The event from the pygame.event.get loop.
• volume_adjustment: The amount to increase or decrease the volume. The default is 0.05.
• seek_adjustment: The amount to skip forward or backward in the video, used with next or previous. The default is 5 seconds.

quit

Exits, cleans up, and frees the video while also deleting the temporary audio file __temp__.mp3 / __temp_X__.mp3. The show_log parameter to determine whether to display error messages or not during the video closing process.

close

This method is identical to quit.

__getitem__

This method is used to retrieve a frame by its index. Similar to the get_frame method, it also returns a frame surface, but the index is a regular index instead of a time-based one. Additionally, this method supports slice indexing. It does not use caching to retrieve the video frame, so it may take some time. Below are some usage examples:

# To get the first frame
first_frame = video[0]

# To get the last frame
last_frame = video[-1]

# To get frames from the 10th to the 100th
scene_frames = video[10:100]

# To get frames from the 10th to the 100th, with every 5th frame
scene_frames = video[10:100:5]

__enter__ and __exit__

These methods are part of the Python context manager syntax, which simplifies resource management. With this, the Video class will automatically close or exit when outside the with block, or if an exception occurs inside it. This helps prevent file or memory leaks and releases the video file properly. Here are some usage examples:

import random

with pygvideo.Video('myvideo.mp4') as video:
    random_index = random.randint(0, video.get_total_frame())
    thumbnail = video[random_index]

__iter__ and __next__

These methods implement the iterator protocol, allowing the Video class to loop using the for keyword and yielding frame surfaces. Here is an example:

for frame in video:
    screen.blit(frame, (0, 0))

Comparison Operators

Several comparison operators such as __lt__, __gt__, __le__, and __ge__ are also available in the Video class. The comparison is not based on object comparison or other criteria (valid on version 1.0.1 and below for __eq__ and __ne__ methods), but rather on the video duration. For example, if you want to compare the duration of an intro and outro video, you can use the following code:

intro = pygvideo.Video('intro.mp4')
outro = pygvideo.Video('outro.mp4')

print(intro == intro) # same like `intro is intro`
print(intro == outro) # same like `intro is outro`

if intro > outro:
    print('Intro is bigger duration than outro')

In addition to comparing the Video instances, you can also compare with other objects such as VideoFileClip from MoviePy or with integers or floats representing milliseconds:

clip = VideoFileClip('someclip.mp4')

if outro < clip:
    print('Outro is small than someclip duration')

__bool__, __list__, __tuple__, __len__, __repr__, __str__, and __copy__

The remaining methods have the following functions:

• __bool__: Returns True if the video is initialized.
• __list__: Returns a list of all video frames. (Not nearly all of them).
• __tuple__: Similar to __list__, but returns a tuple instead.
• __len__: Returns the total number of video frames.
• __repr__: Returns the string repr of the object.
• __str__: Returns brief information about the video.
• __copy__: For copying using the copy method.

Function quit

Exits, cleans up, and releases the video globally. All the videos you have loaded will be released. This function is highly recommended once you no longer need the video or when you exit the Pygame window.

Function close

This function is the same as the quit function.

Environment Variables

These are the environment variables from the os.environ module.

PYGAME_VIDEO_HIDE_SUPPORT_PROMPT

Set this environment variable to hide the support prompt in the console. This must be set before importing PyGVideo.

PYGAME_VIDEO_TEMP

Set this environment variable to specify the directory path where audio or any temporary files are stored. For example, if you have a folder ./temp, set this environment variable to ./temp.

PYGAME_VIDEO_USED

This variable checks whether a video is in use or not. It will have the value '1' when a video is being used and '0' when none are in use. This changes when the methods prepare and release are called. For safety and to avoid exceptions, do not alter this value manually.

Additional Information

What's new in version 1.3.0?

• Bug fixes and documentation.
• Method changes.
• Previewer updates.
• Addition of new methods.
• The clip attribute is now implemented as a property decorator.
• Methods is_cache_full, is_ready, is_pause, is_play, is_mute, is_quit, and is_close are now property decorators, starting from version 1.3.0.

Kredit

• Me (AzzamMuhyala)
• ChatGPT -- LOL