13 WTHR local Station web-dl issue

[doctorm]

I've live download from this network before. I can't remember if I'm doing anything different so I don't know if this is my fault or if they've changed something.

https://www.wthr.com/watch

I've tried using n_m3u8dl-re and yt-dlp. It seems successful, but it goes wrong when they have a commercial.

The video downloads during the commercials, but the audio track doesn't. So playing it back the audio of the program continues to play during the commercial break, and of course when the show returns the audio is out of sync.

0.m3u8 is the 1080p version of the stream and index.m3u8 has the same result.

Any suggestions?

[LZAA]

n_m3u8dl-re --live-real-time-merge "https://video.tegnaone.com/wthr/live/v1/manifest/f9c1bf9ffd6ac86b6173a7c169ff6e3f4efbd693/WTHR-Production/fca34a67-d7bd-4404-a76a-4982c4c15a74/0.m3u8"

[doctorm]

Edit:
Ah. The issue was that I had "-M format=mp4". It 'fixes' the timing issues, which is what causes the problem. It apparently has to remain .ts(?)

Playing it back and the time on my media player goes nuts.

Is this them, me, and/or something that can be repaired?

[LZAA]

What's the problem?

[doctorm]

The audio goes out of sync in mp4 format.
Kept in .ts format it freezes occasionally on playback and when a commercial comes up the time on my media player can jump from 00:06:30 and to 02:12:00 or 530:00:00 while the commercial is playing without sound. The time codes are really messed up.

[LZAA]

n_m3u8dl-re --live-real-time-merge "https://video.tegnaone.com/wthr/live/v1/manifest/f9c1bf9ffd6ac86b6173a7c169ff6e3f4efbd693/WTHR-Production/fca34a67-d7bd-4404-a76a-4982c4c15a74/0.m3u8"

[doctorm]

Have you tried downloading for 10-15 minutes and recording commercial breaks? This just doesn't come out as having good timing and audio sync once there is a commercial.
The commercials don't even have any audio.

[Asdfgf]

https://gofile.io/d/VcpKNv

The commercials don´t show video, only an static image with audio, but when it starts again the show, the audio is okay. (See it yourself, advance to 02:50 on the video file that I uploaded)
Try the chrome extension called "Stre am Recor der - HLS & m3u8 Video Downloader" (without spaces in its name) to do it.

[doctorm]

https://gofile.io/d/VcpKNv

The commercials don´t show video, only an static image with audio, but when it starts again the show, the audio is okay. (See it yourself, advance to 02:50 on the video file that I uploaded)
Try the chrome extension called "Stre am Recor der - HLS & m3u8 Video Downloader" (without spaces in its name) to do it.

Thanks. Feels like a cheat, but it seems to work well.
Here's an example of what n_m3u8dl-re was doing (and I think yt-dlp was as well). While this has the audio go out of sync, in .ts format the video played back awkward and the time positions were all over the place. Trying to manually fix the timing issue with ffmpeg had the same result.
https://gofile.io/d/Yr1D8T

Edit: Tested this extension multiple times and it works like a charm. I don't love that it downloads the stream to their servers and then you download it from them, but it works. I wish I knew how it did it though.

[LZAA]

1. Download.

n_m3u8dl-re "https://video.tegnaone.com/wthr/live/v1/manifest/f9c1bf9ffd6ac86b6173a7c169ff6e3f4efbd693/WTHR-Production/fca34a67-d7bd-4404-a76a-4982c4c15a74/0.m3u8"

2. Saving.

a) Ctrl+C
b) In the '0____' folder there are 'ts' files with an eight-character name (1323xxxx) and a ten-character name (17796xxxxx).
c) Delete all files with an eight-character name (1323xxxx).
d) In the '0____' folder, start a command prompt and run:
copy /b *ts video.ts

[doctorm]

Is there a windows command that will delete files based on number of characters? I've tried ????????.ts but that doesn't work right.

Edit: Testing some things, I mistyped something and it worked...

copy /b *~?.ts video.ts

I think I get it though.

If you use:

dir /x

it lists 8 character files untouched. 10 character files are named xxxxxx~1, xxxxxx~2, etc.
By using a question mark after the tilde, instead of a number, it uses ALL files of that length.
Pretty narrow use, but perfect in this case.

Edit again: btw, that 0.m3u8 has expired.

[LZAA]

1. Delete:

del /q 13*

2. I used 'xxxxx.0.m3u8' for example only.

[xangetsue]

Firstly, I would like to thank LZAA for getting the solution to this problem

Secondly, with the help of gemini we wrote this python script and I packaged it into .exe for simplicity

It automatas the recording and will do the merging and conversion and cleanup automatically

Here is the script in case you want to edit/add anything

import os
import subprocess
import glob
import shutil

def convert_minutes_to_time_str(minutes):
    """Converts minutes (e.g., 30) into HH:MM:SS format (e.g., 00:30:00)"""
    total_seconds = int(minutes) * 60
    hours = total_seconds // 3600
    minutes = (total_seconds % 3600) // 60
    seconds = total_seconds % 60
    return f"{hours:02d}:{minutes:02d}:{seconds:02d}"

def main():
    # 1. Gather User Inputs
    m3u8_url = input("Enter m3u8 URL: ").strip()
    filename = input("Enter output filename (without extension): ").strip()
    video_format = input("Enter video format (e.g., mkv, mp4): ").strip().replace(".", "")
    duration_mins = input("Enter duration of recording in minutes: ").strip()
    
    # Format the duration for N_m3u8dl-re
    time_limit_str = convert_minutes_to_time_str(duration_mins)
    
    # 2. Run N_m3u8dl-re recording command
    print("\n--- Starting Recording ---")
    record_cmd = [
        "N_m3u8dl-re", 
        m3u8_url, 
        "--save-name", filename, 
        "--live-wait-time", "5", 
        "--live-record-limit", time_limit_str
    ]
    subprocess.run(record_cmd)
    
    # 3. Navigate into the output directory created by N_m3u8dl-re
    target_dir = os.path.abspath(filename)
    if not os.path.exists(target_dir):
        print(f"Error: Expected directory '{target_dir}' was not found.")
        return

    # Find the subfolder inside
    subfolders = [d for d in glob.glob(os.path.join(target_dir, "*")) if os.path.isdir(d)]
    if not subfolders:
        print("Error: No subdirectory found inside the recording folder.")
        return
    
    working_dir = subfolders[0]
    print(f"Working inside segment folder: {working_dir}")
    
    # 4. Filter and delete files (8-digit OR less than 2MB)
    print("Filtering segments (removing ads & short chunks)...")
    
    # 2 MB in bytes = 2 * 1024 * 1024
    TWO_MB_IN_BYTES = 2097152 
    
    for file_path in glob.glob(os.path.join(working_dir, "*.*.ts")) + glob.glob(os.path.join(working_dir, "*.ts")):
        # Ensure we don't accidentally target a file that doesn't exist anymore
        if not os.path.exists(file_path):
            continue
            
        base_name = os.path.splitext(os.path.basename(file_path))[0]
        
        try:
            file_size = os.path.getsize(file_path)
            
            # Condition 1: It's an 8-digit filename
            is_8_digits = base_name.isdigit() and len(base_name) == 8
            
            # Condition 2: File size is strictly less than 2 MB
            is_under_2mb = file_size < TWO_MB_IN_BYTES
            
            if is_8_digits or is_under_2mb:
                # Print a log so you know exactly what's being tossed out
                reason = "8-digit name" if is_8_digits else f"under 2MB ({file_size / (1024*1024):.2f} MB)"
                print(f"Deleting ad/garbage segment: {os.path.basename(file_path)} -> Reason: {reason}")
                os.remove(file_path)
                
        except Exception as e:
            print(f"Could not process/delete {os.path.basename(file_path)}: {e}")

    # Change current working directory to execute copy /b smoothly
    original_cwd = os.getcwd()
    os.chdir(working_dir)

    # Double check if there are any .ts files left to merge
    remaining_ts = glob.glob("*.ts")
    if not remaining_ts:
        print("Error: No .ts files left after filtering! Adjust your size threshold.")
        os.chdir(original_cwd)
        return

    # 5. Merge the remaining valid .ts files
    print("Merging remaining segments...")
    subprocess.run("copy /b *ts video.ts", shell=True)
    
    # 6. Convert to user's desired format using FFmpeg
    output_video_name = f"{filename}.{video_format}"
    print(f"Converting to {output_video_name}...")
    ffmpeg_cmd = ["ffmpeg", "-i", "video.ts", "-c", "copy", output_video_name]
    subprocess.run(ffmpeg_cmd)
    
    # 7. Cleanup and Move final file back to root
    os.makedirs(os.path.join(original_cwd, "recordings"), exist_ok=True)
    final_output_path = os.path.join(original_cwd, "recordings", output_video_name)
    
    if os.path.exists(output_video_name):
        shutil.move(output_video_name, final_output_path)
        print(f"Success! Saved final video to: {final_output_path}")
    else:
        print("Error: Final video conversion failed.")

    # Switch back to main script folder and wipe temp files
    os.chdir(original_cwd)
    print("Cleaning up temporary recording folders...")
    try:
        shutil.rmtree(target_dir)
    except Exception as e:
        print(f"Warning: Could not completely remove temp folder {target_dir}: {e}")

if __name__ == "__main__":
    while True:
        main()
        cont = input("\nDo you want to record another stream? (y/n): ").strip().lower()
        if cont != 'y':
            print("Exiting the recorder. Goodbye!")
            break

The m3u8 url I used is the index.m3u8

https://video.tegnaone.com/wthr/live/v1/master/f9c1bf9ffd6ac86b6173a7c169ff6e3f4efbd693/WTHR-Production/live/index.m3u8

Here is a video I tested recording for (40 minutes) but after removing ads its become ~26 minutes.. just in case you want to check the result
https://gofile.io/d/XCG9IY

Finally, here it is
https://files.videohelp.com/u/315187/recorder.exe

Cheers

[doctorm]

Sounds good, xangetsue, but I use windows batch scripts and wrote that last night.

As far as "del /q 13*" goes, I don't know if all their commercials will always be 13*, 8 digit names or what. I didn't want separate based on 13* and 17* which is what I was currently downloading.
For now it should work fine though.

[xangetsue]

Sounds good, xangetsue, but I use windows batch scripts and wrote that last night.

As far as "del /q 13*" goes, I don't know if all their commercials will always be 13*, 8 digit names or what. I didn't want separate based on 13* and 17* which is what I was currently downloading.
For now it should work fine though.

Unfortunately, I noticed a 10-digit .ts file that turned out to be an ad. To fix this, I updated the script to search for filenames with fewer than 10 digits, as well as files under 2 MB, which are highly likely to be ads. It worked perfectly