FreeLCS User Manual

Introduction

Features

Requirements

Debian modifications needed before installation

Ubuntu Server modifications needed before FreeLCS installation

Installation

Installing FreeLCS on a computer without a graphical user interface

FFmpeg / Libav-tools allowed wrapper and codec formats

Files that FreeLCS creates in your system

Usage

Loudness History Graphics

Loudness history graphics when target loudness is set to -16, -23 and -31 LUFS

Machine readable results

MXF audio remixing

Mounting the server HotFolder from OS X

Making an permanent alias to the mount from OS X

Mounting the server HotFolder from windows 10

Making a mount permanent in windows 10

Html Progress Report

Static ip - address for the server

Static ip - address for Ubuntu Server and Debian without a GUI

Optimizing processing speed

Configuring software Raid

How to partition disks bigger than 2 TB on Linux

Notes on loudness correction usage

Setting up automatic security updates

Automatic security updates on Ubuntu Server and Debian

Synchronizing clock to a NTP time server

Cloning a FreeLCS installation to another machine (without a display)

How to report a bug

Run FreeLCS programs so that possible error messages can be seen

FreeLCS development

FAQ

Introduction

Features

• Automatically loudness corrects audio files to conform with EBU R128 loudness recommendation.
• Very easy to use, just drop in your file and copy back the loudness corrected version.
• Creates loudness history graphics that allows the user to see how loudness varies inside a file (example picture below).
• Supports channel counts from mono to 5.1.
• Supports EBU TruePeak measurement.
• Uses a protective limiter to keep peaks in the allowed range where needed.
• Takes advantage of multiple processor cores to run calculations simultaneously.
• User is able to define how many audio files are processed at the same time, limited only by available cpu processing power and hard disk bandwidth.
• Supports writing loudness measurement results to a machine readable file.
• Modest hardware requirements, runs fine with only 4 GB of ram.
• Can be run inside a virtual machine, no need for physical hardware.
• Fast processing. On typical hardware you can expect processing to be 10 - 20 faster than realtime.
• Cleans up automatically by deleting all files after a set time delay.
• Sends possible error messages to the admin by email.
• If user installs FFmpeg or libav-tools then additional features becomes available:
• The wide range of FFmpeg / libav-tools supported formats can be processed.
• Support for multistream files, all audio streams are extracted and processed individually.
• Extract audio from wrappers containing both video and audio.
• Remix audio from MXF files to required mixes before loudness processing.
• All software is free and Open Source. Install in as many computers as you like.
• Runs on the free Linux operating system, no need to pay for proprietary os installation.
• Easy to integrate in many workflows.
• Written in Python 3.
• More specific information can be found in the FAQ.

Requirements

• You need a computer with at least 2 processor cores.
• 4 GB of RAM  (there probably is not much advantage in having more).
• Ubuntu Linux 18.04 LTS, 16.04 LTS, Debian 10 or 9.
• A person that is able to set up and update an Linux server.
• 1 Gbit ethernet is recommended.
•  The FAQ discusses things like supported formats, channel orders, etc.
  

Installation

• Always use a 64 bit Linux distro. FreeLCS uses external helper programs and one of those (mediainfo) in it's 32 bit version is not able to handle big audio files. The 64 bit version of mediainfo works fine with big files.
• A good starting point for a company server running in a virtual machine might be assigning 4 GB of ram and 6 processor cores for the virtual machine running FreeLCS. All cores will only be used when there are 3 files to process simultaneously. When FreeLCS virtual server is idle, the cores are available to other virtual machines running on the same hardware.
• It is recommended to use Ext4 or XFS filesystems for all disks. These are now very stable and much faster than the JFS filesystem I used to recommend earlier.
• During the installation of Ubuntu and Debian, you are prompted to create the first user. It is important to use a password for this account, do not leave the password empty.
• After installation of Ubuntu, log in as the first user you created during installation. (Debian: log in as the non root user you created during installation (the second user you created)).
• Next you must install python3 and idle3 giving a terminal command.
• Press the "windows" key on the keyboard to bring up Ubuntu search function, write: term
  You should now see an icon for "Terminal", start it up by clicking on it.

• Write the following commands in the terminal window (you can copy and paste them from below. Note that that keyboard command for paste in Terminal is: shift + ctrl + v):

sudo apt-get update

sudo apt-get -y install python3 idle3

• Press enter after each line to run the commands.
  

• The first command will ask for your password. Type it in and press enter.
• The first command-line (apt-get update) fetches up to date information about what software package versions are available for installation. If the package information is out of date then FreeLCS installation may fail. The second line installs python3 programming language.
• Now that python3 is installed you can start the FreeLCS installation program. The installer will ask you some details about how you would like the loudness correction server to work and installs the rest of the programs FreeLCS needs.
  
• Next unzip the FreeLCS - package by clicking it with the right mouse button and selecting " Extract Here". A new directory will be created and all FreeLCS files will be unzipped there.
  

• Next we need to make the installer program executable. Navigate to the directory that unzip created and right click the "installer.py" program and select "Properties" from the menu.
  

• Click the "Permissions" tab and select "Allow executing file as program" and the click "Close". The installer program can now be run.
  

• Double click the "installer.py" - file. A window will appear, click "Run".
  

If the installer program opens up in a text editor instead of showing you the window above, then the file manager you are using prevents the program from running. This happens at least on Ubuntu 14.04.

• If this happens then you need to change a setting in your file manager. First open up it's preferences.

• Then open up the tab named "Behavior" and activate "Run executable text files when they are opened". Now you can run the installer by double clicking on it.

• The installer program starts. The first screen asks you to accept the license agreement.
• FreeLCS is published under a permissive license that lets you use, modify and distribute the program.
• This license lets you use FreeLCS in your business or home free of any charges. You are allowed to install FreeLCS in as many computers as you want. FreeLCS is free software.

• Click "Next" when you have ticked "I accept the license agreement".
  

• The next window asks for your password. On Ubuntu type in the password for the first user that you created during Ubuntu installation.
• (Debian: type in the password for the non privileged user you created during installation (the second user you created)).
• The installer uses this password to get root - proviledes. Root permissions are needed when FreeLCS program files are installed into the Linux system.

• The next screen will ask some details about how you would like loudness correction to work.
• The first section will allow you to define the path to the target directory. Some directories are automatically created inside this target directory, the most important of these is the 'HotFolder'. HotFolder is the folder where users can drop in files for processing. HotFolder is by default located in: 'Target Directory / LoudnessCorrection'. Hotfolder name depends on your selected target loudness, this if further discussed below.
• The second section allows you to select English or Finnish as the language for directories and text in loudness history graphics files.
• 'Number of processor cores to use' lets you define how many processor cores are used in loudness calculation and file processing. When a file is being processed it always ties up two processor cores. The speed of processing is most often dependent of the raw speed of disks. If your HotFolder is on an fast disk for example a Raid device, then it often speeds up processing if you select here more processor cores than you actually have. For example for our test server that has one Intel Core 2 Duo (dual core) processor, 4 GB of Ram, 3 x 1 TB Sata - disks in Raid 0 for HotFolder, the sweet spot for speed is six cores even though we only have two. See: Optimizing processing speed.
  
• The next section of the window lets you define the time files are allowed to stay on the server before they are automatically deleted. The default value here is 8 hours (480 minutes).
• The last section lets you define the target loudness level. Files processed with FreeLCS will be at this loudness level. EBU R128 standard target loudness is -23 LUFS and the installer defaults to this value. If you tick the box on this section you can change target loudness in range of -31 LUFS to -12 LUFS. Note that using high target loudness levels puts some additional requirements for the source files, these must be mixed with reduced loudness range or you might introduce clipping. This is further discussed here. Hotfolder name will be just "LoudnessCorrection" when using EBU target level -23 LUFS, but if you change it then the target level is added to the hotfolder name, for example "LoudnessCorrection_-16_LUFS". Hotfolder name is part of the mount command and this will help users to see which target level service they are using in case you have one instance processing files with -23 LUFS and another instance processing with -16 LUFS.

• If you tick the box on this section you can change target loudness in range of -31 LUFS to -12 LUFS.

• Hotfolder name will be just "LoudnessCorrection" when using EBU target level -23 LUFS, but if you change it then the target level is added to the hotfolder name, for example "LoudnessCorrection_-16_LUFS". Hotfolder name is part of the mount command and this will help users to see which target level service they are using in case you have one instance processing files with -23 LUFS and another instance processing with -16 LUFS.

• The next window lets you select if you want to email possible error messages to the admin.
  
• Input your email details here. Smtp - server is the email server you use to send email out to the world. (Note that if you use Gmail, the settings to use are displayed printed in light gray on the second text line in the window).
  
• You can add max 5 target addresses. When you have filled in email details, click the "Send" - button to test that email sending really works. The bottom of the window displays a message if email sending is succesful.
• IMPORTANT NOTE: Gmail two-factor authentication is not supported. Google calls this feature two-step verification.
  
• HeartBeat Checker is a separate program that monitors the health of the loudness correction program. If Loudness correction crashes for some reason, HeartBeat Checker sends you email. If your loudness correction server is mission critical then enable this option.
• (Note that the password you give here is not encrypted when it is stored in the file "/etc/LoudnessCorrection_Settings.pickle". Scripts can not securely encrypt anything since the scripts themselves can be easily read and any encryption method used can be seen by examining the source code :)
  

• The next window lets you select if you want "Html progress report" written to the HotFolder. It is an html-page that can be viewed with a web-browser and it displays the loudness calculation queue inside the server. On the page you can see how many files are waiting to be processed and names of the last 100 processed files and times they were processed. The "html progress report" is a tool for the admin to see what's going on inside the server, this feature is not needed by normal users. Current time is displayed on the page also and if it is advancing then the server is alive and well :) The default path for the "Html progress report" is "Target Directory / LoudnessCorrection / 00-Calculation_Queue_Information / 00-Calculation_Queue_Information.html".
  
• The "Html progress report" is updated once in every 5 seconds. During heavy file processing the physical harddisk can become so busy that the progress report can't be written to it often enough. When this happens the progress report seems to be frozen. This problem can be solved by creating a ram-disk and writing the "Html progress report" there instead of the physical hard disk. Ram-disks can be written and read very quickly and traffic jams on hard disks don't affect them at all.
• If you wan't the progress report to be written on a ram-disk, then you can select which of the operating systems ram-devices are used to create the ram disk. You need to change the default setting of "/dev/ram1" here only if you know you have created other ram-disks for other purposes and you know that device "/dev/ram1" is already used for something else.
• Then you can select which user account is used when loudness correction and heartbeat checker programs are run. These two programs are always run as an "unprivileged" user. If you only created one user, then this user is used by default.
  
• If you are not sure what to select for options described above, then just leave the defaults as they are. Defaults are ok for most users :)
• The last option on this window lets you select which peak measurement method you want to use: sample peak or TruePeak. TruePeak is a method that tries to find invisible peaks that might exist between audio samples. These peaks can go above the maximum allowed 0 dBFS and cause distortion when the digital signal is converted back to analog. TruePeak measurement calculates new samples in between existing audio samples and finds these hidden peaks. EBU R128 recommends always using TruePeak method to measure audio peaks. TruePeak calculation is about four time slower than finding the highest existing audio sample in a file.
  
• FreeLCS uses peak limiting in cases where loudness is below -23 LUFS and peaks are so high that distortion would be introduced when file loudness is adjusted up. If highest peak value after gain correction would go over a set limit (-4 dBFS for sample peaks and -2 dBFS for TruePeaks), then a peak limiter is used before loudness correction. The resulting peaks after gain correction will be max. -3 dBFS when using sample peak measurement and -1 dBTP when using TruePeak measurement. Note that if you use sample peak measurement, there will always be 3 dB headroom for the invisible peaks to exist and distortion is very unlikely.
  

• The next window will allow you to define how FreeLCS ouputs its results. Most users will use the defaults meaning FreeLCS creates loudness corrected audio files, history graphics files and deletes multistream files immediately after audiostreams have been extracted from them (multistream extraction requires you to install FFmpeg or libav-tools). This is how FreeLCS versions up to 2.4 worked by default. If you want FreeLCS to work just like it did in versions up to 2.4, then make sure "Use FreeLCS file output defaults" is "Yes" and go to the next screen.
• If you however would like to integrate FreeLCS as a loudness measurement step in your automated file processing workflow, then you might want to change the defaults. FreeLCS 3.0 is able to output loudness measurement results in a machine readable file. Read more about it here.

• The following window allows you to define how you would like audio inside a MXF wrapped file to be recombined to new mixes before loudness processing. Read more about it here. Most users will leave "Enable MXF Audio Remixing" to "No".
• Starting from FreeLCS 3.0 it is possible to restrict FFmpeg / libav-tools to only process a small set of formats or allow it to process all formats it supports. This feature is mainly designed to allow the user to restrict processing only to patent free formats even when FFmpeg or libav-tools is installed. Format patentability varies from country to country, see more about it in FAQ.
• As of 2018 the patents for formats: Dolby AC-3, Mpeg 1 Layers 1, 2 and 3 (mp1, mp2, mp3) have expired. These formats are now patent free and can be used without paying license fees. More of this and links to the current patent situation of these format in FAQ here. These formats are now enabled if you enable free formats in this window. FFmpeg or libav-tools is required to use these formats.
• Settings for allowed FFmpeg / libav-tools formats are located at the bottom of this window. These settings will have no effect if you don't install FFmpeg /  libav-tools.

• In the following picture FFmpeg / libav-tools processing has been restricted to wrapper formats: Wav, Flac, Ogg, Matroska, AC-3, Mpeg 1 Layers 1, 2 and 3, MXF, Webm  and codec formats: PCM, Flac, Vorbis, Opus, AC-3, Mpeg 1 layers 1, 2 and 3.

• The option "Mp4" enables wrapper formats: mp4, m4v, m4a and formats based on mp4: mov, 3gp, 3g2, mj2.

• The next window will let you share the "HotFolder" to the network. Select "Yes" if you want users to be able to drop in files to the HotFolder through the network.
• Samba-server is responsible for creating the network share. The default FreeLCS samba configuration is shown on the window and you can edit it if you like. Only change the configuration if you know your settings will work in your network. In all other cases just leave the configuration as it is and click "Next".
• If you want to disable SMB protocol version 1 (cifs), then remove the character # from the beginning of line: # server min protocol = SMB2

• The next window lists all external programs that FreeLCS needs. Just click the "Install" - button and the installer will download and install everything for you.
• The "Toggle" button can be used to force reinstallation of all programs. You might need this if for example libebur128 has just released an updated version which implements latest changes in EBU R128 recommendation and you want this version right away. The Toggle - button changes the state of all programs to "Not Installed" and clicking it again changes the state back.
  
• The installation takes a couple of minutes and the window freezes while the installer launches external programs to do the actual installation. So don't try to click on anything on the window while the installation is going on. The installer shows messages on the bottom of the screen telling what it is doing at the moment.

• When installation has finished successfully the window looks like the screenshot below. Click the "Next" -  button.
  

• The next window shows how you can install FFmpeg / libav-tools if you need it.

• The last window shows the terminal command to start LoudnessCorrection immediately. The picture below is from a systemd based operating system (Debian 8, Ubuntu 16.04 and later).
• The easiest way to start loudness correction however is to restart the computer.

Files that FreeLCS creates on your system

Installed programs:

• Mediainfo
• Sox
• Gnuplot
• samba
• Libebur128

Files created on a systemd based os (Debian 8, Ubuntu 16.04 and newer):

Explanation of FreeLCS startup files:

• This is the main program of FreeLCS. It responsible for processing all audio files.

• HeartBeat Checker monitors the health of the LoudnessCorrection.py - program and sends email to the admin if LoudnessCorrection.py crashes. This program is started only if you enabled it in the installer.

• The settings that you defined when running the FreeLCS installer are stored in this file. LoudnessCorrection.py and HeartBeat_Checker.py both read their settings from this file when they start up.

• This script is resposible for first setting up things FreeLCS needs and then starting up FreeLCS. When this program starts it first creates a ram disk, then creates directories in the HotFolder and sets their permissions correctly. Then it starts up LoudnessCorrection.py and HeartBeat_Checker.py - programs.

• This file describes what systemd needs to do to start up FreeLCS and what services needs to be up before starting it. Based on this information systemd starts: /etc/systemd/system/loudnesscorrection_init_script

• This is the configuration file for the samba server that shares the FreeLCS HotFolder to the network. Hotfolder is shared to the network only if you selected "Share HotFolder to the network with samba"  =  "yes" in the installer.

Usage

1. First you mount FreeLCS HotFolder through the network.
2. Then you drag some files to the HotFolder .
3. A couple of minutes later you can drag loudness corrected files back from "HotFolder / 00-Corrected_Files" to your computer .

• When you mount the HotFolder your operating system automatically opens a window that shows the contents of the HotFolder. The examples below are from Mac OS X, but the same principles apply to all operating systems.
• The root of HotFolder is where you drag files you want to be loudness corrected.
• Inside the root - folder there are two other folders: "00-Corrected_Files" is where corrected files and loudness calculation history graphics appear. The folder "00-Calculation_Queue_Information" is used to store two files that normally aren't of interest to normal users. The first one is "00-Calculation_Queue_Information.html". That is an html-file that can be opened in a web-browser. The page is updated every 5 seconds and it shows how many files are waiting to be processed, which files are currently worked upon and so on. The second file that is stored in this folder is "00-HeartBeat.pickle". LoudnessCorrection writes periodically timestamps and other information in this file and HeartBeat_Checker.py monitors this file. If information in the file stops updating then LoudnessCorrection has crashed and HeartBeat_Checker sends email to the admin.
  

• For normal users the only folders of interest are the root of the HotFolder and the "HotFolder / 00-Corrected_Files".
• To process a file drag it in the root of the HotFolder.
  

• The file will be copied over the network to the FreeLCS - server.
  

• When the file is copied over navigate to "00-Corrected_Files".
  

• In a while files start to appear there. The first file will be the loudness calculation history graphics file and some time after that the loudness corrected sound file appears. Drag the corrected sound files back to your computer. Each corrected file will have the string  "_-23_LUFS" added at the end of its name.
• All files in the root of the "HotFolder" and  "00-Corrected_Files" will be automatically deleted after a time period that you defined in the installer program. By default it is 8 hours.
  

Loudness History Graphics

• Source file loudness variations and loudness calculation results are shown on the loudness history graphics file.
• Short term loudness variations over time are shown on the picture as a brown line.
• The bright green line represents the target loudness when it is set to EBU R128 standard -23 LUFS. Color of this line is blue if target loudness is set to a value lower that -23 LUFS and red if it is set higher than -23 LUFS.
• The dark green horizontal line shows the integrated loudness of the source file. Integrated loudness averages loudness across the length of your whole file into a single number.

Loudness history graphics when target loudness is set to -16, -23 and -31 LUFS (from top to bottom)

The color for the line for target loudness is:

• red when target loudness is higher than -23 LUFS
• light green when target loudness is EBU R128 standard -23 LUFS (0 LU)
• blue when target loudness is lower than -23 LUFS

Mounting the server HotFolder from OS X

• Note !!!!!!! There are connection problems on OS X 10.7 - 10.10 due to Apple's broken smb - implementation. Read more about it here.
• First click an empty spot on your desktop, this activates the Finder - program.
• The keyboard shortcut to bring up the "Connect to Server" - window is cmd + k, so keep the cmd pressed and press k.
• Now you need to have your servers ip-address. Write "smb://your servers ip address" in the prompt and press "Connect".
  

• The next window lets you choose between logging in anonymously or as a registered user. The network share on the FreeLCS - server is configured to accept only anonymous logins. So don't write in any username or password, just click "Guest".
  

• Now the window changes and looks like the one below. Click "Connect".
  

• The FreeLCS network share is mounted and Finder opens a window showing contents of the HotFolder. Now you can start using the server :)

Making an permanent alias to the HotFolder in OS X

When you have mounted the HotFolder you can make an alias for it that stays on your desktop. You can then skip all the steps above and mount the HotFolder just by double clicking on the alias.

• First you must have mounted the HotFolder.
• Open a Finder - window and scroll to the bottom of the device list on the left panel. You should see an entry there that has the name of your computer, by default it is "admin's Mac". Click on it.
• Now you should see all mounted devices in the Finder window and "LoudnessCorrection" should be among them. Start dragging "LoudnessCorrection" out of the finder window to the desktop and before you let go of the mouse button press and hold cmd and alt - buttons. You should see a curved arrow on top of the LoudnessCorrection - icon. Keep pressing cmd and alt and let go of the mouse button.

• Now on your desktop there should be a "LoudnessCorrection" - icon with a curved arrow on top of it. An icon like this is called an alias.
  
• From now on you can mount the LoudnessCorrection HotFolder just by double clicking on the alias.
  

Mounting the server HotFolder from windows 10

• Click on the "Windows" - button and click on the entrybox on the bottom of the window. If your installation has a target loudness different from standard -23 LUFS then the target loudness is part of the mount path. For example if your FreeLCS installation has a target loudness of -16 LUFS, then the mount path is something like: \\192.168.100.109\LoudnessCorrection_-16_LUFS
  

• Windows mounts the HotFolder on the server and opens it in a file explorer window. Now you can start using the server :)

Making the mount permanent in windows 10

• Open windows file explorer and right - click on "This PC". Select "Map Network Drive" from the menu.

• Write the address of the server in the window. If your installation has a target loudness different from standard -23 LUFS then the target loudness is part of the mount path. For example if your FreeLCS installation has a target loudness of -16 LUFS, then the mount path is something like: \\192.168.100.109\LoudnessCorrection_-16_LUFS. Make sure that the option "Reconnect at sign-in" is tagged, without it windows won't make the mount when windows starts up. Click "Finnish".

Html Progress Report

Static ip - address for the server

• Click on the double arrow on the top right corner of Ubuntu's desktop, select "Edit Connections".

• (Debian: the icon looks a bit different on Debian): 

• Click on tab "Wired" and activate the "Wired connection 1" by clicking on it and click "Edit".
  

• Click the "IPv4 Settings" - tab and then the text "Automatic (DHCP)".
  

• Select "Manual" from the drop down list.
  

• Click "Add" and type in the details of your network connection. You need to know the following information which the network admin will be able to give you.
  
• The static address for the server.
• Network mask.
• The ip-address of the gateway machine. Gateway machine is the machine that is connected to your internal network and internet. Gateway routes traffic from the internal network to internet.
  
• The ip-addresses of your network name servers (DNS).

Optimizing processing speed

1. Delete all files from the HotFolder.
   
2. Run the FreeLCS installer and set the "Number Of Processor Cores To Use" to the number of physical processor cores you have and reboot.
3. Wait for FreeLCS to start up.
4. Open the html-progress report in your web-browser and see that the time displayed on the page is advancing. If not then LoudnessCorrection is not running yet.
   
5. Now you need to make the test files appear in the HotFolder. The quickest way to make this happen is to create hard links for the test files in the HotFolder. This makes the files appear immediately without making physical copies of the files.
   

date  ;  ln   /path_to_test_files/*   /path_to_hotfolder/

1. This command displays the current time which also is the time files were linked to the HotFolder, take note of the time.
2. Wait for processing to finish and take a look at the html-progress report and see the time the last files processing finished.
3. Calculate the time it took.
4. Repeat the process from step 1, but add two to the "Number Of Processor Cores To Use" in the installer.

You should find the fastest processing time in 2 - 4 rounds of the test.

Our test servers sweet spot is to use six cores even though there is only two physical processor cores. Our test server specs are:

1 x Intel Core 2 Duo Processor
2 GB Ram
1 TB hard disk for Ubuntu
3 x 1 TB hard disks in a software raid 0 array for the "HotFolder".
1 Gbit Ethernet

• Use 2 Cores: 20 minutes 11 seconds
• Use 4 Cores: 14 minutes 10 seconds
• Use 6 Cores: 13 minutes 09 seconds

With more than 6 cores the processing time started to increase rapidly.

Configuring software Raid

• Install 3 Sata disks in the computer. Ubuntu gave our disks device-names: sdb, sdc , sdd.
  
• Get root-permission with the command: sudo su
  
• Partition each disk as type: Primary,   fd (Linux raid autodetect). If your disks are bigger than 2 TB then see here how to partition them.
• Install Linux raid software: apt-get install mdadm
• Create a raid 0 device from the three disks: mdadm --verbose --create /dev/md0 --level=0 --raid-devices=3 /dev/sdb1 /dev/sdc1 /dev/sdd1       (Warning: don't give this command on an existing raid, this will wipe out the raid partition table clearing out all data on the disks)
  
• If mdadm complains that: "Device Or Resource Busy", then mdadm may already be using some of the disks and you need to tell it not to. The command: cat /proc/mdstat  lists currently existing raid-devices, stop them with the command (device name is md127 in the example, check the real device name from the output of: cat /proc/mdstat): mdadm --stop /dev/md127    now try creating the raid-device again with the command on the previous line.
• When raid-device creation succeeds without error messages, format the raid device with Ext4 or XFS filesystem (in this example we use ext4): mkfs.ext4 /dev/md0       (Warning: this will wipe out all data on the disks)
  
• Now the raid device exists, but Linux forgets about it in the next boot unless we tell it to configure it again when booting up. The next commands will save the configuration:

cd /etc/mdadm

mv mdadm.conf mdadm.conf.orig

echo "DEVICE partitions" > /etc/mdadm/mdadm.conf

mdadm --detail --scan >> /etc/mdadm/mdadm.conf

• Next make a directory for mounting the raid: mkdir  -p  /mountpoint/raid-disk    and make it readable and writable for everyone: chmod -R 777   /mountpoint
• Get the Universally Unique IDentifier (UUID) for the raid disk with command: blkid /dev/md0        the command will show output like below:

/dev/md0: UUID="229f0670-c8c2-4145-b2f4-b1100d9cb918" TYPE="ext4"

• Add the device to your /etc/fstab so that it gets automatically mounted next time the computer starts up. It is important to use the UUID instead of the device name (/dev/md0) to identify your raid disk in fstab. When the kernel gets updated then your raid device number may change (for example from: /dev/md0  to  /dev/md127) and if you used the device name + number in fstab then the disk won't mount anymore. The UUID of a device will not change and it can be used to identify the disk instead of its device name. (Your disk gets a randomly selected new UUID every time it is reformatted).
• Add the following line in /etc/fstab (change the UUID to match your disk):
  

UUID=229f0670-c8c2-4145-b2f4-b1100d9cb918    /mountpoint/raid-disk        auto    defaults    0 0

• Now reboot the machine and if the raid-disk was successfully mounted then make the directory structure on the raid-device readable and writable for everyone: chmod -R 777   /mountpoint

How to partition disks bigger than 2 TB on Linux

• If you use hard disks bigger than 2 TB, then the disks must be partitioned using the GPT format. Use parted to create GPT partitions on the disks. The easiest way is to install and use the graphical version of parted called:  gparted
• The commandline procedure to use the text based parted to create a partition spanning the whole disk /dev/sdb and setting the partition type to Raid is:

•     Note: the command "set 1 raid on" is used only if you want this disk to be part of an raid array.
  

Notes on loudness correction usage

• Never process audio channels from the same mix separately from each other. Processing left and right channels of a stereo mix separately results in loudness being 3 dBs too high. All audio channels of a mix must be in the same file or audiostream of a file.
• Silence does not affect loudness calculation. You can cut off silence after loudness correction without affecting the integrated loudness of a file.
  
• It might sometimes be useful to 'normalize' raw material early in the video editing phase of a project. You might have a project that is made up of several independent parts that need to be loudness corrected separately. An example of this might be a program where a studio host talks and plays cartoons for kids. The studio sections of the program and each cartoon will have different loudness characteristics.  After you have loudness corrected each section separately there might not be any need for audio post production anymore.
  
• Loudness correction does not affect internal dynamics of a file. If you have too wide loudness variations in a file they will be there also after loudness correction. Correction only adjusts the 'main level' of a file. If integrated loudness of a file is -19 LUFS, loudness correction will make all parts of the file 4 dBs quieter and relative loudness variations inside the file will remain the same. This is how EBU R128 is designed to work, you can still have loud and soft parts in your mix, but it is your responsibility to use them wisely :)
  

Setting up automatic security updates

• Making the server boot automatically after an power outage.
• Making the server install all security updates automatically.
• Making the server automatically reboot weekly so that all security updates are applied.
  

1. Press the "windows" key on the keyboard to bring up Ubuntu search function, write: update
2. You should now see an icon for "Update Manager", start it up by clicking on it.

3. Click on "Settings".
   

4. Click on the "Updates" - tab and then on the box "When there are security updates:".

5. Select "Download and install automatically" from the drop down list.

• (Debian: "Automatically Check For Updates" is by default "Never". You probably might want to change this to "Daily", otherwise new updates are never downloaded).

6. Next the OS asks for your password.
   

7. Click on "Close".
   

Now security updates are downloaded and installed automatically. Some of them are applied immediately but some of them (new versions of the Linux kernel) may come into effect only after a reboot.

• Next we need to make the computer restart automatically once a week. A good time to reboot might be for example: every monday morning at 07:00 am.
• On Ubuntu 12.04, 14.04 and Debian 7 automatic tasks are created with Crontab. On Debian 8, Ubuntu 16.04 and newer these tasks are created with systemd. Below there are instructions for both of these methods.

Automatic weekly reboot with Crontab

• Cron can be used to run commands at preset times. The configuration for Cron is called Crontab. It is a simple text file and we add our commands to it using a text editor.
  
• Open a terminal an write in the command: (asks for root password):

sudo crontab -e

Ubuntu will prompt:

no crontab for root - using an empty one

Select an editor.  To change later, run 'select-editor'.
  1. /bin/ed
  2. /bin/nano        <---- easiest
  3. /usr/bin/vim.basic
  4. /usr/bin/vim.tiny

Choose 1-4 [2]:

This means that Ubuntu wants to know which text based text editor you would like to use to edit the crontab - file, press "2". This will select the "nano" - text editor.

Now the text editor opens up in Terminal and shows some text that Ubuntu automatically inserted into the crontab, navigate to the bottom of it and add the following line:

00 07 * * 1 /sbin/reboot

Automatic weekly reboot with systemd

sudo nano /etc/systemd/system/reboot_weekly.service

Add the following content to the file:

[Unit]
Desciption=Reboot Weekly

[Service]
Type=simple

ExecStart=/sbin/reboot

Now create the file "reboot_weekly.timer" that instructs systemd when to run the task (every Monday morning at 07:00). Startup the nano - editor to create the file with the command:

sudo nano /etc/systemd/system/reboot_weekly.timer

Add the following content to the file:

[Unit]
Description=Reboot Weekly

[Timer]
OnCalendar=Mon *-*-* 00:07:00
Unit=reboot_weekly.service

[Install]
WantedBy=timers.target

Then we need to enable the timer with the command:

sudo systemctl enable reboot_weekly.timer  &&  systemctl start reboot_weekly.timer

Synchronizing clock to a NTP time server

Get time with Crontab

Get time with systemd

sudo nano /etc/systemd/system/synchronize_clock.service

Add the following content to the file:

[Unit]
Desciption=Synchronize Clock

[Service]
Type=simple

ExecStart=/usr/sbin/ntpdate 192.168.1.110

Now create the file "synchronize_clock.timer" that instructs systemd when to run the task (every day at 06:25 am). Startup the nano - editor to create the file with the command:

sudo nano /etc/systemd/system/synchronize_clock.timer

Add the following content to the file:

[Unit]
Description=Synchronize Clock

[Timer]
OnCalendar=*-*-* 00:06:25
Unit=synchronize_clock.service

[Install]
WantedBy=timers.target

Then we need to enable the timer with the command:

sudo systemctl enable synchronize_clock.timer  &&  systemctl start synchronize_clock.timer

Cloning a FreeLCS installation to another machine (without a display)

Usage

1. Install FreeLCS with the normal graphical installer and adjust settings the way you want.
2. Then run '00-backup_freelcs_configuration.sh' to make a backup of the installation settings. Files needed for restoration and a restoration script will be created in directory 'freelcs_backup'.
3. Copy the contents of dir 'freelcs_backup' to all machines that you want to have the same configuration and run the restoration script '00-restore_freelcs_configuration.sh' on each.
4. The restoration script will install all programs and scripts that FreeLCS needs to function. After restoration all machines will have identical FreeLCS installation.
5. Reboot the machines or start loudness correction scripts manually (restoration script displays the startup command at the end of the run).
   

Important notes:

• It is very important that you restore FreeLCS onto the same Linux distro and version that it was originally installed on. There are some differences between Linux versions and FreeLCS adjusts the installation according to these differences. FreeLCS might not work correctly if restored onto a wrong distro or version.
• When you first run the FreeLCS installer you define which user account is used to run FreeLCS programs. It is important that you create this same user on all machines where you restore a FreeLCS backup. If the user account defined during FreeLCS installation does not exist then FreeLCS programs fail to run.
• FreeLCS installation requires either a connection to the internet or access to a local Debian / Ubuntu package repository.

How to report a bug

• The operating system version that you use.
• The version of the FreeLCS - package that you use.
  
• The steps that needs to be done in order to make the bug appear.
• If a file causes the bug and you want to send the file to the developer, then store the file in a cloud file storage system for example DropBox's Public - folder and add the download link to the bug report.
• Error messages are good sources for information when hunting a bug, so if you get one please add it to the bug report.
• If a program just stops working, then it has crashed. In most cases the program is not able to show any error message to the user because the program runs in the background. An error message holds valuable information about what went wrong and if you want to help to sort out the cause of the error you can run the program manually in Terminal. When a program is run manually in Terminal the user can see all possible error messages. Click here to see how a program can be manually run in the Terminal.
  

Run FreeLCS programs so that possible error messages can be seen

LoudnessCorrection.py

• Open the Terminal and stop all FreeLCS programs with the command (asks for root password): sudo systemctl stop freelcs
  

Now there are two ways to run LoudnessCorrection.py, the first one discards all settings user defined in the installer and runs with default settings. When you run the program in this mode you need to add the full path to the target directory at the end of the command. The target directory is the directory which has Hotfolder "LoudnessCorrection" inside it. Note that running the program like this the language defaults to english meaning the program expects the HotFolder name to be "LoudnessCorrection".

• Run the program with default settings: /usr/bin/LoudnessCorrection.py /mountpoint/

Note that when running in this mode the program outputs a lot of text, it tells all the time what it is doing.

The second way to run LoudnessCorrection.py is to use the settings that user defined in the installer. The settings are stored in file /etc/Loudness_Correction_Settings.pickle

• Run the program with user defined settings: /usr/bin/LoudnessCorrection.py -configfile /etc/Loudness_Correction_Settings.pickle
  

HeartBeat_Checker.py

• If you have not already done so open the Terminal and stop all FreeLCS programs (asks for root password): sudo systemctl stop freelcs
  
• Run HeartBeat_Checker manually in Terminal: /usr/bin/HeartBeat_Checker.py -configfile /etc/Loudness_Correction_Settings.pickle

Installer.py

• Open Terminal and navigate to the directory where installer.py is located.
• Make installer executable with this command: chmod 755 installer.py
• Run installer with command:  ./installer.py

Installer only displays something if it crashes.

Typical Python error message

FreeLCS development

• Open the Terminal and get root - permissions (asks for root password): sudo su
  
• Install git - software: apt-get -y install git
  
• Navigate to the directory where you want to Download the latest FreeLCS files.
• Download the git-repository with the command: git clone http://github.com/mhartzel/freelcs.git
  
• This will create a new directory 'freelcs' and put all files inside it.