Adult Swim / Warner

[doctorm]

In the past, ripping Adult Swim videos was simple. Yt-dlp worked fine since there was no encryption.

They've now gone to widevine. While getting a key is easy enough, I'm missing something.
After the first commercial break, the download stops even though it continues in the browser.

The playlists are warner media at

https://vod-media-lln.warnermediacdn.com/adultswim/cenc/

master_wv_de.m3u8 (there is no mpd) contains the different versions. So the 'layer6_wv.m3u8' will do 1080p.

Grabbing that playlist I see the segments are wm_6_0_0.mp4 through wm_6_0_106.mp4.
Then at the commercial break there is:

#EXT-X-DISCONTINUITY
#EXT-X-CUE-OUT:0
#EXT-X-MAP:URI="wm_6_1.mp4"
#EXTINF:6.006,

Then the segments are wm_6_1_*.mp4 instead of wm_6_0_*.mp4.

N_m3u8dl-re doesn't seem to know to resume after the next section of segments. Thoughts?

I don't know what else to provide.

[2nHxWW6GkN1l916N3ayz8HQoi]

You could provide a video URL? I'm asking because for this
https://www.adultswim.com/videos/apollo-gauntlet/eros

I can download it easily with no problems like you mentioned. So a URL would be nice to have

[doctorm]

That video is short and probably doesn't have a commercial break.

https://www.adultswim.com/videos/demon-slayer/someones-dream

Playlist: https://vod-media-aka.warnermediacdn.com/adultswim/cenc/19d98c54340a04225722750ab11016...ame=top-2.15.0

Key=

7883dc804772f614d235809d00000000:f8d5c5c5840e9e888b584a1f3b5fcfde

[aqzs]

That video is short and probably doesn't have a commercial break.
https://www.adultswim.com/videos/demon-slayer/someones-dream

When fetching API by hand (my browser doesn't let me play this one for some reason) I got that video id :

AZEtqNWaSaru9Cdsy7g4

And got this media_id :

85dd75a69f06a2a9846c7a92e11eada896564ca8

Which returned 3 manifest (PR, FP, WV) :

https://vod-media-aka.warnermediacdn.com/adultswim/cenc/19d98c54340a04225722750ab11016f400000000/master_wv_de.m3u8

But I'm getting "08:18:55.616 ERROR: Input not supported" even with headers..

import requests
from pprint import pprint

id = "AZEtqNWaSaru9Cdsy7g4"
print('id: ', id)

params = (('fields', 'title,type,duration,collection_title,poster,stream,title_id,auth,media_id,launch_date,auth_launch_date,disable_ads,season_number,episode_number'),)
response = requests.get(f'https://www.adultswim.com/api/shows/v1/videos/{id}', params=params).json()
media_id = response['data']['video']['media_id']
print('media_id: ', media_id)
season_number = response['data']['video']['season_number']
episode_number = response['data']['video']['episode_number']
title = response['data']['video']['title']

params = (('appId', 'eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJuZXR3b3JrIjoiYWR1bHRzd2ltIiwicHJvZHVjdCI6InR2ZSIsInBsYXRmb3JtIjoid2ViLXRvcDIiLCJhcHBJZCI6ImFkdWx0c3dpbS10dmUtd2ViLXRvcDItZWphdWdvIn0.fjf3vyXbgvxlOQo8RBmic9FQtIUvE_oxVunWFpOiLwQ'),)
response = requests.get(f'https://medium.ngtv.io/v2/media/{media_id}/desktop', params=params).json()
mpd_url = response['media']['desktop']['widevine']['cenc']['url']
print('mpd_url: ', mpd_url)

id:  AZEtqNWaSaru9Cdsy7g4
media_id:  85dd75a69f06a2a9846c7a92e11eada896564ca8
mpd_url:  https://vod-media-aka.warnermediacdn.com/adultswim/cenc/19d98c54340a04225722750ab11016f400000000/master_wv_de.m3u8

appId token doesn't contains sensitive data and is static :

var i = Cn.PlayerConfigBuilder.forSetup().withContainer(e.container).withMetadata(Cr({}, Object.assign({}, {
    env: "production" === window.AS.Environment.getEnv() ? "prod" : "ite",
    companyId: "e1b9076f-e876-436d-ba0a-d0fac24ff03e",
    appId: "eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJuZXR3b3JrIjoiYWR1bHRzd2ltIiwicHJvZHVjdCI6InR2ZSIsInBsYXRmb3JtIjoid2ViLXRvcDIiLCJhcHBJZCI6ImFkdWx0c3dpbS10dmUtd2ViLXRvcDItZWphdWdvIn0.fjf3vyXbgvxlOQo8RBmic9FQtIUvE_oxVunWFpOiLwQ",
    adType: "SSAI"
}

[Frieren]

Manually remove the ads from m3u8 then download with this command

N_m3u8DL-RE --check-segments-count false --base-url https://vod-media-aka.warnermediacdn.com/adultswim/cenc/19d98c54340a04225722750ab11016f400000000/layer6/ layer6_wv.m3u8 --key 7883dc804772f614d235809d00000000:f8d5c5c5840e9e888b584a1f3b5fcfde --save-name hentai

Modified m3u8

https://gofile.io/d/htPMSA

[2nHxWW6GkN1l916N3ayz8HQoi]

Manually remove the ads from m3u8 then download with this command

Downloading the video (only) using your command and edited m3u8 works fine. It has 47:30. But after muxing it to mkv, around ~30 min mark, there's an encrypted section.

Edit: @OP, I don't think there's a straight solution for this. You could take @frieren's solution and just manually cut out the encrypted parts. Those are most likely the remaining ads.

Edit 2: here's a download link to the full uncorrupted video https://www.transfernow.net/dl/20240819qViM8cNv

[larley]

After extensive consultation with Pepe, we were able to come up with this:

The reason the video was not correctly downloaded is because N_m3u8DL-RE gives up an '#EXT-X-DISCONTINUITY' HLS tag.
This tag is used when there are changes in the video fragments (a discontinuity). In this case, the tag only changed the initialization fragment, which would have been fine, if one the init fragments didn't have a single byte difference. This is the reason N_m3u8DL-RE stopped processing the manifest.
So 'forcing' the complete download of the manifest by removing the discontinuity tags would still leave the one part encrypted.

What had to be done is splitting the manifest before and after the discontinuity section, which left three separate manifests that had to be downloaded individually.
I also removed the first fragment of parts 2 and 3 so the text at the start of the video wouldn't repeat itself in the middle. There is still an about one second long black screen between part transitions.

Merging there manifests results in this file:
https://www.swisstransfer.com/d/5c0bb331-238b-4760-8648-fb63a1fe939b

@2nHxWW6GkN1l916N3ayz8HQoi:
Your download seems to be unavailable.

[2nHxWW6GkN1l916N3ayz8HQoi]

Your download seems to be unavailable.

Weird. I just tested it in Tor browser with a random country and works. I guess I should stop using transfernow if it fails

[2nHxWW6GkN1l916N3ayz8HQoi]

Until N_m3u8DL-RE fixes their program, here's a script that does all the steps that were explained here

import re
from os import makedirs, listdir
from os.path import join, exists
from subprocess import run

import requests

MASTER_M3U8 = "https://vod-media-aka.warnermediacdn.com/adultswim/cenc/19d98c54340a04225722750ab11016f400000000/master_wv_de.m3u8?playername=top-2.15.0"
KEY = "7883dc804772f614d235809d00000000:f8d5c5c5840e9e888b584a1f3b5fcfde"
VIDEO_RESOLUTION = ""
AUDIO_BANDWIDTH = ""

TEMP_FOLDER = "temp"
if not exists(TEMP_FOLDER):
    makedirs(TEMP_FOLDER)


def split_m3u8(m3u8_name, m3u8_url):
    response = requests.get(m3u8_url).text
    m3u8_url = m3u8_url.split("/")
    m3u8_url.pop()
    init_m3u8 = response.split("#EXT-X-MAP:")[0]
    end_m3u8 = "#EXT-X-ENDLIST"

    m3u8_maps = response.split('#EXT-X-MAP:')[1:]
    m3u8_maps = [f'#EXT-X-MAP:{m.strip()}' for m in m3u8_maps if m.strip()]
    fragment_name = m3u8_name.split("_")[0] + "_"

    fragment = 0
    commands = []
    for current_map in m3u8_maps:
        fragment += 1

        current_map = current_map.split(end_m3u8)[0]
        current_map = current_map.splitlines()
        last_lines = current_map[-2:]
        has_disco = False
        for line in last_lines:
            for tag in ['#EXT-X-DISCONTINUITY', '#EXT-X-CUE-OUT']:
                if tag in line:
                    has_disco = True
                    break
            if has_disco:
                break

        if has_disco:
            current_map.pop()
            current_map.pop()

        for i in range(0, len(current_map)):
            if "#EXTINF:" in current_map[i]:
                current_map[i + 1] = "/".join(m3u8_url + [current_map[i + 1]])

            elif "#EXT-X-MAP:" in current_map[i]:
                uri = re.findall(r'URI="([^"]+)"', current_map[i])[0]
                new_uri = "/".join(m3u8_url + [uri])
                current_map[i] = current_map[i].replace(uri, new_uri)

        current_map = "\n".join(current_map)
        current_map = "\n".join([init_m3u8, current_map, end_m3u8])
        m_name = f'{m3u8_name}_part_{fragment}.m3u8'

        output_path = join(TEMP_FOLDER, m_name)
        with open(output_path, "w") as f:
            f.write(current_map)

        output_name = m3u8_name.split("_")[0] + "_" + str(fragment)
        commands.append(
            f'N_m3u8DL-RE "{output_path}" '
            f'--key {KEY} -M format=mkv '
            f'--save-name "{output_name}"'
        )

    for c in commands:
        run(c)

    files = listdir(".")
    ffmpeg_txt = ""
    for f in files:
        if f.startswith(fragment_name) and ".MUX." not in f and "_dec" not in fragment_name:
            if "_full" not in f:
                ffmpeg_txt += f"file '{f}'\n"

    fragment_name = fragment_name.split("_")[0]
    ffmpeg_txt_path = "ffmpeg_txt.txt"
    with open(ffmpeg_txt_path, "w") as f:
        f.write(ffmpeg_txt)

    full_name = f'{fragment_name}_full.mkv'
    run(f'ffmpeg -f concat -i "{ffmpeg_txt_path}" -c copy {full_name}')
    return full_name


def process_master_m3u8():
    response = requests.get(MASTER_M3U8).text
    lines = response.splitlines()
    stream_inf_tuples = []

    for i in range(len(lines)):
        line = lines[i].strip()
        if line.startswith("#EXT-X-STREAM-INF"):
            m3u8_path = lines[i + 1].strip()
            stream_inf_tuples.append((line, m3u8_path))

    video_tuples = []
    audio_tuples = []

    for stream_inf, m3u8_path in stream_inf_tuples:
        paths = MASTER_M3U8.split("?")[0].split("/")
        paths.pop()
        paths.extend(m3u8_path.split("/"))
        m3u8_path = "/".join(paths)

        if "RESOLUTION=" in stream_inf:
            resolution = re.findall(r'RESOLUTION=\d+x(\d+)', stream_inf)
            video_tuples.append((int(resolution[0]), m3u8_path))
        elif "BANDWIDTH=" in stream_inf:
            bandwidth = re.findall(r'BANDWIDTH=(\d+)', stream_inf)
            audio_tuples.append((int(bandwidth[0]), m3u8_path))

    video_tuples = sorted(video_tuples, key=lambda v: v[0], reverse=True)
    if VIDEO_RESOLUTION not in [None, ""]:
        video_tuples = list(filter(lambda v: str(v[0]) == VIDEO_RESOLUTION, video_tuples))
    video_tuple = video_tuples[0]

    audio_tuples = sorted(audio_tuples, key=lambda a: a[0], reverse=True)
    if AUDIO_BANDWIDTH not in [None, ""]:
        audio_tuples = list(filter(lambda a: str(a[0]) == AUDIO_BANDWIDTH, audio_tuples))
    audio_tuple = audio_tuples[0]

    fulls = []
    fulls += [split_m3u8(f'video_res_{video_tuple[0]}', video_tuple[1])]
    fulls += [split_m3u8(f'audio_band_{audio_tuple[0]}', audio_tuple[1])]

    mkvmerge = f'mkvmerge {" ".join(fulls)} -o content_full.mkv'
    run(mkvmerge)


if __name__ == '__main__':
    process_master_m3u8()

Run it on a empty folder. You're gonna need N_m3u8, mkvmerge, ffmpeg, mp4decrypt. The final output is "content_full.mkv". Don't forget to delete the temp files. The input is the master m3u8 URL and the decryption key.

[doctorm]

Manually remove the ads from m3u8 then download with this command

N_m3u8DL-RE --check-segments-count false --base-url https://vod-media-aka.warnermediacdn.com/adultswim/cenc/19d98c54340a04225722750ab11016f400000000/layer6/ layer6_wv.m3u8 --key 7883dc804772f614d235809d00000000:f8d5c5c5840e9e888b584a1f3b5fcfde --save-name hentai

Modified m3u8

https://gofile.io/d/htPMSA

Edited a bunch...:
Another problem is the audio is in Layer7, so you'd have to edit both playlists.

But it still doesn't work. The file size is increased, but nothing after the break plays. Remuxing even removes the data after the break.

Tried yt-dlp with ffmpeg as downloader and then mp4decrypt. It seems to download the master playlist audio and video okay (based on size), but doesn't decrypt.

Mind you, yt-dlp w/o ffmpeg errors that 'initialization fragment found after media fragment' and ffmpeg errors A LOT 'Failed to seek for auxiliary info, will only parse senc atoms for encryption info'.

Edit... and so...:
Just to prove this was doable, you need to edit the audio and video m3u8 files and SPLIT them at the commercial breaks. So in this case layer6_wv is split into L6a and L6b. layer7_wv_0 is L7a and L7b.

N_m3u8DL-RE.exe -M format=mp4 --check-segments-count false --key a8099fd53564a923956eca3000000000:e52353c682f180e478712f2237df91aa --base-url "https://vod-media-cf.warnermediacdn.com/adultswim/cenc/b69aa9fd1a102d803a270d13b8f753a300000000/layer6/" l6a.m3u8 --save-name 6a
N_m3u8DL-RE.exe -M format=mp4 --check-segments-count false --key a8099fd53564a923956eca3000000000:e52353c682f180e478712f2237df91aa --base-url "https://vod-media-cf.warnermediacdn.com/adultswim/cenc/b69aa9fd1a102d803a270d13b8f753a300000000/layer6/" l6b.m3u8 --save-name 6b
N_m3u8DL-RE.exe -M format=mp4 --check-segments-count false --key a8099fd53564a923956eca3000000000:e52353c682f180e478712f2237df91aa --base-url "https://vod-media-cf.warnermediacdn.com/adultswim/cenc/b69aa9fd1a102d803a270d13b8f753a300000000/layer7/" l7a.m3u8 --save-name 7a
N_m3u8DL-RE.exe -M format=mp4 --check-segments-count false --key a8099fd53564a923956eca3000000000:e52353c682f180e478712f2237df91aa --base-url "https://vod-media-cf.warnermediacdn.com/adultswim/cenc/b69aa9fd1a102d803a270d13b8f753a300000000/layer7/" l7b.m3u8 --save-name 7b
(echo file '6a.mp4' & echo file '6b.mp4' )>listv.txt
ffmpeg -safe 0 -f concat -i listv.txt -c copy test.mp4
(echo file '7a.mp4' & echo file '7b.mp4' )>lista.txt
ffmpeg -safe 0 -f concat -i lista.txt -c copy test-a.mp4
ffmpeg.exe -i test.mp4 -i test-a.mp4 -c copy -disposition:s:0 default -metadata:s:1 language=eng Complete.mp4

...never bothering with this again. Ugh.

[2nHxWW6GkN1l916N3ayz8HQoi]

I think I found an easy fix for this site that skips over their broken ad manifest. Do you have another video URL where you get ads? The demon slayer one is gone.

[doctorm]

I don't know what info you need. Here's an episode: https://www.adultswim.com/videos/common-side-effects/pilot

N_m3u8DL-RE "https://vod-manifests-fly.warnermediacdn.com/csm/builder/734364842,proxy.1,735184591,729790619,735061713,734609818,734487686,672057476,734778148,proxy.2.m3u8?yo.p.si=csm-e-cewbdause1vprlf-0bd7db27f8a5204a6-45-1741292602151&yo.p.hn=csm-e-cewbdause1vprlf-0bd7db27f8a5204a6.bln1.yospace.com&yo.p.ci=478924639&yo.p.fn=aHR0cHM6Ly92b2QtbWVkaWEtZmx5Lndhcm5lcm1lZGlhY2RuLmNvbS9hZHVsdHN3aW0vY2VuYy83ZDFiODM3NWNjZWU1NTJhYWJiYTJjNzhlYWU2MGVhZDAwMDAwMDAwL21hc3Rlcl9tZC5tM3U4&yo.p.bp=MANIFEST&yo.p.cu=Q3dRQg==&_fw_ae=&user_ids=&nw=48804&yo.me=true&afid=325214640&prof=48804%3Aadultswim_web_vod&yo.oh=Y3NtLWUtd2JkLXZsZi1lYi50bHMxLnlvc3BhY2UuY29t&yo.kf=wv-cenc&yo.up=https%3A%2F%2Fvod-media-fly.warnermediacdn.com&_fw_ar=&yo.av=3&caid=51470adab5b4ff0235de4c378ddefe616d01011e&yo.aas=true&conf_csid=as.com_vod_web&context=478924639&playername=top-2.14.0&vdur=1289.054&vip=184.164.175.47"  --key 2995ba01b09e5cede9700f7700000000:9cd3d6f7ecc0b58a258f4f9ad021e2b1 --use-shaka-packager -M format=mkv

There's about 10 manifests with this to select from.

[2nHxWW6GkN1l916N3ayz8HQoi]

Link was enough. Thanks for the url. Interesting, seems the trick can't be used for all videos. Good to know. For example for this video
https://www.adultswim.com/videos/dragon-ball-z-kai/a-boundary-pushing-brawl-goku-friez...nd-ginyu-again

You can download the video in 1 easy command by skipping over the drm.

yt-dlp "http://amd.cdn.turner.com/adultswim/episodes/us_geo/dragonballzkai/2015/11/dragonballzkai_cc_044/hls/dragonballzkai_cc_044_stream_tv.m3u8"

Or if you're outside the US

yt-dlp "http://amd.cdn.turner.com/adultswim/episodes/us_geo/dragonballzkai/2015/11/dragonballzkai_cc_044/hls/dragonballzkai_cc_044_stream_tv.m3u8" --add-headers "X-Forwarded-For: 143.231.249.138"

But the link you sent can't be used for that.

A custom script for splitting the manifest still seems to be better suited for this site in the end since you can't do it like that for all videos.

[doctorm]

Are you saying you can't get Common Side Effects to rip? I was working on a script for something else and found that rips easy. Just let yt-dlp do all the work:

yt-dlp.exe "https://www.adultswim.com/videos/common-side-effects/pilot" -o "Common_Side_Effects_S01E01.mp4"

Could be because the first few episodes don't require a log in.

Edit: If your provider is supported, you can look up the ap_mso with

yt-dlp.exe --ap-list-mso

and then add

--ap-mso MSO --ap-username USERNAME --ap-password PASSWORD

[doctorm]

Sorry to double post, but found something interesting. With these streams, yt-dlp can download them properly, but can't decrypt if your cable provider isn't supported. N_m3u8dl-RE can decrypt them, but can't download them properly (it cuts off at the first commercial break).

Is it possible to get n_m3u8dl-re.exe to use yt-dlp to perform the download? Or have yt-dlp.exe download but get n_m3u8dl-re to decrypt?

[2nHxWW6GkN1l916N3ayz8HQoi]

yt-dlp is not a suitable program for adultswim drm content. You can see for yourself

yt-dlp -f 1110 "https://vod-media-aka.warnermediacdn.com/adultswim/cenc/19d98c54340a04225722750ab11016f400000000/master_wv_de.m3u8?playername=top-2.15.0"

Will download encrypted 360p for that demon slayer episode. Even though the page is gone, the m3u8 is still on their servers. Then for decrypting you can try

mp4decrypt --key 7883dc804772f614d235809d00000000:f8d5c5c5840e9e888b584a1f3b5fcfde video.mp4 video_decrypted.mp4

or

shaka-packager in=video.mp4,stream=video,output=video_decrypted.mp4 --enable_raw_key_decryption --keys key_id=7883dc804772f614d235809d00000000:key=f8d5c5c5840e9e888b584a1f3b5fcfde

In both cases it will result in an unplayable video. So yt-dlp is useless for this particular case even if it can download the full encrypted video (it's still broken because of issues that were already explained). I suppose you could attempt with a plugin for yt-dlp that has mp4decrypt included. But if base yt-dlp+mp4decrypt/shaka doesn't work, then I doubt a plugin would be more effective.
https://github.com/pratikpatel8982/yt-dlp-mp4decrypt
or
https://github.com/aarubui/yt-dlp-mp4decrypt

If you're curious, you can give it a shot

Edit: in your case it seems it found a drm free stream, that's why it worked. I don't think it's guaranteed that you'll find one for all videos though