Updated README for HLS-16
Updated README for HLS-16

#!/bin/bash #!/bin/bash
# #
# A very simple BASH script to take an input video and split it down into Segments # A very simple BASH script to take an input video and split it down into Segments
# before creating an M3U8 Playlist, allowing the file to be served using HLS # before creating an M3U8 Playlist, allowing the file to be served using HLS
# #
# #
   
###################################################################################### ######################################################################################
# #
# Copyright (c) 2013, Ben Tasker # Copyright (c) 2013, Ben Tasker
# All rights reserved. # All rights reserved.
# #
# Redistribution and use in source and binary forms, with or without modification, # Redistribution and use in source and binary forms, with or without modification,
# are permitted provided that the following conditions are met: # are permitted provided that the following conditions are met:
# #
# Redistributions of source code must retain the above copyright notice, this # Redistributions of source code must retain the above copyright notice, this
# list of conditions and the following disclaimer. # list of conditions and the following disclaimer.
# #
# Redistributions in binary form must reproduce the above copyright notice, this # Redistributions in binary form must reproduce the above copyright notice, this
# list of conditions and the following disclaimer in the documentation and/or # list of conditions and the following disclaimer in the documentation and/or
# other materials provided with the distribution. # other materials provided with the distribution.
# #
# Neither the name of the {organization} nor the names of its # Neither the name of Ben Tasker nor the names of his
# contributors may be used to endorse or promote products derived from # contributors may be used to endorse or promote products derived from
# this software without specific prior written permission. # this software without specific prior written permission.
# #
# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND # THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED # ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
# WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE # WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR # DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR
# ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES # ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
# (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; # (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON # LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
# ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT # ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
# SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. # SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
# #
###################################################################################### ######################################################################################
   
# Basic config # Basic config
OUTPUT_DIRECTORY='./output' OUTPUT_DIRECTORY=${OUTPUT_DIRECTORY:-'./output'}
   
# Change this if you want to specify a path to use a specific version of FFMPeg # Change this if you want to specify a path to use a specific version of FFMPeg
FFMPEG='ffmpeg' FFMPEG=${FFMPEG:-'ffmpeg'}
   
  # Number of threads which will be used for transcoding. With newer FFMPEGs and x264
  # encoders "0" means "optimal". This is normally the number of CPU cores.
  NUMTHREADS=${NUMTHREADS:-"0"}
   
  # Video codec for the output video. Will be used as an value for the -vcodec argument
  VIDEO_CODEC=${VIDEO_CODEC:-"libx264"}
   
  # Video codec for the output video. Will be used as an value for the -acodec argument
  AUDIO_CODEC=${AUDIO_CODEC:-"libfdk_aac"}
   
  # Additional flags for ffmpeg
  FFMPEG_FLAGS=${FFMPEG_FLAGS:-""}
   
  # If the input is a live stream (i.e. linear video) this should be 1
  LIVE_STREAM=${LIVE_STREAM:-0}
   
  # Video bitrates to use in output (comma seperated list if you want to create an adaptive stream.)
  # leave null to use the input bitrate
  OP_BITRATES=${OP_BITRATES:-''}
   
  # Determines whether the processing for adaptive streams should run sequentially or not
  NO_FORK=${NO_FORK:-0}
   
# Lets put our functions here # Lets put our functions here
   
   
## Output the script's CLI Usage ## Output the script's CLI Usage
# #
# #
function print_usage(){ function print_usage(){
   
cat << EOM cat << EOM
HTTP Live Stream Creator HTTP Live Stream Creator
Version 1 Version 1
   
Copyright (C) 2013 B Tasker Copyright (C) 2013 B Tasker, D Atanasov
Released under BSD 3 Clause License Released under BSD 3 Clause License
See LICENSE See LICENSE
   
   
Usage: HLS-Stream-Creator.sh inputfile segmentlength(seconds) [outputdir='./output'] Usage: HLS-Stream-Creator.sh -[lf] [-c segmentcount] -i [inputfile] -s [segmentlength(seconds)] -o [outputdir] -b [bitrates] [-p filename]
   
  -i Input file
  -s Segment length (seconds)
  -o Output directory (default: ./output)
  -l Input is a live stream
  -c Number of segments to include in playlist (live streams only) - 0 is no limit
  -b Output video Bitrates (comma seperated list for adaptive streams)
  -f Foreground encoding only (don't fork the encoding processes into the background - adaptive non-live streams only)
  -p Playlist filename
  -t Segment filename prefix
  -S Segment directory name (default none)
   
  Deprecated Legacy usage:
  HLS-Stream-Creator.sh inputfile segmentlength(seconds) [outputdir='./output']
   
EOM EOM
   
exit exit
   
} }
   
   
## Create the Initial M3U8 file including the requisite headers function createStream(){
# # For VoD and single bitrate streams the variables we need will exist in Global scope.
# Usage: create_m3u8 streamname segmentlength # for live adaptive streams though, that won't be the case, so we need to take them as arguments
# # Some are global though, so we'll leave those as is
function create_m3u8(){  
  playlist_name="$1"
# We'll add some more headers in a later version, basic support is all we need for now output_name="$2"
# The draft says we need CRLF so we'll use SED to ensure that happens bitrate="$3"
cat << EOM | sed 's/$/\r/g' > $OUTPUT_DIRECTORY/$1.m3u8 infile="$4"
#EXTM3U  
#EXT-X-TARGETDURATION:$2 $FFMPEG -i "$infile" \
#EXT-X-MEDIA-SEQUENCE:0 -loglevel error -y \
#EXT-X-VERSION:3 -vcodec "$VIDEO_CODEC" \
  -acodec "$AUDIO_CODEC" \
  -threads "$NUMTHREADS" \
  -map 0 \
  -flags \
  -global_header \
  -f segment \
  -segment_list "$playlist_name" \
  -segment_time "$SEGLENGTH" \
  -segment_format mpeg_ts \
  $bitrate \
  $FFMPEG_ADDITIONAL \
  $FFMPEG_FLAGS \
  "$OUTPUT_DIRECTORY/$output_name"
  }
   
   
  function createVariantPlaylist(){
  playlist_name="$1"
  echo "#EXTM3U" > "$playlist_name"
  }
   
   
  function appendVariantPlaylistentry(){
  playlist_name=$1
  playlist_path=$2
  playlist_bw=$(( $3 * 1000 )) # bits not bytes :)
   
  cat << EOM >> "$playlist_name"
  #EXT-X-STREAM-INF:BANDWIDTH=$playlist_bw
  $playlist_path
EOM EOM
}  
  }
   
## Append a movie segment to the M3U8  
# function awaitCompletion(){
# Usage: append_segment streamname SegmentLength(Seconds) SegmentFilename # Monitor the encoding pids for their completion status
# while [ ${#PIDS[@]} -ne 0 ]; do
function append_segment(){ # Calculate the length of the array
  pid_length=$((${#PIDS[@]} - 1))
cat << EOM | sed 's/$/\r/g' >> $OUTPUT_DIRECTORY/$1.m3u8  
#EXTINF:$2 # Check each PID in the array
$3 for i in `seq 0 $pid_length`
EOM do
} # Test whether the pid is still active
  if ! kill -0 ${PIDS[$i]} 2> /dev/null
  then
## Close the M3U8 file echo "Encoding for bitrate ${BITRATE_PROCESSES[$i]}k completed"
# unset BITRATE_PROCESSES[$i]
# Found that ffplay skips the first few segments if this isn't included. unset PIDS[$i]
# fi
# Usage: close_m3u8 streamname done
# PIDS=("${PIDS[@]}") # remove any nulls
function close_m3u8(){ sleep 1
cat << EOM | sed 's/$/\r/g' >> $OUTPUT_DIRECTORY/$1.m3u8 done
#EXT-X-ENDLIST }
EOM  
}  
  # This is used internally, if the user wants to specify their own flags they should be
  # setting FFMPEG_FLAGS
# The fun begins! Think of this as function main FFMPEG_ADDITIONAL=''
  LIVE_SEGMENT_COUNT=0
  IS_FIFO=0
  TMPDIR=${TMPDIR:-"/tmp"}
  MYPID=$$
# Get the input data # Get the input data
   
# Basic Usage is going to be # This exists to maintain b/c
# cmd.sh inputfile segmentlength LEGACY_ARGS=1
   
INPUTFILE=$1 # If even one argument is supplied, switch off legacy argument style
SEGLENGTH=$2 while getopts "i:o:s:c:b:p:t:S:lf" flag
if ! [ -z "$3" ] do
then LEGACY_ARGS=0
OUTPUT_DIRECTORY=$3 case "$flag" in
fi i) INPUTFILE="$OPTARG";;
  o) OUTPUT_DIRECTORY="$OPTARG";;
  s) SEGLENGTH="$OPTARG";;
  l) LIVE_STREAM=1;;
  c) LIVE_SEGMENT_COUNT="$OPTARG";;
  b) OP_BITRATES="$OPTARG";;
  f) NO_FORK=1;;
  p) PLAYLIST_PREFIX="$OPTARG";;
  t) SEGMENT_PREFIX="$OPTARG";;
  S) SEGMENT_DIRECTORY="$OPTARG";;
  esac
  done
   
   
  if [ "$LEGACY_ARGS" == "1" ]
  then
  # Old Basic Usage is
  # cmd.sh inputfile segmentlength
   
  INPUTFILE=${INPUTFILE:-$1}
  SEGLENGTH=${SEGLENGTH:-$2}
  if ! [ -z "$3" ]
  then
  OUTPUT_DIRECTORY=$3
  fi
  fi
   
   
# Check we've got the arguments we need # Check we've got the arguments we need
if [ "$INPUTFILE" == "" ] || [ "$SEGLENGTH" == "" ] if [ "$INPUTFILE" == "" ] || [ "$SEGLENGTH" == "" ]
then then
print_usage print_usage
fi fi
   
   
   
# FFMpeg is a pre-requisite, so let check for it # FFMpeg is a pre-requisite, so let check for it
if hash ffmpeg 2> /dev/null if hash $FFMPEG 2> /dev/null
then then
# FFMpeg exists # FFMpeg exists
echo "ffmpeg command found.... continuing" echo "ffmpeg command found.... continuing"
else else
# FFMPeg doesn't exist, uh-oh! # FFMPeg doesn't exist, uh-oh!
echo "Error: FFmpeg doesn't appear to exist in your PATH. Please addresss and try again" echo "Error: FFmpeg doesn't appear to exist in your PATH. Please addresss and try again"
exit 1 exit 1
fi fi
   
  # Check whether the input is a named pipe
# Now we want to make sure out input file actually exists if [ -p "$INPUTFILE" ]
if ! [ -f "$INPUTFILE" ] then
then echo "Warning: Input is FIFO - EXPERIMENTAL"
echo "Error: You gave me an incorrect filename. Please re-run specifying something that actually exists!" IS_FIFO=1
exit 1  
fi fi
   
   
   
# OK, so from here, what we want to do is to split the file into appropriately sized chunks,  
# re-encoding each to H.264 with MP3 audio, all to go into an MPEG2TS container  
#  
# The protocol appears to support MP4 as well though, so we may well look at that later.  
#  
# Essentially we want to create the chunks by running  
#  
# ffmpeg -i "$INPUTFILE" -vcodec libx264 -acodec mp3 -ss "START_POINT" -t "$SEGLENGTH" -f mpegts $OUTPUT_DIRECTORY/"$INPUTFILE"_"$N".ts  
   
# First we need the duration of the video  
DURATION=$($FFMPEG -i "$INPUTFILE" 2>&1 | grep Duration | cut -f 4 -d ' ')  
   
# Now we need to break out the duration into a time we can use  
DUR_H=$(echo "$DURATION" | cut -d ':' -f 1)  
DUR_M=$(echo "$DURATION" | cut -d ':' -f 2)  
DUR_X=$(echo "$DURATION" | cut -d ':' -f 3 | cut -d '.' -f 1)  
   
# Calculate the duration in seconds  
DURATION_S=$(expr \( $DUR_H \* 60 + $DUR_M \) \* 60 + $DUR_X)  
   
   
# Check we've not got empty media  
if [ "$DURATION_S" == "0" ]  
then  
echo "You've given me an empty media file!"  
exit 1  
fi  
   
   
# Now we've got our Duration, we need to work out how many segments to create  
N='1'  
START_POS='0'  
let 'N_FILES = DURATION_S / SEGLENGTH + 1'  
   
# Check output directory exists otherwise create it # Check output directory exists otherwise create it
if [ ! -w $OUTPUT_DIRECTORY ] if [ ! -w $OUTPUT_DIRECTORY ]
then then
echo "Creating $OUTPUT_DIRECTORY" echo "Creating $OUTPUT_DIRECTORY"
mkdir -p $OUTPUT_DIRECTORY mkdir -p $OUTPUT_DIRECTORY
fi fi
   
  if [ "$LIVE_STREAM" == "1" ]
  then
  FFMPEG_ADDITIONAL+="-segment_list_flags +live"
   
  if [ "$LIVE_SEGMENT_COUNT" -gt 0 ]
  then
  FFMPEG_ADDITIONAL+=" -segment_list_size $LIVE_SEGMENT_COUNT"
  fi
  fi
   
   
# Pulls file name from INPUTFILE which may be an absolute or relative path. # Pulls file name from INPUTFILE which may be an absolute or relative path.
INPUTFILENAME=${INPUTFILE##*/} INPUTFILENAME=${INPUTFILE##*/}
   
# Create the M3U8 file # If a prefix hasn't been specified, use the input filename
create_m3u8 "$INPUTFILENAME" "$SEGLENGTH" PLAYLIST_PREFIX=${PLAYLIST_PREFIX:-$INPUTFILENAME}
  SEGMENT_PREFIX=${SEGMENT_PREFIX:-$PLAYLIST_PREFIX}
# Finally, lets build the output filename format  
OUT_NAME=$INPUTFILENAME"_%03d.ts" # The 'S' option allows segments and bitrate specific manifests to be placed in a subdir
  SEGMENT_DIRECTORY=${SEGMENT_DIRECTORY:-''}
# Processing Starts  
  if [ ! "$SEGMENT_DIRECTORY" == "" ]
while [ "$START_POS" -lt "$DURATION_S" ] then
do  
  if [ ! -d "${OUTPUT_DIRECTORY}/${SEGMENT_DIRECTORY}" ]
OUTPUT=$( printf "$OUT_NAME" "$N" ) then
echo "Creating $OUTPUT ($N/$N_FILES)..." mkdir "${OUTPUT_DIRECTORY}/${SEGMENT_DIRECTORY}"
$FFMPEG -i "$INPUTFILE" -loglevel quiet -vcodec libx264 -acodec mp3 -ss "$START_POS" -t "$SEGLENGTH" -f mpegts $OUTPUT_DIRECTORY/"$OUTPUT" fi
   
let "N = N + 1" SEGMENT_DIRECTORY+="/"
let "START_POS = START_POS + SEGLENGTH" fi
   
# If we're on the last segment, the duration may be less than the seglenth, so we need to reflect this in the m3u8 # Set the bitrate
if ! [ "$START_POS" -lt "$DURATION_S" ] if [ ! "$OP_BITRATES" == "" ]
then then
SEG_DURATION=$($FFMPEG -i $OUTPUT_DIRECTORY/"$OUTPUT" 2>&1 | grep Duration | cut -f 4 -d ' ') # Make the bitrate list easier to parse
# Now we need to break out the duration into a time we can use OP_BITRATES=${OP_BITRATES//,/$'\n'}
DUR_H=$(echo "$SEG_DURATION" | cut -d ':' -f 1)  
DUR_M=$(echo "$SEG_DURATION" | cut -d ':' -f 2) # Create an array to house the pids for backgrounded tasks
DUR_X=$(echo "$SEG_DURATION" | cut -d ':' -f 3 | cut -d '.' -f 1) declare -a PIDS
  declare -a BITRATE_PROCESSES
# Calculate the duration in seconds  
let "SEGLENGTH = ( DUR_H * 60 + DUR_M ) * 60 + DUR_X" # Get the variant playlist created
fi createVariantPlaylist "$OUTPUT_DIRECTORY/${PLAYLIST_PREFIX}_master.m3u8"
  for br in $OP_BITRATES
# Append the file reference to the M3U8 do
append_segment "$INPUTFILENAME" "$SEGLENGTH" "$OUTPUT" appendVariantPlaylistentry "$OUTPUT_DIRECTORY/${SEGMENT_DIRECTORY}${PLAYLIST_PREFIX}_master.m3u8" "${PLAYLIST_PREFIX}_${br}.m3u8" "$br"
  done
done  
  OUTPUT_DIRECTORY+=$SEGMENT_DIRECTORY
# Add the close tag (ffplay gives some weird behaviour without this!)  
close_m3u8 "$INPUTFILENAME" # Now for the longer running bit, transcode the video
  for br in $OP_BITRATES
  do
  BITRATE="-b:v ${br}k -bufsize ${br}k"
  # Finally, lets build the output filename format
  OUT_NAME="${SEGMENT_PREFIX}_${br}_%05d.ts"
  PLAYLIST_NAME="$OUTPUT_DIRECTORY/${PLAYLIST_PREFIX}_${br}.m3u8"
  SOURCE_FILE="$INPUTFILE"
  echo "Generating HLS segments for bitrate ${br}k - this may take some time"
   
  if [ "$NO_FORK" == "0" ] || [ "$LIVE_STREAM" == "1" ]
  then
  # Processing Starts
  if [ "$IS_FIFO" == "1" ]
  then
  # Create a FIFO specially for this bitrate
  SOURCE_FILE="$TMPDIR/hlsc.encode.$MYPID.$br"
  mknod "$SOURCE_FILE" p
  fi
   
  # Schedule the encode
  createStream "$PLAYLIST_NAME" "$OUT_NAME" "$BITRATE" "$SOURCE_FILE" &
  PID=$!
  PIDS=(${PIDS[@]} $PID)
  BITRATE_PROCESSES=(${BITRATE_PROCESSES[@]} $br)
  else
  createStream "$PLAYLIST_NAME" "$OUT_NAME" "$BITRATE" "$SOURCE_FILE"
  fi
   
  done
   
  if [ "$IS_FIFO" == "1" ]
  then
  # If the input was a FIFO we need to read from it and push into the new FIFOs
  cat "$INPUTFILE" | tee $(for br in $OP_BITRATES; do echo "$TMPDIR/hlsc.encode.$MYPID.$br"; done) > /dev/null &
  TEE_PID=$!
  fi
   
  if [ "$NO_FORK" == "0" ] || [ "$LIVE_STREAM" == "1" ]
  then
  # Monitor the background tasks for completion
  echo "All transcoding processes started, awaiting completion"
  awaitCompletion
  fi
   
  if [ "$IS_FIFO" == "1" ]
  then
  for br in $OP_BITRATES
  do
  rm -f "$TMPDIR/hlsc.encode.$MYPID.$br";
  done
  # If we were interrupted, tee may still be running
  kill $TEE_PID 2> /dev/null
  fi
   
  else
   
  OUTPUT_DIRECTORY+=$SEGMENT_DIRECTORY
  # No bitrate specified
   
  # Finally, lets build the output filename format
  OUT_NAME="${SEGMENT_PREFIX}_%05d.ts"
  PLAYLIST_NAME="$OUTPUT_DIRECTORY/${PLAYLIST_PREFIX}.m3u8"
   
  echo "Generating HLS segments - this may take some time"
   
  # Processing Starts
   
  createStream "$PLAYLIST_NAME" "$OUT_NAME" "$BITRATE" "$INPUTFILE"
   
   
  fi
   
file:a/README.md -> file:b/README.md
HLS-Stream-Creator HLS-Stream-Creator
================== ==================
   
Introduction Introduction
------------- -------------
   
HLS-Stream-Creator is a simple BASH Script designed to take a media file, segment it and create an M3U8 playlist for serving using HLS. HLS-Stream-Creator is a simple BASH Script designed to take a media file, segment it and create an M3U8 playlist for serving using HLS.
There are numerous tools out there which are far better suited to the task, and offer many more options. This project only exists because I was asked to look There are numerous tools out there which are far better suited to the task, and offer many more options. This project only exists because I was asked to look
into HTTP Live Streaming in depth, so after reading the [IETF Draft](http://tools.ietf.org/html/draft-pantos-http-live-streaming-11 "HLS on IETF") I figured I'd start with the basics by creating a script to encode arbitrary video into a VOD style HLS feed. into HTTP Live Streaming in depth, so after reading the [IETF Draft](http://tools.ietf.org/html/draft-pantos-http-live-streaming-11 "HLS on IETF") I figured I'd start with the basics by creating a script to encode arbitrary video into a VOD style HLS feed.
   
   
   
Usage Usage
------ ------
   
Usage is incredibly simple Usage is incredibly simple
   
``` ```
./HLS-Stream-Creator.sh inputfile segmentlength(seconds) [outputdir='./output'] ./HLS-Stream-Creator.sh -[lf] [-c segmentcount] -i [inputfile] -s [segmentlength(seconds)] -o [outputdir] -b [bitrates]
   
   
  Deprecated Legacy usage:
  HLS-Stream-Creator.sh inputfile segmentlength(seconds) [outputdir='./output']
   
``` ```
   
So to split a video file called *example.avi* into segments of 10 seconds, we'd run So to split a video file called *example.avi* into segments of 10 seconds, we'd run
   
``` ```
./HLS-Stream-Creator.sh example.avi 10 ./HLS-Stream-Creator.sh -i example.avi -s 10
``` ```
   
  **Arguments**
   
  ```
  Mandatory Arguments:
   
  -i [file] Input file
  -s [s] Segment length (seconds)
   
  Optional Arguments:
   
  -o [directory] Output directory (default: ./output)
  -c [count] Number of segments to include in playlist (live streams only) - 0 is no limit
  -b [bitrates] Output video Bitrates in kb/s (comma seperated list for adaptive streams)
  -p [name] Playlist filename prefix
  -t [name] Segment filename prefix
  -l Input is a live stream
  -f Foreground encoding only (adaptive non-live streams only)
  -S Name of a subdirectory to put segments into
  ```
   
   
  Adaptive Streams
  ------------------
   
  As of [HLS-6](http://projects.bentasker.co.uk/jira_projects/browse/HLS-6.html) the script can now generate adaptive streams with a top-level variant playlist for both VoD and Linear input streams.
   
  In order to create seperate bitrate streams, pass a comma seperated list in with the *-b* option
   
  ```
  ./HLS-Stream-Creator.sh -i example.avi -s 10 -b 28,64,128,256
  ```
   
  By default, transcoding for each bitrate will be forked into the background - if you wish to process the bitrates sequentially, pass the *-f* option
   
  ```
  ./HLS-Stream-Creator.sh -i example.avi -s 10 -b 28,64,128,256 -f
  ```
   
  In either case, in accordance with the HLS spec, the audio bitrate will remain unchanged
   
   
Output Output
------- -------
   
As of version 1, the HLS resources will be output to the directory *output*. These will consist of video segments encoded in H.264 with MP3 audio (should be AAC really, but I'd compiled *ffmpeg* without) and an m3u8 file in the format As of version 1, the HLS resources will be output to the directory *output*. These will consist of video segments encoded in H.264 with AAC audio and an m3u8 file in the format
   
>\#EXTM3U >\#EXTM3U
>\#EXT-X-MEDIA-SEQUENCE:0 >\#EXT-X-MEDIA-SEQUENCE:0
>\#EXT-X-VERSION:3 >\#EXT-X-VERSION:3
>\#EXT-X-TARGETDURATION:10 >\#EXT-X-TARGETDURATION:10
>\#EXTINF:10, no desc >\#EXTINF:10, no desc
>example_001.ts >example_00001.ts
>\#EXTINF:10, no desc >\#EXTINF:10, no desc
>example_002.ts >example_00002.ts
>\#EXTINF:10, no desc >\#EXTINF:10, no desc
>example_003.ts >example_00003.ts
>\#EXTINF:5, no desc >\#EXTINF:5, no desc
>example_004.ts >example_00004.ts
>\#EXT-X-ENDLIST >\#EXT-X-ENDLIST
   
   
   
   
   
   
Using a Specific FFMPEG binary Using a Specific FFMPEG binary
------------------------------- -------------------------------
   
There may be occasions where you don't want to use the *ffmpeg* that appears in PATH. At the top of the script, the ffmpeg command is defined, so change this to suit your needs There may be occasions where you don't want to use the *ffmpeg* that appears in PATH. At the top of the script, the ffmpeg command is defined, so change this to suit your needs
   
``` ```
FFMPEG='/path/to/different/ffmpeg' FFMPEG='/path/to/different/ffmpeg'
``` ```
   
   
  Additional Environment Variables
  -------------------------------
   
  There are few environment variables which can control the ffmpeg behaviour.
   
  * `VIDEO_CODEC` - The encoder which will be used by ffmpeg for video streams. Examples: _libx264_, _nvenc_
  * `AUDIO_CODEC` - Encoder for the audio streams. Examples: _aac_, _libfdk_acc_, _mp3_, _libfaac_
  * `NUMTHREADS` - A number which will be passed to the `-threads` argument of ffmpeg. Newer ffmpegs with modern libx264 encoders will use the optimal number of threads by default.
  * `FFMPEG_FLAGS` - Additional flags for ffmpeg. They will be passed without any modification.
   
  Example usage:
   
  ```
  export VIDEO_CODEC="nvenc"
  export FFMPEG_FLAGS="-pix_fmt yuv420p -profile:v"
  ./HLS-Stream-Creator.sh example.avi 10
  ```
   
License License
-------- --------
   
HLS-Stream-Creator is licensed under the [BSD 3 Clause License](http://opensource.org/licenses/BSD-3-Clause) and is Copyright (C) 2013 [Ben Tasker](http://www.bentasker.co.uk) HLS-Stream-Creator is licensed under the [BSD 3 Clause License](http://opensource.org/licenses/BSD-3-Clause) and is Copyright (C) 2013 [Ben Tasker](http://www.bentasker.co.uk)
   
   
  Issue Tracking
  ----------------
   
  Although the Github issue tracker can be used, the bulk of project management (such as it is) happens in JIRA. See [projects.bentasker.co.uk](http://projects.bentasker.co.uk/jira_projects/browse/HLS.html) for a HTML mirror of the tracking.