Corrupted PNG in .resize()

[Marcelo-Marinho]

So, I was trying to make a "auto editor", but in process I made a "auto shorts video", but the image of the pngtuber just corrupts when I try to give it a zoom in effect with this: lambda t: 1 + 0.01 * t . Someone can help me?

Expected Behavior

The image should zoom in and rotate a little without problems

Actual Behavior

The png corrupts and becomes transparent where it shouldn't
Video (in portuguese) showing the problem: https://www.reddit.com/r/moviepy/comments/1gmfzd5/corrupted_png_in_resize/ (yes, I posted in reddit)

Steps to Reproduce the Problem

The problem just appear when I put it in my code .resize(lambda t: 1 + 0.01 * t)

Function video_gen() code

def video_gen():
    #def move_img(time):
    #    return('center', 'top')

    speak = (
        AudioFileClip("assets/gen/fala_google_IA.wav")
        .fx(afx.audio_fadein, 1)
    )

    tempo_vid = speak.duration

    vids_possible = os.listdir("./assets/background-vid")
    bg_vid = (
        VideoFileClip("./assets/background-vid/" + str(random.choice(vids_possible)))
        .subclip(0, int(tempo_vid)+1)
        .fx(vfx.fadein, 1)
        .fx(vfx.colorx, 1)
    )

    try:
        music_bg = (
            AudioFileClip("assets/audio/" + bg_music())
            .subclip(0, int(tempo_vid)+1)
            .fx(afx.volumex, 0.25)
            .fx(afx.audio_fadein, 1)
            #.fx(afx.audio_normalize)
        )
    except:
        print("Error in AI Choice, random choice now")
        musics = os.listdir("./assets/audio")
        music_bg = (
            AudioFileClip("assets/audio/" + random.choice(musics))
            .subclip(0, int(tempo_vid)+1)
            .fx(afx.volumex, 0.25)
            .fx(afx.audio_fadein, 1)
            #.fx(afx.audio_normalize)
        )

    happy_img = (
        ImageClip("assets/img/feliz.png", duration=2)
        .resize((640,640))
        .resize(lambda t: 1 + 0.01 * t) # problem here
        .rotate(lambda t: 1 + 0.75 * t)
        .set_position(('center', 1024-620))
    )
    angry_img = (
        ImageClip("assets/img/bravo.png", duration=2)
        .resize((640,640))
        .resize(lambda t: 1 + 0.01 * t) # problem here
        .rotate(lambda t: 1 + 0.75 * t)
        .set_position(('center', 1024-620))
    )
    normal_img = (
        ImageClip("assets/img/serio.png", duration=2)
        .resize((640,640))
        .resize(lambda t: 1 + 0.01 * t) # problem here
        .rotate(lambda t: 1 + 0.75 * t)
        .set_position(('center', 1024-620))
    )
    dont_like_img = (
        ImageClip("assets/img/desgosto.png", duration=2)
        .resize((640,640))
        .resize(lambda t: 1 + 0.01 * t) # problem here
        .rotate(lambda t: 1 + 0.75 * t)
        .set_position(('center', 1024-620))
    )

    c_img_1 = (
        ImageClip("assets/img/img_1.jpg", duration=2)
        .resize((512, 512))
        .set_position(('center', 'top'))
        .fx(vfx.fadein, 1)
    )

    c_img_2 = (
        ImageClip("assets/img/img_2.jpg", duration=2)
        .resize((512, 512))
        .set_position(('center', 'top'))
    )

    c_img_3 = (
        ImageClip("assets/img/img_3.jpg", duration=2)
        .resize((512, 512))
        .set_position(('center', 'top'))
    )

    c_img_4 = (
        ImageClip("assets/img/img_4.jpg", duration=2)
        .resize((512, 512))
        .set_position(('center', 'top'))
    )


    imgs_array = []
    try:
        subs = pysrt.open('./assets/gen/roteiro.srt')
        temp = 0
        for sub in subs:
            text = sub.text.lower()
            start_time = sub.start
            end_time = sub.end

            txt = text.split(" ")
            file = open('./assets/gen/' + 'roteiro.txt', 'r', encoding='utf-8')
            texto_de_fala = replace_special_chars(file.read()).replace(",", ".").replace("!", ".").replace("?", ".").lower()

            a = list(filter(None, texto_de_fala.split(" ")[temp:len(txt) + temp]))

            if not a == text.split(" "):
                for i in a:
                    if (i.lower() in lista_png):
                        print("png find!", i)
                        temp += 1
                        if i == "[alegre]":
                            imgs_array.append(happy_img.set_start(convert_time(str(start_time))).set_duration(random.randint(2, 5)))
                        elif i == "[brava]":
                            imgs_array.append(angry_img.set_start(convert_time(str(start_time))).set_duration(random.randint(2, 5)))
                        elif i == "[desgosto]":
                            imgs_array.append(dont_like_img.set_start(convert_time(str(start_time))).set_duration(random.randint(2, 5)))
                        else:
                            imgs_array.append(normal_img.set_start(convert_time(str(start_time))).set_duration(random.randint(2, 5)))
                        break
                    elif (not (i.lower() in lista_png)) and (i ==  a[len(a)-1]):
                        print("png doesnt find! Last Word: ", i)
                        temp += 1
            temp = temp + len(txt)

            if temp > len(texto_de_fala.split(" ")):
                temp = len(texto_de_fala.split(" ")) - 6

            print(start_time, " : ", end_time)
    except:
        print("Error, generating one image each 5 seconds")
        print(tempo_vid)
        x_imgs = int(tempo_vid) // 5
        for i in range(0, x_imgs):
            ch = random.randint(0, 3)
            if ch == 0:
                imgs_array.append(happy_img.set_start(i * 5).set_duration(random.randint(2, 5)))
            elif ch == 1:
                imgs_array.append(angry_img.set_start(i * 5).set_duration(random.randint(2, 5)))
            else:
                imgs_array.append(normal_img.set_start(i * 5).set_duration(random.randint(2, 5)))

    generator = lambda text: TextClip(text.encode('utf8'), font='./assets/font/ComicHelvetic_Heavy.otf', fontsize=32, color='yellow', stroke_color='black', stroke_width=0.5, align='center')
    subs = SubtitlesClip('./assets/gen/roteiro.srt', generator)

    #print(TextClip.list('font'))
    bg_music()

    complement_imgs = []
    print("time for each image: ", tempo_vid / 2, " | ", tempo_vid // 2)

    print("bg_vid size:", bg_vid.size)
    print("happy_img size:", happy_img.size)
    print("angry_img size:", angry_img.size)
    print("normal_img size:", normal_img.size)
    print("dont_like_img size:", dont_like_img.size)
    print("c_img_1 size:", c_img_1.size)
    print("c_img_2 size:", c_img_2.size)
    print("c_img_3 size:", c_img_3.size)
    print("c_img_4 size:", c_img_4.size)

    print("happy_img shape:", happy_img.get_frame(0).shape)
    print("angry_img shape:", angry_img.get_frame(0).shape)
    print("normal_img shape:", normal_img.get_frame(0).shape)
    print("dont_like_img shape:", dont_like_img.get_frame(0).shape)
    print("c_img_1 shape:", c_img_1.get_frame(0).shape)
    print("c_img_2 shape:", c_img_2.get_frame(0).shape)
    print("c_img_3 shape:", c_img_3.get_frame(0).shape)
    print("c_img_4 shape:", c_img_4.get_frame(0).shape)

    for i in range(0, 4):
        if i == 0:
            complement_imgs.append(c_img_1.set_start(i * (int(tempo_vid) // 4)).set_duration(tempo_vid // 4))
        elif i == 1:
            complement_imgs.append(c_img_2.set_start(i * (int(tempo_vid) // 4)).set_duration(tempo_vid // 4))
        elif i == 2:
            complement_imgs.append(c_img_3.set_start(i * (int(tempo_vid) // 4)).set_duration(tempo_vid // 4))
        else:
            complement_imgs.append(c_img_4.set_start(i * (int(tempo_vid) // 4)).set_duration(tempo_vid // 4))

    array_comp = [bg_vid]
    for x in complement_imgs:
        array_comp.append(x)
    #imgs_final = concatenate_videoclips(imgs_array[::-1], method='compose')
    #array_comp.append(imgs_final)
    for x in imgs_array[::-1]:
        print("duration: ", x.duration, " | start: ", x.start_time)
        array_comp.append(x)
    array_comp.append(subs.set_position(('center', 'center')))
    
    final_video = CompositeVideoClip(array_comp, size=(576,1024)).subclip(0, int(tempo_vid)+1)
    audio_final = CompositeAudioClip([speak, music_bg])
    final_video.audio = audio_final

    print("final_video size: ", final_video.size)
    final_video.write_videofile("output_videos/video_" + tema + ".mp4")

Specifications

• Python Version: 3.12.3
• MoviePy Version: 1.0.3
• Platform Name: Windows 10
• Platform Version: 22H2