Add a threshold parameter again and copy audio stream (and video/subtitle stream) in case normalization would not be necessary.

For loudnorm, this still requires at least the first pass. For peak and RMS, this only requires the volumedetect filter.

test_mp4.zip

After I added avconv/normalize-audio, I left behind a couple of options:

-l, --level
-m --max

Probably, there is a way to re-implement them!

I just downloaded and installed Python 3.5.1 from the site and got that installed. I got your script using the pip you had listed on the site. That all worked.
Then I tried a test dry run on a wav file I have it less then 30 seconds all told. Here is what I got when I run your script.

C:\Users\micha\Documents\Script sounds>ffmpeg-normalize -n -d -v done.wav 2016-01-29 15:13:03,564 - ffmpeg_normalize - DEBUG - {'--debug': True, '--dry-run': True, '--force': False, '--level': '-26', '--max': False, '--prefix': 'normalized', '--verbose': True, '-m/--max': None, '<input-file>': ['done.wav']} 2016-01-29 15:13:03,565 - ffmpeg_normalize - INFO - reading file done.wav 2016-01-29 15:13:03,565 - ffmpeg_normalize - DEBUG - [command] c:\ffmpeg\bin\ffmpeg -i "done.wav" -filter:a "volumedetect" -vn -sn -f null /dev/null 2016-01-29 15:13:03,655 - ffmpeg_normalize - DEBUG - b"ffmpeg version N-74566-g531b0a3 Copyright (c) 2000-2015 the FFmpeg developers\r\n built with gcc 4.9.3 (GCC)\r\n conf iguration: --enable-gpl --enable-version3 --disable-w32threads --enable-avisynth --enable-bzlib --enable-fontconfig --enable-frei0r --enable-gnutls --enable-iconv --enable-li bass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libdcadec --enable-libfreetype --enable-libgme --enable-libgsm --enable-libilbc --enable-libmodplug --enabl e-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-libopus --enable-librtmp --enable-libschroedinger --enable-libsoxr --enable-l ibspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-aacenc --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enab le-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-lzma --enable-decklink --enable-zlib\r\n libavutil 54. 31.100 / 54. 31.100\r\n libavcodec 56. 58.100 / 56. 58.100\r\n libavformat 56. 40.101 / 56. 40.101\r\n libavdevice 56. 4.100 / 56. 4.100\r\n libavfilter 5. 36.100 / 5. 36.100\r\n libswscale 3. 1.101 / 3. 1.101\r\n libswresample 1. 2.101 / 1. 2.101\r\n libpostproc 53. 3.100 / 53. 3.100\r\nGuessed Channel Layout for Input Stream # 0.0 : stereo\r\nInput #0, wav, from 'done.wav':\r\n Metadata:\r\n encoder : Lavf56.40.101\r\n Duration: 00:00:08.24, bitrate: 705 kb/s\r\n Stream #0:0: Audio: pcm_s16le ([1][0][0][0] / 0x0001), 22050 Hz, 2 channels, s16, 705 kb/s\r\nOutput #0, null, to '/dev/null':\r\n Metadata:\r\n encoder : Lavf56.40.101\r\n Strea m #0:0: Audio: pcm_s16le, 22050 Hz, stereo, s16, 705 kb/s\r\n Metadata:\r\n encoder : Lavc56.58.100 pcm_s16le\r\nStream mapping:\r\n Stream #0:0 -> #0:0 (pcm _s16le (native) -> pcm_s16le (native))\r\nPress [q] to stop, [?] for help\r\nsize=N/A time=00:00:08.23 bitrate=N/A \r\nvideo:0kB audio:710kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown\r\n[Parsed_volumedetect_0 @ 00000000052411c0] n_samples: 363268\r\n[Parsed_volumedetect_0 @ 00000000052411c0] mean_volume: -20.3 dB\r\n[Parsed_volumedetect_0 @ 00000000052411c0] max_volume: -4.9 dB\r\n[Parsed_volumedetect_0 @ 00000000052411c0] histogram_4db: 2\r\n[Parsed_volumedetect_0 @ 00000000052411 c0] histogram_5db: 30\r\n[Parsed_volumedetect_0 @ 00000000052411c0] histogram_6db: 212\r\n[Parsed_volumedetect_0 @ 00000000052411c0] histogram_7db: 472\r\n" Traceback (most recent call last): File "C:\Python35-32\Scripts\ffmpeg-normalize-script.py", line 9, in <module> load_entry_point('ffmpeg-normalize==0.1.3', 'console_scripts', 'ffmpeg-normalize')() File "c:\python35-32\lib\site-packages\ffmpeg_normalize\__main__.py", line 197, in main mean, maximum = ffmpeg_get_mean(input_file) File "c:\python35-32\lib\site-packages\ffmpeg_normalize\__main__.py", line 143, in ffmpeg_get_mean mean_volume_matches = re.findall(r"mean_volume: ([\-\d\.]+) dB", output) File "c:\python35-32\lib\re.py", line 213, in findall return _compile(pattern, flags).findall(string) TypeError: cannot use a string pattern on a bytes-like object

Now I'm new to python but not new to programming, but this debug dump did not help me to try to find what was causing this. Before you ask I'm run windows 10 build 11099.rs1_release.160109-1156 it is a 64 bit build and I have 8GB of RAM. So I hope this helps you in trying to fix the script. I would love to use it as it's something I have need very badly for some time now.
I have added a screen grab of my terminal out put.

I'm not an expert, but shouldn't EBU R128 normalize to -23 dB by default?

Hey,

Your script is working for single file perfectly, but when using ffmpeg-normalize -vuo *.mp4 i get an error 2017-05-12 11:22:09,763 - ffmpeg_normalize - ERROR - file *.mp4 does not exist, will skip.

It's working fine on WSL

I tried ffmpeg-normalize --no-prefix -rMP3 --level=-12 --force *.MP3

to normalize a bunch of files, and just overwrite them after normalizing is done. ffmpeg-normalize says output file is the same as input file, cannot proceed. I guess this cannot work because of the design of ffmpeg itself?

Hi thanks for the software

i use the command to do 1 by 1 but i have all the videos files ina folder over 682 videos and i want to do batch on windows i tried all the command and it doesnt work

can you help me please

ffmpeg-normalize --dir --force --acodec libmp3lame --extra-options "-b:a 320k" -l -10 *.avi
�[1;31mERROR�[1;0m: file *.avi does not exist, will skip

for %%A IN (*.avi) DO ffmpeg-normalize "%%A" --merge --force --acodec libmp3lame --extra-options "-b:a 320k" -l -10

C:\Users***\Documents\Mix 682 Songs>for %%A IN (.avi) DO ffmpeg-normalize --merge --dir --acodec libmp3lame --extra-options "-b:a 320k" -l -10 "%%A"
%%A was unexpected at this time.

C:\Users***\Documents\Mix 682 Songs>for %%A IN (.avi) DO ffmpeg-normalize "%%A" --merge --dir --acodec libmp3lame --extra-options "-b:a 320k" -l -10
%%A was unexpected at this time.

tired all the things and couldnt get it to work please let me know

Would it be possible to have a setting for the temporary file(s) location? The reason I ask is because I use this script on a machine with very little local disk space and I access the files over NFS. I haven't run into an issue yet but I'm only testing with smaller files at the moment.

Thanks for the awesome script, I've been looking for something like this for a while now.
Sorry for opening multiple issues in the last day or so!

The command does not seem to be able to find ffmpeg on Windows (using Cygwin).

1. Please convert all line endings to UNIX format (they are currently in DOS format), otherwise the shebang doesn't work.

2. Line 135: target_level = args.level should be target_level = float(args.level). Otherwise, Python 2.7 complains that it cannot subtract a float from a string on the next line.

3. Please add an option (e.g. --max) to normalize maximum volume instead of mean volume, which would effectively change line 136 from target_level - mean to target_level - maximum. (This is how Audacity normalizes.)

4. Warning on line 74 in ffmpeg_adjust_volume should call print, not print_verbose; I shouldn't have to use -v to see this warning.

Hey,

It's bad practice to use popen with a string instead of a list of arguments, as argument escaping isn't done in that case.

Why is raw=True by default on run_command? This creates all sorts of issues.

You should simply provide a list of arguments, the executable being the first item in that list, to the popen call. That takes care of proper escaping.

That way, this would work:
ffmpeg-normalize --verbose --merge --force --acodec libfdk_aac --ebu --extra-options "-threads 4 -b:a 224k -filter:a compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0"

Instead now I have to single quote the compand filter to not get a wrong call to ffmpeg.

If you do not want that (due to popen replacing os.system perhaps?), you can use this to sanitize the argument list to a string:

import sys
mswindows = (sys.platform == "win32")

if mswindows:
    from subprocess import list2cmdline
    quote_args = list2cmdline
else:
    # POSIX
    from pipes import quote

    def quote_args(seq):
        return ' '.join(quote(arg) for arg in seq)

The current input/output file handling options are just based on a legacy script.

It'd be better if users supplied their output file name, therefore:

ffmpeg-normalize INPUT [INPUT ...] -o OUTPUT [OUTPUT ...]

This would override the prefix if -o is specified; users can simply use a shell loop:

for f in *.wav; do ffmpeg-normalize "$f" -o "dir/foo-$f"; done

Is there a setting to have the normalized file match the same bit rate as the original file?

The space in [folder name]Program Files broke usage.
Had to move ffmpeg to C:

I was doing a bulk conversion of audio files when I noticed some files hadn't been converted. I ran the tool with verbose output and noticed this output:

2017-04-17 01:58:59,787 - ffmpeg_normalize - INFO - file needs -0.1 dB gain to reach -26 dB
2017-04-17 01:58:59,787 - ffmpeg_normalize - INFO - gain = -0.1, will not adjust file

I appreciate that no unnecessary work will be done here, but for bulk scripting it's sometimes useful to make a copy even if no processing needed to happen. Anyway, not a big deal, but could be a nice feature. Thanks!

Hi, I am not reporting a bug; I just have a usage question about this script. First of all, I am confused about what the value-add is over ffmpeg. But maybe there is one because I have tried to use ffmpeg for volume normalization and the results were not very satisfactory.

My application is that I want to use librivox audiobooks as training data for a machine learning model. These audiobooks are recorded under varying conditions with varying quality.

I tried using ffmpeg with the following option: af 'dynaudnorm=c=1:r=1:p=1'. I found that this seemed to do a bad job in the following sense. After running this, the difference between the max sample and min sample in my audio files is still all over the map (ranging from about .5 to almost 2). Do you know why this happens? Anyway, I found that to my model performs far better if, after running the ffmpeg command, I rescale the audio so that the difference between the max and min samples is the same in each 30 second segment. But this seems like an unprincipled hack, and it introduces DC bias. Would your script, with some particular options, be a better solution?

Also, if you happen to know of similar scripts to remove background noise or other artifacts of amateur recordings, I would be glad to know about them.

I run this command ffmpeg-normalize -u -v -b -a aac -e "-b:a 96k" "D:\seens\Meet The Press - 2017-08-06.mp4" and with just -l -27 not -b. I get a file that has aac encoded audio at a bitrate of 96K but when I run the above code I got a video file with aac audio but the it is not the Sampling rate of : 48.0 kHz it changes the Sampling rate not the bitrate.

Meet The Press - 2017-08-06.mp4.txt
This is the base file info before anything is done to it.
normalized by leveling-Meet The Press - 2017-08-06.mp4.txt
This one is done just useing the same code as above just with -l not -b
normalized-Meet The Press - 2017-08-06.mp4.txt
This is the output of the above code.
As you can see the one done with just the -l flag work just fine it's playable on almost anything. But the one done with -b flag masses around with the Sampling rate and that file can only be played from a computer using VLC or something like it. It will not even play any audio on my Samsung Note 3 run android 7.1.1 using VLC or MXPlayer with the FFMpeg codec bundle installed. But the fist two play just fine on my Note my 2 PS3's and all of the computers.
The ebu leveling shouldn't be messing around with the sampling rate at all.
Please have a look into this. If it's me please let me know how I can fix it as i use 96K bitrate audio to help save space on my NAS.
Thanks your friend in coding,
Michael

Hello,

Seems like I can't normalize 24 bit depth files. Utility still converting them to 16 bit files
Can you have a look, pls?

clear && ffmpeg-normalize -d -f -v -a pcm_s24le 101-MA_5.1.flac

2016-08-25 18:36:24,551 - ffmpeg_normalize - DEBUG - {'--acodec': 'pcm_s24le',
 '--debug': True,
 '--dir': False,
 '--dry-run': False,
 '--extra-options': None,
 '--force': True,
 '--level': '-26',
 '--max': False,
 '--merge': False,
 '--prefix': 'normalized',
 '--threshold': '0.5',
 '--verbose': True,
 '<input-file>': ['101-MA_5.1.flac']}
2016-08-25 18:36:24,551 - ffmpeg_normalize - DEBUG - writing result in normalized-101-MA_5.1.wav
2016-08-25 18:36:24,551 - ffmpeg_normalize - INFO - reading file 1 of 1 - 101-MA_5.1.flac
2016-08-25 18:36:24,551 - ffmpeg_normalize - DEBUG - [command] "/usr/local/bin/ffmpeg" -i "101-MA_5.1.flac" -filter:a "volumedetect" -vn -sn -f null /dev/null
2016-08-25 18:36:24,674 - ffmpeg_normalize - DEBUG - ffmpeg version 3.1.2-tessus Copyright (c) 2000-2016 the FFmpeg developers
  built with Apple LLVM version 6.0 (clang-600.0.57) (based on LLVM 3.5svn)
  configuration: --cc=/usr/bin/clang --prefix=/opt/ffmpeg --as=yasm --extra-version=tessus --enable-avisynth --enable-fontconfig --enable-gpl --enable-libass --enable-libbluray --enable-libfreetype --enable-libgsm --enable-libmodplug --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopus --enable-libschroedinger --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libvidstab --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzmq --enable-version3 --disable-ffplay --disable-indev=qtkit --disable-indev=x11grab_xcb
  libavutil      55. 28.100 / 55. 28.100
  libavcodec     57. 48.101 / 57. 48.101
  libavformat    57. 41.100 / 57. 41.100
  libavdevice    57.  0.101 / 57.  0.101
  libavfilter     6. 47.100 /  6. 47.100
  libswscale      4.  1.100 /  4.  1.100
  libswresample   2.  1.100 /  2.  1.100
  libpostproc    54.  0.100 / 54.  0.100
Input #0, flac, from '101-MA_5.1.flac':
  Metadata:
    ENCODER         : Lavf57.41.100
  Duration: 00:00:10.00, start: 0.000000, bitrate: 3263 kb/s
    Stream #0:0: Audio: flac, 48000 Hz, 5.1, s32 (24 bit)
[null @ 0x7f960b000000] Using AVStream.codec to pass codec parameters to muxers is deprecated, use AVStream.codecpar instead.
Output #0, null, to '/dev/null':
  Metadata:
    encoder         : Lavf57.41.100
    Stream #0:0: Audio: pcm_s16le, 48000 Hz, 5.1, s16 (24 bit), 4608 kb/s
    Metadata:
      encoder         : Lavc57.48.101 pcm_s16le
Stream mapping:
  Stream #0:0 -> #0:0 (flac (native) -> pcm_s16le (native))
Press [q] to stop, [?] for help
size=N/A time=00:00:10.00 bitrate=N/A speed= 454x
video:0kB audio:5625kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: unknown
[Parsed_volumedetect_0 @ 0x7f9608401e60] n_samples: 2880000
[Parsed_volumedetect_0 @ 0x7f9608401e60] mean_volume: -37.7 dB
[Parsed_volumedetect_0 @ 0x7f9608401e60] max_volume: -16.2 dB
[Parsed_volumedetect_0 @ 0x7f9608401e60] histogram_16db: 36
[Parsed_volumedetect_0 @ 0x7f9608401e60] histogram_17db: 158
[Parsed_volumedetect_0 @ 0x7f9608401e60] histogram_18db: 473
[Parsed_volumedetect_0 @ 0x7f9608401e60] histogram_19db: 901
[Parsed_volumedetect_0 @ 0x7f9608401e60] histogram_20db: 1640

2016-08-25 18:36:24,675 - ffmpeg_normalize - INFO - mean volume: -37.7
2016-08-25 18:36:24,675 - ffmpeg_normalize - INFO - max volume: -16.2
2016-08-25 18:36:24,675 - ffmpeg_normalize - INFO - file needs 11.7 dB gain to reach -26 dB
2016-08-25 18:36:24,675 - ffmpeg_normalize - DEBUG - [command] /usr/local/bin/ffmpeg -y -i "101-MA_5.1.flac" -vn -sn -filter:a "volume=11.7dB" -c:a pcm_s16le "normalized-101-MA_5.1.wav"
2016-08-25 18:36:24,735 - ffmpeg_normalize - INFO - normalized file written to normalized-101-MA_5.1.wav

No issues with running DEBUG command with 24 bit set directly:

 /usr/local/bin/ffmpeg -y -i "101-MA_5.1.flac" -vn -sn -filter:a "volume=11.7dB" -c:a pcm_s24le "normalized-101-MA_5.1.wav"

ffmpeg version 3.1.2-tessus Copyright (c) 2000-2016 the FFmpeg developers
  built with Apple LLVM version 6.0 (clang-600.0.57) (based on LLVM 3.5svn)
  configuration: --cc=/usr/bin/clang --prefix=/opt/ffmpeg --as=yasm --extra-version=tessus --enable-avisynth --enable-fontconfig --enable-gpl --enable-libass --enable-libbluray --enable-libfreetype --enable-libgsm --enable-libmodplug --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopus --enable-libschroedinger --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libvidstab --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzmq --enable-version3 --disable-ffplay --disable-indev=qtkit --disable-indev=x11grab_xcb
  libavutil      55. 28.100 / 55. 28.100
  libavcodec     57. 48.101 / 57. 48.101
  libavformat    57. 41.100 / 57. 41.100
  libavdevice    57.  0.101 / 57.  0.101
  libavfilter     6. 47.100 /  6. 47.100
  libswscale      4.  1.100 /  4.  1.100
  libswresample   2.  1.100 /  2.  1.100
  libpostproc    54.  0.100 / 54.  0.100
Input #0, flac, from '101-MA_5.1.flac':
  Metadata:
    ENCODER         : Lavf57.41.100
  Duration: 00:00:10.00, start: 0.000000, bitrate: 3263 kb/s
    Stream #0:0: Audio: flac, 48000 Hz, 5.1, s32 (24 bit)
[wav @ 0x7fdeed814000] Using AVStream.codec to pass codec parameters to muxers is deprecated, use AVStream.codecpar instead.
Output #0, wav, to 'normalized-101-MA_5.1.wav':
  Metadata:
    ISFT            : Lavf57.41.100
    Stream #0:0: Audio: pcm_s24le ([1][0][0][0] / 0x0001), 48000 Hz, 5.1, s32 (24 bit), 6912 kb/s
    Metadata:
      encoder         : Lavc57.48.101 pcm_s24le
Stream mapping:
  Stream #0:0 -> #0:0 (flac (native) -> pcm_s24le (native))
Press [q] to stop, [?] for help
size=    8438kB time=00:00:10.00 bitrate=6912.1kbits/s speed=69.8x
video:0kB audio:8438kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.001181%

I am trying to normalize a directory of .mp4 files to 0 db so I dont have to have my receiver cranked up. When it reaches a file that its says is below minimum to normalize it stops. I really want it to go through the entire directory and normalize everything needed. I am using Ubuntu 16.04 on an Odroid to perform this.

Is this normal behavior or something wrong?

It would be awesome if I could use a flag to set the path to the ffmpeg binary instead of modifying my path temporarily on each run.

Hi,

Currently, when ffmpeg-normalize processes a video file with several audio tracks (for example with different language), the code does consider only one track so that the output misses the other audio tracks.

Would it be possible for ffmpeg-normalize to apply the desired normalization to all the audio tracks within a video file?

Thanks

I'm a little confused on how to run it against multiple files at one time, I thought all I needed to do was *.mp4 if I wanted to normalize all mp4 files in a directory but it just does this

Running on linux, if I pass a filename in with multiple consecutive spaces, ffmpeg gets a filename with the block if spaces converted to a single space and it then cannot find the file.
e.g. 'test name' gets passed to ffmpeg as 'test file'

Is there any reason why overwriting the input file with the output file is currently forbidden?

I've used this for an old series which has really bad sound mixing (starts with Doctor) on the remastered BluRay/DVD releases:

ffmpeg-normalize --verbose --merge --force --acodec libfdk_aac --ebu --extra-options "-threads 4 -b:a 224k -filter:a 'compand=0|0:1|1:-90/-900|-70/-70|-30/-9|0/-3:6:0:0:0'" *.mkv

Perhaps this helps someone who doesn't only want to normalize, but to even out peaks and lows, also.

So I had a problem in the beginning with not processing certain videos. It appears you fixed that however, now it produces videos without ANY sound. I am using the below command. So either something is wrong with the command or with the program. Please advise.

ffmpeg-normalize -v -f -nt peak -t 0 -c:a aac -ext mp4 *.mp4

For example, ffmpeg.exe resides in /home/<username>/bin and $PATH is defined in ~/.bashrc as follows:
PATH="~/bin:$PATH"

Results in the following error:
ERROR: Could not find ffmpeg in your $PATH. Please install ffmpeg from http://ffmpeg.org

I'd love if it was possible to provide an easier import functionality for the program, where you could just simply do:

from ffmpeg_normalize import FFmpegNormalize, InputFile

... and then:

ff = FFmpegNormalize()
input_file = InputFile(...)
ff.add_input_file(input_file)
ff.run()

or something.

I am trying to pip install ffmpeg-normalize in Python 3.6.1. This used to work fine in Python 2.7.13 but not in the latest version, it seems.

In an elevated command prompt i type: pip install ffmpeg-normalize

I continue to get Command "python setup.py egg_info" failed with error code 1 for some reason and I am not sure what this means...

    C:\Users\Arete>python --version
    Python 3.6.1
    
    C:\Users\Arete>pip install ffmpeg-normalize
    Collecting ffmpeg-normalize
      Using cached ffmpeg-normalize-0.4.3.tar.gz
        Complete output from command python setup.py egg_info:
        Traceback (most recent call last):
          File "<string>", line 1, in <module>
          File "C:\Users\Arete\AppData\Local\Temp\pip-build-rcxpzvv4\ffmpeg-normalize\setup.py", line 7, in <module>
            readme = readme_file.read()
          File "c:\program files\python36\lib\encodings\cp1252.py", line 23, in decode
            return codecs.charmap_decode(input,self.errors,decoding_table)[0]
        UnicodeDecodeError: 'charmap' codec can't decode byte 0x9d in position 2117: character maps to <undefined>
    
        ----------------------------------------
    Command "python setup.py egg_info" failed with error code 1 in C:\Users\Arete\AppData\Local\Temp\pip-build-rcxpzvv4\ffmpeg-normalize\
    
    C:\Users\Arete>

I am using Windows 10. and I have already tried the accepted answer on a very similar question without any luck...

What is causing the problem here and how can I get ffmpeg-normalize installed?

Just testing on a sample mkv file...

ffmpeg-normalize test.mkv -c:a aac -b:a 192k -f

...getting...
PermissionError: [WinError 32] The process cannot access the file because it is being used by another process: 'C:\\Users\\username\\AppData\\Local\\Temp\\tmpxcqezk6w.mkv' -> 'output.mkv'

-i is not needed and -v should go in its space, not trailing.
proper example would read:
ffmpeg-normalize -v filenamehere

Hello
I am trying to normalize mp3 files (Stereo, 32 bits, 44 MHz).
When I do so with ffmpeg-normalize, the new normalized files are up to 10 times bigger in term of size.
Is there a way to avoid this?
Thank you
Fabrice

I am sorry I am new to all this.

I have successfully installed the script using the following command:

pip install ffmpeg-normalize

I was just wondering how to update the same script.

Like we can do with apt-get update

have tried pip update, but there is no such command.

Thanks for the great script!

As my post says the new ver does not work right on windows any more.
As far as I can see it's because everything now has '' around it now and windows don't understand those both in PS or CL. Windows has to have "" around anything that gets passed to it. Oh just let you know my ffmpeg was just built from git just last night so it's not ffmpeg here it the syntax that is being used.
Thought I would let you know. You might want to think about forking a new build for windows as a lot of the stuff your trying to do may work in a *inux systems but don't work right under windows.

Hi, I have setup a bat with this command:

ffmpeg-normalize "f:\_convert\input.mkv" -o "f:\_convert\normalized.mkv" -c:a aac -b:a 192k -v -d

I can't understand if it's working or not. No errors are thrown, but "normalized.mkv" is never created. The Task Manager says it's using some CPU, so maybe it's doing something.

Is there any way to get some sort of percentage of the running task?

Have ffmpeg in my system path in windows 8.1. Windows command line sees it fine.

Using version 1.05 of your script finds ffmpeg and works fine. Version 1.06/1.07 both complain during install that it can't find it. Version 1.08 installs, but it says that the ffmpeg executable is not a valid file, the path that it lists is the correct path.

I downgraded back to 1.05, which is still working.

Issue

The operation errors out if no prefix is given

Expected output

When no prefix is given, the default "normalized" should be assumed

Steps to reproduce

1. ffmpeg-normalize.exe --merge --prefix --verbose VID_20170119_110840.mp4

Console output

ffmpeg-normalize.exe --merge --prefix --verbose VID_20170119_110840.mp4
2017-01-19 14:08:30,884 - ffmpeg_normalize - ERROR - error running command: "C:\Users\Dheeraj\include_in_path\ffmpeg" -y -i "VID_20170119_110840.mp4" -strict -2 -vcodec copy -af "volume=11.4dB" "--verbose-VID_20170119_110840.mp4"
2017-01-19 14:08:30,884 - ffmpeg_normalize - ERROR - ffmpeg version N-83152-gf7e9275 Copyright (c) 2000-2017 the FFmpeg developers
  built with gcc 5.4.0 (GCC)
  configuration: --enable-gpl --enable-version3 --enable-cuda --enable-cuvid --enable-d3d11va --enable-dxva2 --enable-libmfx --enable-nvenc --enable-avisynth --enable-bzlib --enable-fontconfig --enable-frei0r --enable-gnutls --enable-iconv --enable-libass --enable-libbluray --enable-libbs2b --enable-libcaca --enable-libfreetype --enable-libgme --enable-libgsm --enable-libilbc --enable-libmodplug --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenh264 --enable-libopenjpeg --enable-libopus --enable-librtmp --enable-libsnappy --enable-libsoxr --enable-libspeex --enable-libtheora --enable-libtwolame --enable-libvidstab --enable-libvo-amrwbenc --enable-libvorbis --enable-libvpx --enable-libwavpack --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxavs --enable-libxvid --enable-libzimg --enable-lzma --enable-decklink --enable-zlib
  libavutil      55. 43.100 / 55. 43.100
  libavcodec     57. 74.100 / 57. 74.100
  libavformat    57. 62.100 / 57. 62.100
  libavdevice    57.  2.100 / 57.  2.100
  libavfilter     6. 69.100 /  6. 69.100
  libswscale      4.  3.101 /  4.  3.101
  libswresample   2.  4.100 /  2.  4.100
  libpostproc    54.  2.100 / 54.  2.100
Unrecognized option '-verbose-VID_20170119_110840.mp4'.
Error splitting the argument list: Option not found

I would like to normalize audio contained in video files and I want the process to merge the audio back into the video container and then remove the old (not normalized) file.

This is all I get when I run it

I'm new to the whole Python programming game, but where in the code is the -28 or 26 dB hard coded into the code? I can't find it and been over all of the files on here. I would like to know where that is set because I would like to mod it so i can have it ask for a level and remember that and not have to use a over ride eatch I make a call to the program in my batch files.

As mentioned here: #43 (comment)

Allow still copying the file to handle the case where a bunch of files are converted.

I'm running this command on my xubuntu box. The /tmp and /media/pool (where my target files are) are on separate file systems.

$ ffmpeg-normalize -vuxf "20170927--Why_Farming_Is_Broken_And_Always_Has_Been--UkMZJrbCRdQ.mp4"
INFO: reading file 1 of 1 - 20170927--Why_Farming_Is_Broken_And_Always_Has_Been--UkMZJrbCRdQ.mp4
INFO: mean volume: -18.8
INFO: max volume: 0.0
INFO: file needs -7.199999999999999 dB gain to reach -26.0 dB
WARNING: Merging audio with the original file, but encoder was automatically chosen. Set '--acodec' to overwrite.
Traceback (most recent call last):
  File "/usr/local/bin/ffmpeg-normalize", line 11, in <module>
    load_entry_point('ffmpeg-normalize==0.7.2', 'console_scripts', 'ffmpeg-normalize')()
  File "/usr/local/lib/python3.5/dist-packages/ffmpeg_normalize/__main__.py", line 479, in main
    ffmpeg_normalize.run()
  File "/usr/local/lib/python3.5/dist-packages/ffmpeg_normalize/__main__.py", line 448, in run
    input_file.move_tmp_file()
  File "/usr/local/lib/python3.5/dist-packages/ffmpeg_normalize/__main__.py", line 364, in move_tmp_file
    os.rename(self.output_file, self.input_file)
OSError: [Errno 18] Invalid cross-device link: '/tmp/tmpf2hta6d7.mp4' -> '20170927--Why_Farming_Is_Broken_And_Always_Has_Been--UkMZJrbCRdQ.mp4'

From what I gather, the os.rename method can't be used across different file systems.

Don't know what is going on but this is what I keep getting when I run the script with the "-v" flage.
Traceback (most recent call last):
File "C:\Python27\Scripts\ffmpeg-normalize-script.py", line 9, in
load_entry_point('ffmpeg-normalize==0.2.2', 'console_scripts', 'ffmpeg-normalize')()
File "c:\python27\lib\site-packages\ffmpeg_normalize__main__.py", line 247, in main
target_level = float(args['--level'])
ValueError: could not convert string to float: -v

Could someone provide an example setting the ffmpeg_path variable?

When trying sudo pip install ffmpeg-normalize on OS X with Python 2.7, I get this error message:

Collecting ffmpeg-normalize
  Downloading ffmpeg-normalize-0.1.tar.gz
    Complete output from command python setup.py egg_info:
    Traceback (most recent call last):
      File "<string>", line 20, in <module>
      File "/private/tmp/pip-build-ab9SEV/ffmpeg-normalize/setup.py", line 9, in <module>
        with open('HISTORY.rst') as history_file:
    IOError: [Errno 2] No such file or directory: 'HISTORY.rst'

    ----------------------------------------
Command "python setup.py egg_info" failed with error code 1 in /private/tmp/pip-build-ab9SEV/ffmpeg-normalize

I haven't had this problem before when installing libraries so I'm not sure what to make of it.

I have a folder of music from different albums all at varying volumes. Is it possible to use ffmpeg-normalize to analyze the RMS volume in all files and increase/alter the volume accordingly?

Thanks!

Add a progress bar that shows ffmpeg progress in time. The overall duration can be estimated from the stream parsing command.

Code that could serve as a starting point:

• https://stackoverflow.com/questions/7632589/getting-realtime-output-from-ffmpeg-to-be-used-in-progress-bar-pyqt4-stdout
• https://github.com/althonos/ffpb
• https://searchcode.com/codesearch/view/29976718/

$ ffmpeg-normalize -h
Traceback (most recent call last):
File "c:\python33\lib\runpy.py", line 160, in run_module_as_main
"main", fname, loader, pkg_name)
File "c:\python33\lib\runpy.py", line 73, in run_code
exec(code, run_globals)
File "c:\Python33\Scripts\ffmpeg-normalize.exe__main.py", line 5, in
File "c:\python33\lib\site-packages\ffmpeg_normalize__main__.py", line 96, in
if 'avconv' in FFMPEG_CMD:
TypeError: argument of type 'NoneType' is not iterable

When running your utility with --no-prefix, the scripts generates the normalized file in /tmp and then overrides the old file with the new file. This works well for most of my needs. However, some video files on my system are ~8 GB. My free space on my / partition is around 6 GB. This means I am unable to use the --no-prefix flag on any files which are larger than around 6 GB. Is there any way to get around this limitation using ffmpeg-normalize itself?

I ended up writing a script to rename the normalized files, but it would have been nice not to need to. :-)