Provided by: ripit_3.9.0-2ubuntu1_all 
NAME:
ripit - Perl script to create flac, ogg, mp3, m4a (aac), als (mp4), mpc, wv or other files
from an audio CD.
SYNOPSIS:
ripit [options]
DESCRIPTION:
This Perl script makes it a lot easier to create compressed sound files from an audio CD.
RipIT supports Flac, Lame, Oggenc, Faac, mp4als, Musepack, Wavpack and ffmpeg. Artist and
song titles are retrieved either with the CDDB_get.pm from freedb.org or using
WebService::MusicBrainz.pm from MusicBrainz.org.
It is possible to submit and edit CDDB entries at freedb.org, submission of MusicBrainz
data need a login. Hidden tracks and ghost songs are detected and split into chunks of
sound, a toc, cue or inf files permit to burn the WAV files with text and no gaps in DAO
mode. Several encoder formats and qualities can be used in the same run (even the same
encoder) and encoded into different directories. Distributed ripping and launch of more
than one encoder process in the same time is possible.
Several manipulations on tags are supported, e.g. adding coverart, detection of Various
Artists style at different levels or album gain; external commands can be launched before
and after ripping / encoding is done with all CDDB information, e.g. to ensure presence of
coverart, calculation of md5sums and playlistfiles can be done.
VERSION:
RipIT Version 3.9.0
OPTIONS:
[track_selection]
Tracks to rip from. If not specified, all tracks will be ripped. Specify a single
track by using a single number, or a selection of tracks using numbers separated by
commas or hyphens, e. g. 2,6,10, or 3,5,7-9. Using a number followed by a comma or
hyphen will rip from that track to the end of the CD i. e. 3,5- rips track 3 and
from track 5 to the last one. Default: not set.
--merge ordered list of comma separated intervals
Place a hyphen (-) or a plus (+) between first and last track number to be merged.
Omitting the last number will merge all subsequent tracks. Separate several
intervals with commas. Default: not set.
-I, --span start-endtime
Define an interval to be ripped when using cdda2wav or cdparanoia. The hyphen is
mandatory. The format of start and end-time is according to cdparanoia hh:mm:ss.ff
in hours/minutes/seconds/frames, but precision is set to seconds, i. e. frames
will be ignored. Zero fields need not be specified and the value of seconds may be
larger than 60. When using cdda2wav this option translates the values to the -o and
-d options of cdda2wav. When using option merge, endtime is related to the last
track in the selection. Default: not set.
-o, --outputdir dir
Where the sound should go. If not set, $HOME will be used. Default: not set.
-d, --device cddevice
Path of audio CD device, default: /dev/cdrom.
--scsidevice cddevice
Device name for special devices if the non ripping commands (except eject commands)
shall be executed on a different device node. This might be useful for some old
SCSI devices. If not set the cddevice of option --device will be used. Default: not
set
-r, --ripper number
CD ripper to use, 0 - dagrab, 1 - cdparanoia, 2 - cdda2wav, 3 - tosha, 4 - cdd.
Because cdparanoia and dagrab are the only rippers in this list that immediately
fail on data tracks, RipIT can create an error.log if problems are encountered and
continues to rip and encode without endless looping! Please use dagrab or
cdparanoia, otherwise RipIT might hang while waiting for data tracks. Default: 1 -
cdparanoia.
--ripopt options
User definable options for specific ripper. Quote them if necessary. Default: not
set.
--nicerip number
Set niceness of ripping process, default: 0.
-Z, --disable-paranoia [number]
When using dagrab, the number of retries will be set to 3, with cdparanoia this
option is equal to the -Z option of cdparanoia. The number can be set to 0
(paranoia on), to 1 (paranoia off) or to 2 to switch paranoia off if a track fails
or gets a timeout. RipIT will retry that failed track without paranoia or give up.
Ripping will then continue with paranoia on. Default: 0 ‐ off, i. e. use paranoia
mode.
--ghost
Analyze each wav for gaps and split it into chunks if gaps are detected and/or try
to trim lead-in/out. This may override option merge. Delete blank tracks if only
silence ("zero bytes") are found. Experimental! Default: off.
--extend seconds
When using option --ghost and only if a chunk of sound is detected, enlarge chunk
by amount of seconds at end if possible, or track may be trimmed if value is small
(e.g. 0.3), use with caution. Default: 2.0.
--prepend seconds
When using option --ghost and only if a chunk of sound is detected, enlarge chunk
by amount of seconds at beginning, if possible, or track may be trimmed if value is
small (e.g. 0.3), use with caution. Default: 2.0.
-c, --coder number
Encoder(s) to use, 0 - Lame (for mp3), 1 - Oggenc, 2 - Flac, 3 - Faac, 4 - mp4als,
5 - Musepack, 6 - Wavpack, 7 - ffmpeg, a comma separated list, or as an array. The
same encoder may be stated more than once, but different output-directories must be
specified stating the --ditemplate for each encoder or using the $quality parameter
in option --dirtemplate or --tracktemplate. See below. Default: 0.
--faacopt Faac-options
Pass other options to the encoder. If the same encoder has been stated more than
once, use a comma separated list. Quote them with double quotes if needed; default:
not set.
--flacopt Flac-options
Pass other options to the encoder. If the same encoder has been stated more than
once, use a comma separated list. Quote them with double quotes if needed; default:
not set.
--lameopt Lame-options
Pass other options to the encoder. If the same encoder has been stated more than
once, use a comma separated list. Quote them with double quotes if needed; default:
not set.
--museopt Musepack-options
Pass other options to the encoder. If the same encoder has been stated more than
once, use a comma separated list. Quote them with double quotes if needed; default:
not set.
--musenc Musepack-encoder
Choose Musepack encoder to be used, either mpcenc or mppenc for older versions;
default: mpcenc.
--mp4opt Mp4als-options
Pass other options to the encoder. Option -MP4 of encoder mp4als is mandatory to
enable tagging of sound files. Use the appropriate options to alter compression
level and speed. If the same encoder has been stated more than once, use a comma
separated list. Quote them with double quotes if needed; default: not set.
--oggencopt Oggenc-options
Pass other options to the encoder. If the same encoder has been stated more than
once, use a comma separated list. Quote them with double quotes if needed; default:
not set.
--wavpackopt Wavpack-options
Pass other options to the encoder. If the same encoder has been stated more than
once, use a comma separated list. Quote them with double quotes if needed; default:
not set.
--ffmpegopt ffmpeg-options
Pass other options to the encoder. If the ffmpeg has been stated more than once,
use a comma separated list. Quote them with double quotes if needed; default: not
set.
--ffmpegsuffix ffmpeg-suffix
The suffix of the formats to be encoded. If ffmpeg has been stated more than once,
use a comma separated list. Quote them with double quotes if needed; default: not
set.
-q, --quality quality
A comma separated list of values or the word "off", passed in the same order as the
list of encoders! If no encoders passed, follow the order of the config file!
Quality for oggenc: (1-10), highest = 10; compression level for Flac: (0-8), lowest
compression = 0; quality for Lame in vbr mode: (0-9), best quality = 0; or quality
for Faac: (10-500), highest = 500; mp4als seems not to accept any compression
level, set to zero; quality for Musepack: (0-10), best quality = 10; default:
5,3,5,100,0,5. Wavpack and ffmpeg have no quality switch, use encoder options
instead. The value "off" turns option quality off.
E. g. --coder 2,0,1,3 --quality 3,4,5,300 sets compression level for Flac to 3,
Lame-quality to 4, Oggenc-quality to 5 and Faac quality to 300.
-v, --vbrmode mode
Variable bitrate, only used with Lame, mode is new or old, see the Lame manpage.
The Lame-option quality will be changed to -V instead of -q if vbr-mode is used;
default: not set.
-b, --bitrate rate
Encode "mp3" at this bitrate for Lame. If option --vbrmode used, bitrate is equal
to the -b option, so one might want to set it "off"; or set if "off" if all options
are passed to --lameopt! Default: 128.
-B, --maxrate rate
maxrate (Bitrate) for Lame using --vbrmode is equal to the -B option in Lame or the
-M option in Oggenc, default: 0.
-S, --preset mode
Use the preset switch when encoding with Lame. Combined with option --vbrmode new
the preset fast will be used. Use --vbrmode old not to use fast VBR. Mode is one
of: insane (320 kbps @ CBR), extreme (256 kbps), standard (192 kbps) or medium (160
kbps) or any other valid bitrate. Default: off.
-W, --chars [list]
Exclude special characters in file names and path. The argument is optional. The
argument accepts a list of any characters. One might escape the characters or use
double quotes to pass them on the command line. The argument also accepts the words
NTFS, HFS or off. Following characters will be erased, if no argument stated or the
word NTFS is passed: |\:*?$ plus blanks and periods at beginning and end of file
names and directories. When using the word HFS, only colons will be purged plus
blanks and periods at beginning of file names and directories. Pass the word "off"
as argument to set the option off again. Default: off.
--comment comment
Specify a comment for the comment-tag for Lame, Faac and Musepack, or the
description-tag for Oggenc and Flac. The special comment cddbid or discid will
place the cddbid (used by freedb) or the discid (used for MusicBrainz) into the
comment. Default: not set.
-g, --genre genre
Specify (and override CDDB) genre, must be a valid ID3-genre if using Lame, can
(but shouldn't) be anything if using other encoders, default: not set.
-y, --year year
Specify (and override CDDB) the year of year-tag for Lame, Faac and Musepack, or
the date-tag for Oggenc and Flac. Default: not set.
--coverart number
a comma separated list of numbers (if several encoders are used) according to
option --coder with values 0 (no) or 1 (yes), default 0.
--coverpath path
Path to the coverart picture to be included in the metadata of the sound files. The
picture must be present or e.g. acquired using the --precmd option. Default: not
set.
--mp3tags FRAME=tag
Additional tags to be added to the mp3 file(s) if encoder does not support the
frame or if some unofficial FRAMEs shall be used (e.g. a compilation tag set to 1:
TCMP=1). More than one --mp3tags can be used if several tags shall be added.
Default: not set.
--vatag number
Analyze tracknames for "various artists" style and split the metadata in case one
of the delimeters (colon, hyphen, slash or parenthesis) are found. Use unpair
numbers for the scheme "artist ? tracktitle" and pair numbers in the opposite case.
The artist will be compared to the argument of option --vastring (see below). If
the artist must be like --vastring and each track have a delimeter, use 1 (2), if
the artist must be like --vastring while only some tracks contain the delimeter,
use 3 (4), if no restrictions apply for the artist but all tracknames must have a
delimeter, use 5 (6) and finally, if only a few tracks contain a delimeter to be
used as splitting point, set --vatag to 7 (8). Default: not set.
--vastring string
A string (regular expression) that defines the "various artists" style, e.g.
Varios|VA. Default: \bVA\b|Variou*s|Various\sArtists
-D, --dirtemplate '"foo parameters"'
Use single and double quotes to pass the parameters of the templates! Allowed are
any legal characters, slashes for multi-level directories and following parameters:
$album, $artist, $iletter, $genre, $quality, $suffix, $trackname, $tracknum, $year
and $trackno; example: '"$suffix $quality/my_snd_dir/$iletter/$artist/$year ‐
$album"'. The parameter $iletter is the initial letter of the artist name;
$trackno is the total number of tracks of the release, this might be helpful to
distinguish different releases with exactly the same name; $quality is needed, if
one wants to encode the same format at different qualities in the same run. Note
that $quality reflects the encoder options and not the arguments of option
--quality which might be set to "off". Or state different dirtemplates: the option
is an array, so one can use several different dirtemplates in the same order as the
encoders. If less dirtemplates than encoders are stated, the last dirtemplate will
be used for subsequent encoders and the wav-directory, see EXAMPLES below.
Default: '"$artist - $album"'.
-T, --tracktemplate '"foo parameters"'
See above. The tracktemplate is not an array, only one can be stated. Default:
'"$tracknum $trackname"'.
--trackoffset number
Give a trackoffset to be added to the $tracknum variable (see above). Default: 0.
--sshlist list
Comma separated list of remote machines where RipIT should encode. The output path
must be the same for all machines. Specify the login (login@machine) only if not
the same for the remote machine. Else just state the machine names. See EXAMPLES
for more information, default: not set.
--scp If the file system can not be accessed on the remote machines, scp (copy) the WAV
files to the remote machines, default: off.
--local
Only used with option --sshlist; if all encoding shall be done on remote machines,
use --nolocal, default: on.
--mb Access MusicBrainz DB via WebService::MusicBrainz module instead of the http
protocol (see below). One needs the discid calculation perl module or command to be
installed and in the path. Negate to switch option off again. A non interacitve
freedb lookup will be done to retrieve a genre. Default: off.
--mbname login
Give the users MusicBrainz login name to submitt ISRCs to the database. Saved in
plain when using a config, default not set.
--mbpass password
Give the users MusicBrainz password to submitt ISRCs to the database. Saved in
plain when using a config, default not set.
--isrc number
Enable ISRC detection with number = 1 (using icedax) and submission to MusicBrainz;
default: 0
-C, --cddbserver server
CDDB server, either freedb.org, freedb2.org or musicbrainz.org. Default
freedb.org. Note, the full address is "mirror".freedb.org, i. e. default is
freedb.freedb.org. Exception: freedb2.org and musicbrainz.org have no mirrors,
their full addresses are: freedb2.org and freedb.musicbrainz.org on port 80 (see
transfer mode).
-t, --transfer mode
Transfer mode, cddb or http, will set default port to 8880 or 80 (for http),
default: cddb. Note: use http mode with CDDB servers freedb2.org or
musicbrainz.org.
-m, --mirror mirror
Choose "freedb" or one of the possible freedb mirrors, default: freedb. For more
information check the webpage www.freedb.org. There is no list of mirrors
available these days. This option might be obsolete as freedb uses random servers,
but the default setting is mandatory.
-L, --protocol level
CDDB protocol level for CDDB query. Level = 6 supports UTF-8 and level = 5 not. Use
level = 5 to suppress UTF-8. Cf. option --utftag below. Default: 6.
-P, --proxy address
The http proxy to use when accessing the cddb server. The CDDB protocol must be
http! Default: not set.
-n, --nice number
Set niceness of encoding process, default: 0.
-a, --archive
Read and save CDDB files in $HOME/.cddb/"category" directory, where the "category"
is one of the 11 CDDB categories. Default: off.
-e, --eject
Ejects the CD when finished, if hardware supports it. Use --noeject to switch it
off. Default: off.
--ejectcmd command
Command to be used for ejection or loading of CD tray. Should be specified within
FreeBSD even if option --eject is not used to enable RipIT to load the CD tray. One
does not need to mention the words close or eject or options like -t, RipIT does
it. Default: eject.
--ejectopt options
Options for ejection command to be used for ejection or loading of CD tray.
Default: the path to the CD device.
--halt Powers off the machine when finished if configuration supports it, default: off.
-s, --submission
Specify --nosubmission if the computer is offline and the created file cddb.toc
shall be saved in the home directory instead of being submitted. With option
--archive it will also be saved in the $HOME/.cddb directory. Note: it is really
easy to resubmit incomplete CDDB entries! One can confirm each existing field with
Enter and add a missing genre or year. The purpose of this option is also to permit
the user to edit the CDDB data for the own filenames & tags and not to overwrite
the original CDDB entry! Default: on.
-M, --mail address
Users return email address, needed for submitting an entry to freedb.org. Default:
not set.
-p, --playlist number
Create the m3u playlist file, or use --playlist 0. For filenames without full path
use --playlist 2. Default is full path for filenames. Default 1 - on.
-A, --book number
Create an audiobook file, i. e. merge all tracks into one sinlge file, option
--ghost will be switched off and file suffix will be m4b instead of m4a. Make sure
to use encoder faac, ripit will not check for the encoder used. The installed
version must support the -w option to write tags. A chapter file will be written
for chapter marks. Default: off.
-i, --interaction
Specify --nointeraction if ripit shall take the first CDDB entry found and rip
without any questioning. Default: on.
--lcd Use lcdproc to display status, default: not set.
--lcdhost
Specify the lcdproc host, default: localhost.
--lcdport
Specify the lcdport, default: 13666.
--infolog file
Log operations (system calls, file/directory creation) to file given with full
path; default: not set.
-l, --lowercase
Lowercase filenames, default: off.
-u, --underscore
Use underscores (_) instead of spaces in filenames, default: off.
--uppercasefirst
Uppercase first letter of each word in filenames and tags. Use this option if the
CDDB entry is uppercase only. Default: off.
-U, --utftag
Keep Lame-tags in UTF-8 or decode them (but not the filenames) from UTF-8 to
ISO8859-1. Applies only to mp3 (Lame). May be useful for tags with special
characters and if mp3-player doesn't support Unicode. Using Musepack one might
want to pass the Musepack option --unicode to option museopt. Default: off.
--rip Rip the CD, to be used as --norip if wav-files are present. Used for debugging.
Default: not set.
--encode
Do encode the wavs. If only the wav-files shall be created, prevent encoding with
--noencode (and use option --wav, see below). Default: on.
-w, --wav
Keep the wav files after encoding instead of deleting them, default: off.
-N, --normalize
Normalizes the wave-files to a given dB-value (default: -12dB). Default: off.
This is useful when encoding
* for hardware or software that doesn't support the replaygain‐feature
* to audio formats without replaygain capabilities.
--normcmd
Command to use for normalizing, default: normalize.
-z, --normopt
Options to pass to normalize. Possible values:
-a -nndB : Normalize to -nn dB, default is -12dB,
Value range: All values <= 0dB
Example : normalize -a -20dB *.wav
-b : Batch mode ‐ loudness differences between
individual tracks of a CD are maintained
-m : Mix mode ‐ all track are normalized to the
same loudness
-v : Verbose operation
-q : Quiet operation
For further options see normalize documentation (http://normalize.nongnu.org).
Please note that normalizing to a high dB value, e.g. -6dB or less will lead to
severe compression of the audio signal, thus losing sound quality. Keeping the
default value of -12dB is recommended.
Default: -b Option v (q) will be set according to verbosity chosen.
--cdtoc number
Number n = 1 to create a toc file to burn the wavs with CD-Text using cdrdao or
cdrecord. Default: off.
--inf number
Number n = 1 to create inf files to burn the wavs with CD-Text using wodim or
cdrecord. Default: off.
--cdcue number
Number n = 1 or 2 to create cue sheets to burn the wavs with CD-Text. The cue
sheets will be copied to the other file formats. Number 1 will swith option --ghost
on to eliminate chunks of silence. Number 2 is experimental but does not switch
option --ghost on and may be used if number 1 fails. Default: off.
-h, --help
Print this and exit.
-V, --version
Print version and exit.
-x, --verbose number
Run silent (do not output comments, status etc.) (n = 0), with minimal output (n =
1), normal without encoder messages (n = 2), normal (n = 3), verbose (n = 4),
extremely verbose (n = 5). Default 3.
--config
Read parameters from config file or specify --noconfig to prevent reading it. A
local config file in $HOME/.ripit/ will have higher priority, only if the local
file does not exist, ripit will try to read /etc/ripit/config. Default: on.
--save Add parameters passed on command line to an existing config file. This does not
overwrite other existing settings. Options of an existing config file will be
adopted and the old file saved as config.old. Default: off.
--savenew
Save all parameters passed on command line to a new config file, backup an existing
file to config.old. Creates a default config file if no options are passed on
command line. Default: off.
--loop value
Continue ripping and encoding as soon as the previous CD finished. This option
forces ejection (--eject). The value is either 0 for normal use or 1 for standard
looping. Experimental value 2 migth be used to force immediate restart of ripping
process. This will spoil the output, this will show the prompt although a child
process is running (waiting for a new CD) and must be terminated manually. Default:
off.
--quitnodb value
Give up CD if no CDDB entry found. Useful if option --loop or --nointeraction are
on. Default behaviour is to let operator enter data or to use default artist, album
and track names. Possible values: 0 - off, 1 - on, default: off
--resume
Resume a previously started session. Default: off.
-O, --overwrite argument
Default behaviour of Ripit is not to overwrite existing directories, a suffix will
be added if directory name exists. Use option overwrite to prevent this and either
overwrite a previous rip (y) or force Ripit to quit (q) or even eject the disc (e).
If ejection is chosen, the disc will be ejected even if option --eject has not been
switched on. Default: off (n)
--dpermission number
Change the directory permission to number. Default: 0755.
--fpermission number
Change the file permission to number. Default: 0644.
--md5sum
Create a MD5-sum file for each type of sound files.
--threads number
Comma separated list of numbers giving maximum of allowed encoders to run at the
same time. In conjunction with option sshlist the number for the local machine is
mandatory and must be stated first. Default: 1.
-X, --execmd command
State a command to be executed when ripit finshed. The same variables as in the
dirtemplate can be stated. When using MusicBrainz one can try to use $cd{asin} to
get the ASIN. Make sure to escape the command if needed. Default: not set.
--precmd command
State a command to be executed before ripping started. The same variables as in the
dirtemplate can be stated. When using MusicBrainz one can try to use $cd{asin} to
get the ASIN. Make sure to escape the command if needed. Can be used to execute a
script to retrieve coverart. Default: not set.
EXAMPLES
To specify a CD device, type
ripit --device /dev/sr1
To specify the output directory, type
ripit --outputdir /foo/paths/
To rip and encode a special track selection, type
ripit 1,3-6,8-11
To use several encoders in the same run, type
ripit --coder 1,0,2 --quality 3,5,6
To use Lame with variable bitrate (VBR), type
ripit --vbrmode new --bitrate 0
Note, one should reset the --bitrate to 0 (zero) if the -b option of Lame is not desired.
According to VBR mode in Lame, use "new" or "old".
It is recommended to use the preset switches for Lame, (see Lame man-page) and specify
fast encoding with --vbrmode new (use --vbrmode old to switch off the fast encoding)
ripit --preset extreme --vbrmode new
To pass options related to the encoder use the according encoder option switch: e.g. to
use the managed mode with oggenc:
ripit -c 1 --oggencopt "--managed -b 192 -M 224 -m 96" --quality "off"
Normalize your wave-files prior to encoding:
ripit --normalize
Normalize your files to -18dB using "mix mode" (see normalize documentation)
ripit --normalize --normopt "-m -a -18dB"
To define a directory template where the sound files should go, type
ripit --dirtemplate '"$artist - $year"'
To create 2 directory levels, as used in other environments, type
ripit --dirtemplate '"$artist/$album"'
To save a config file in ~/home/.ripit/ with options: to use Lame and Oggenc, don't create
a m3u file, archive the CDDB entry files in ~/.cddb/"category"/ and to eject CD when done,
type
ripit --coder 0,1 --playlist 0 --archive --eject --save
To do the job without any interaction, type
ripit --nointeraction
To use a network for encoding, make sure that the output paths are equal on all machines!
ripit --sshlist sun,saturn,earth
where sun, saturn and earth are remote machines on which a user can login via ssh without
entering a password or passphrase! Note that the paths must be equal for the user on all
remote machines! If the login is different on some machines, try
ripit --sshlist login1@sun,login2@saturn,login3@earth
If there is "no" identical path on the remote machines, then the user might enter e.g.
/tmp/ as output directory. If the file-system is not mounted on each remote machine, one
can try to copy the wavs to the remote machines using option --scp.
ripit --sshlist sun,saturn,earth --scp
Check for ghost songs and trim very long lead-in/out:
ripit --ghost --prepend 1.0 --extend 1.0
Rip only first 30 seconds of each track (useful for testing)
ripit --span 0-30
Encode the same format in 2 qualities, at low quality for a portable, at high quality for
home:
ripit --dirtemplate '"cbr/$artist/$year ‐ $album"' \
--dirtemplate '"preset/$artist ‐ $album"' \
--coder 0,0 --bitrate off --quality off,off \
--lameopt "-b 128 -q 2;--preset extreme"
Force RipIT to use one and the same directory for all sound files:
ripit --dirtemplate '"/"' --outputdir="/media/sound" \
--resume
FILES
User config file (/etc/ripit/config will only be used if there is no $HOME/.ripit/config):
$HOME/.ripit/config
/etc/ripit/config
/usr/share/doc/packages/ripit/README
/usr/share/doc/packages/ripit/HISTORY
/usr/share/doc/packages/ripit/LICENSE
BUGS
Probably there are more than some.
SEE ALSO
cdparanoia(1), lame(1), oggenc(1), flac(1), normalize(1), cdda2wav(1), ffmpeg(1)
AUTHORS
RipIT is now maintained by Felix Suwald, please send bugs, wishes comments to
ripit_[at]_suwald_[dot]_com with a concise subject.
For bugs, wishes and comments about lcdproc, please contact
max.kaesbauer_[at]_gmail_[dot]_com.
Former maintainer: Mads Martin Joergensen; RipIT was originally developed by Simon
Quinn.
This manpage was originally created for the Debian GNU/Linux system by help2man 1.36 and
optimized by Elimar Riesebieter <riesebie@lxtec.de>, enhanced and updated by Felix Suwald.