Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
tsdecrypt reads and decrypts CSA encrypted incoming mpeg transport stream over UDP/RTP using code words obtained from OSCAM or similar CAM server. tsdecrypt communicates with CAM server using cs378x (camd35 over tcp) protocol or newcamd protocol.
C Makefile C++ Shell
Branch: master

Send ecm pid over camd35 protocol.

Since r10358 oscam expects ECM pid to be send over the camd3 protocol.
This allows ECMs to be retransmitted over gbox.

This change adds support for sending ecm pid.
latest commit f69577a82b
@gfto authored
Failed to load latest commit information.
FFdecsa Add FFdecsa in FFdecsa/ directory.
libfuncs @ 0dc8d87 libfuncs: Update the library.
libtsfuncs @ 45e7e99 libtsfuncs: Update the library.
.gitignore makefile: Track header dependancies in order to rebuild correct objects.
.gitmodules Move libtsfuncs to libtsfuncs submodule and change libfuncs submodule…
COPYING tsdecrypt is licensed under GPL v2 now.
ChangeLog Release version 10.
FFdecsa_init
FILTERING Add support for EMM filters based on section length.
Makefile build: Introduce LDFLAGS variable.
README Fix syntax error.
RELEASE Release version 10.
TODO TODO: Remove item.
camd-cs378x.c Send ecm pid over camd35 protocol.
camd-newcamd.c Prevent openssl warnings when build on OS X >= 10.7
camd.c Log server address that we connected to.
camd.h Add support for IPv6 and for multiple CAMD addresses.
csa.c csa: check if batch_size is zero.
csa.h No need to track where FSF moves. No code changes.
data.c Add support for Irdeto CHID filtering (tested with Raduga).
data.h Prevent openssl warnings when build on OS X >= 10.7
filter.c Remove dead assignments and increments.
filter.h Add EMM filtering support.
notify-script.example Increase stack size in notify-script.example.
notify.c No need to track where FSF moves. No code changes.
notify.h No need to track where FSF moves. No code changes.
process.c Remove dead assignments and increments.
process.h No need to track where FSF moves. No code changes.
tables.c Remove dead assignments and increments.
tables.h tables: Fix header guard definition.
tsdecrypt.1 Release version 10.
tsdecrypt.c Prevent openssl warnings when build on OS X >= 10.7
udp.c Fix invalid read caused by wrong memcpy size.
udp.h No need to track where FSF moves. No code changes.
util.c Create parse_host_and_port() function.
util.h Create parse_host_and_port() function.

README

tsdecrypt
=========
tsdecrypt reads incoming mpeg transport stream over UDP/RTP and then
decrypts it using code words obtained from OSCAM or similar cam
server. tsdecrypt communicates with CAM server using cs378x (camd35
over tcp) protocol or newcamd protocol.

License
=======
tsdecrypt is released under GNU GPL v2.

Dependencies
============
tsdecrypt depends on openssl. It uses rand, MD5, DES and AES functions.
You probably already have openssl installed. If you are seeing compilation
errors related to missing openssl headers you should install openssl-dev
or similar packet depending on your Linux distribution.

If you are not using shipped FFdecsa then the other tsdecrypt dependency
is libdvbcsa. libdvbcsa can be downloaded from
http://www.videolan.org/developers/libdvbcsa.html

FFdecsa vs libdvbcsa
====================
tsdecrypt supports two different CSA decryption libraries.

libdvbcsa is an external library from VideoLAN project released under
LGPL, FFdecsa is shipped with tsdecrypt.

libdvbcsa
=========
libdvbcsa is better tested and have served tsdcrypt for a long time so
it is still the default decryption library. Also it is faster with file
input.

To compile tsdecrypt with libdvbcsa run 'make' or if you have tested
FFdecsa compilation and want to switch to libdvbcsa run 'make dvbcsa'.
'make distclean' switches decryption library back to libdvbcsa.

FFdecsa
=======
FFdecsa support was added after tsdecrypt v7.0. FFdecsa can be up to 40%
faster than libdvbcsa especially on 32-bit hardware. If you want to use
FFDecsa you have to run 'make ffdecsa'.

The setup script will test which FFDecsa algorithm is the fastest for
your CPU and will configure tsdecrypt compilation to stop using libdvbcsa.
Most likely the script will choose PARALLEL_128_SSE or PARALLEL_128_SSE2.
Once the script have finished it stores its decision in FFdecsa.opts file
and as long as this file exists tsdecrypt will be compiled with FFdecsa.
You can force regeneration by deleting running 'make ffdecsa_force' or
'make distclean ffdecsa'.

FFDecsa is considerably slower than libdvbcsa when tsdecrypt's input
is a file or stdin (single packet decryption mode). If you are
decrypting already recorded files you are better of using libdvbcsa.

When tsdecrypt is run it reports which decryption library is used in its
version string. You can use --bench option to test decryption speed.

Development
===========
The development is done using git. tsdecrypt repository is hosted
at http://github.com/gfto/tsdecrypt

To clone the repository issue the following commands:

   git clone git://github.com/gfto/tsdecrypt.git
   cd tsdecrypt
   git submodule init
   git submodule update
   make

The code is developed and tested under modern Linux. It is also
compiled from time to time under OS X but is not tested there.

To see all Makefile targets run 'make help'.

Updating the code
=================
To update cloned tsdecrypt, go to the directory where the repository
is cloned and run the following commands:

   git fetch origin
   git merge origin/master
   git submodule update
   make clean all

tsdecrypt's master branch should always be useful so it is safe to
use it instead of official release. The master branch will always
be better than any released version.

Command line parameters
=======================
tsdecrypt is controlled using command line parameters. For more information
about the parameters see the man page. Here is a list of supported command
line parameters:

Main options:
 -i --ident <server>        | Format PROVIDER/CHANNEL. Default: empty
 -d --daemon <pidfile>      | Daemonize program and write pid file.
 -N --notify-program <prg>  | Execute <prg> to report events. Default: empty

Input options:
 -I --input <source>        | Where to read from. File or multicast address.
                            .    -I 224.0.0.1:5000    (v4 multicast)
                            .    -I [ff01::1111]:5000 (v6 multicast)
                            .    -I file://in.ts      (read from file)
                            . By default the input is stdin.
 -1 --input-source <ipaddr> | Set multicast input source ip.
 -R --input-rtp             | Enable RTP input
 -z --input-ignore-disc     | Do not report discontinuty errors in input.
 -M --input-service <srvid> | Choose service id when input is MPTS.
 -T --input-buffer <ms>     | Set input buffer time in ms. Default: 0
 -W --input-dump <filename> | Save input stream in file.

Output options:
 -O --output <dest>         | Where to send output. File or multicast address.
                            .    -O 239.0.0.1:5000    (v4 multicast)
                            .    -O [ff01::2222]:5000 (v6 multicast)
                            .    -O file://out.ts     (write to file)
                            . By default the output is stdout.
 -o --output-intf <value>   | Set multicast output interface.
                            . Default for IPv4: 0.0.0.0 (intf addr)
                            . Default for IPv6: -1      (intf number)
 -t --output-ttl <ttl>      | Set multicast ttl. Default: 1
 -r --output-rtp            | Enable RTP output.
 -k --output-rtp-ssrc <id>  | Set RTP SSRC. Default: 0
 -g --output-tos <tos>      | Set TOS value of output packets. Default: none
 -u --no-output-on-error    | Do not output data when the code word is missing.
 -p --no-output-filter      | Disable output filtering. Default: enabled
 -y --output-nit-pass       | Pass through NIT.
 -w --output-eit-pass       | Pass through EIT (EPG).
 -x --output-tdt-pass       | Pass through TDT/TOT.

CA options:
 -c --ca-system <ca_sys>    | Process input EMM/ECM from <ca_sys>.
                            | Valid systems are: CONAX (default), CRYPTOWORKS,
                            .   IRDETO, SECA (MEDIAGUARD), VIACCESS,
                            .   VIDEOGUARD (NDS), NAGRA, DRECRYPT, BULCRYPT,
                            .   GRIFFIN and DGCRYPT.
 -C --caid <caid>           | Set CAID. Default: Taken from --ca-system.
 -Y --const-cw <codeword>   | Set constant code word for decryption.
                            . Example cw: a1a2a3a4a5a6a7a8b1b2b3b4b5b6b7b8
 -Q --biss-key <biss-key>   | Set BISS key for decryption.
                            . Example key: 112233445566

CAMD server options:
 -A --camd-proto <proto>    | Set CAMD network protocol.
                            . Valid protocols are: CS378X (default) and NEWCAMD
 -s --camd-server <host>    | Set CAMD server address. Default port: 2233
                            . Example IPv4 addr and port: 127.0.0.1:2233
                            . Example IPv6 addr and port: [2a00::1014]:2233
                            . Example hostname          : example.com
                            . Example hostname and port : example.com:2233
                            . Example IPv4 hostname     : ipv4.google.com
                            . Example IPv6 hostname     : ipv6.google.com
 -U --camd-user <user>      | Set CAMD server user. Default: user
 -P --camd-pass <pass>      | Set CAMD server password. Default: pass
 -B --camd-des-key <key>    | Set DES key for newcamd protocol.
                            . Default: 0102030405060708091011121314
 -4 --ipv4                  | Use only IPv4 addresses of the camd server.
 -6 --ipv6                  | Use only IPv6 addresses of the camd server.

EMM options:
 -e --emm                   | Enable sending EMM's to CAMD. Default: disabled
 -E --emm-only              | Send only EMMs to CAMD, skipping ECMs and without
                            .   decoding the input stream.
 -Z --emm-pid <pid>         | Force EMM pid. Default: none
 -f --emm-report-time <sec> | Report each <sec> seconds how much EMMs have been
                            .   received/processed. Set <sec> to 0 to disable
                            .   the reports. Default: 60 sec
 -a --emm-filter <filter>   | Add EMM filter defined by <filter>.
                            . This option can be used multiple times (max:16).
                            . See FILTERING file for more info.

ECM options:
 -X --ecm-pid <pid>         | Force ECM pid. Default: none
 -v --ecm-only              | Send only ECMs to CAMD, skipping EMMs and without
                            .   decoding the input stream.
 -H --ecm-report-time <sec> | Report each <sec> how much ECMs and CWs have been
                            .   processed/skipped. Set <sec> to 0 to disable
                            .   the reports. Default: 60 sec
 -G --ecm-irdeto-type <int> | Process IRDETO ECMs with index X /0-255/
                            .   It is better to use --ecm-irdeto-chid option!
 -2 --ecm-irdeto-chid <int> | Set CHID to filter Irdeto ECMs (Default: 0x0000).
 -K --ecm-no-log            | Disable ECM and code words logging.
 -J --cw-warn-time <sec>    | Warn if no valid code word has been received.
                            .   Set <sec> to 0 to disable. Default: 60 sec

 -q --ecm-and-emm-only      | Send ECMs and EMMs to CAMD but do not decode
                            .   the input stream.

Logging options:
 -S --syslog                | Log messages using syslog.
 -l --syslog-host <host>    | Syslog server address. Default: disabled
 -L --syslog-port <port>    | Syslog server port. Default: 514
 -F --log-file <filename>   | Log to file <filename>.
 -D --debug <level>         | Message debug level.
                            .    0 = default messages
                            .    1 = show PSI tables
                            .    2 = show EMMs
                            .    3 = show duplicate ECMs
                            .    4 = packet debug
                            .    5 = packet debug + packet dump

Debugging options:
 -n --ecm-file <file.txt>   | Read ECM from text file.
 -m --emm-file <file.txt>   | Read EMM from text file.

Misc options:
 -j --pid-report            | Report how much packets were received.
 -b --bench                 | Benchmark decrypton.
 -h --help                  | Show help screen.
 -V --version               | Show program version.

Examples
========
To get a quick start here are some example command lines. The default
CA system is set to CONAX, you can change it using --ca-system parameter,
see man page or program help for more options.

Examples:

   # Decrypt multicast stream from 239.0.50.11:5000 using 10.0.1.1:2233
   # as camd server and output decrypted result to 239.78.78.78:5000
   tsdecrypt --camd-server 10.0.1.1 \
       --input 239.0.50.11:5000 --output 239.78.78.78:5000

   # Same as above but enable EMM processing
   tsdecrypt --emm --camd-server 10.0.1.1:2233 \
       --input 239.0.50.11:5000 --output 239.78.78.78:5000

   # Same as above but do not filter output stream thus allowing
   # EIT/TOT/NIT, etc tables to passthrough
   tsdecrypt --no-output-filter --emm --camd-server 10.0.1.1 \
       --input 239.0.50.11:5000 --output 239.78.78.78:5000

   # Choose program/service_id to decrypt. Useful when the input is MPTS
   tsdecrypt --camd-server 10.0.1.1 --input-service 1234 \
       --input 239.0.50.11:5000 --output 239.78.78.78:5000

   # Read stream over RTP and process VIACCESS encoded channel
   tsdecrypt --ca-system VIACCESS --emm --camd-server 10.0.1.1:2233 \
       --input-rtp --input 239.0.50.11:5000 --output 239.78.78.78:5000

   # Decrypt stream encypted with CAID 0x0963 (NDS, sky)
   tsdecrypt --camd-server 10.0.1.1 --ca-system NDS --caid 0x0963 \
       --input 239.0.50.11:5000 --output 239.78.78.78:5000

   # Send only EMMs to OSCAM for CAID 0x0963 (NDS, sky)
   tsdecrypt --camd-server 10.0.1.1 --emm-only --caid 0x0963 \
       --input 239.0.50.11:5000 --output /dev/null

   # Decrypt stream encypted with CAID 0x5581 (Bulcrypt)
   tsdecrypt --camd-server 10.0.1.1 --caid 0x5581 \
       --input 239.0.50.11:5000 --output 239.78.78.78:5000

   # Decrypt BISS encrypted stream
   tsdecrypt --biss-key 0x112233445566 --input 239.0.50.11:5000 \
       --output 239.78.78.78:5000

   # Decrypt file encypted with constant code word
   tsdecrypt --const-cw 0x00000000000000001111111111111111 \
       --input encrypted-file.ts --output file://decrypted-file.ts

   # Send ECM from file
   tsdecrypt --ecm-file ecm.txt --caid 0x5581 --input-service 12345 \
       --camd-server example.com

   # Decrypt IRDETO stream from Raduga (CHID == 0x0015)
   tsdecrypt --input 239.0.50.11:5000 --output 239.78.78.78:5000 \
             --camd-server example.com \
             --ca-system IRDETO --caid 0x0652 --ecm-irdeto-chid 0x0015

OSCAM cs378x configuration
==========================
In order for tsdecrypt to communicate with OSCAM using cs378x (camd35
over tcp) protocol you can use configuration like the examples bellow.

# Example oscam.user file:
[account]
user     = user
pwd      = pass
group    = 1
au       = 1
uniq     = 0
monlevel = 4

# Example part of oscam.conf file:
[cs378x]
port    = 2233

OSCAM newcamd configuration
===========================
In order for tsdecrypt to communicate with OSCAM using newcamd protocol
you can use configuration like the examples bellow.

# Example oscam.user file
[account]
user     = user
pwd      = pass
group    = 1
au       = 1
uniq     = 0
monlevel = 4
ident    = 0604:000000
caid     = 0604

# Example part of oscam.conf file:
[newcamd]
port     = 1122@0604:000000
serverip = 0.0.0.0
key      = 0102030405060708091011121314

The above example allows tsdecrypt clients using newcamd to retrive keys
from OSCAM serving CAID 0x0604 (Irdeto) on port 1122.

Reporting bugs
==============
If you think you have found bug in tsdecrypt, please report it to the
e-mail listed in Contact section (see below) of this README file.

When reporting bugs, please send the contents of tsdecrypt.log file
generated by using the following command (fill in **YOUR_OPTIONS**):

  tsdecrypt --debug 1 --log-file tsdecrypt.log **YOUR_OPTIONS**

Also write the full command line which you used and describe what
you think the tsdecrypt does wrong.

One commonly seen error is not an error in tsdecrypt at all. You
have to make sure that the software that streams channels from the
DVB card is sending ECMs and EMMs along with other streams. If these
streams are missing you'll probably get the following messages from
tsdecrypt:

2012-04-02 22:02:12 | ECM | Received 0 (0 dup) and processed 0 in 3 seconds.
2012-04-02 22:03:09 | CW  | *ERR* No valid code word was received for 60 seconds!
2012-04-02 22:03:12 | ECM | Received 0 (0 dup) and processed 0 in 60 seconds.

Please configure your software to stream ECMs and EMMs. For dvblast
(version 2.0 and above) the correct options are --ecm-passthrough and
--emm-passthrough. Note that dvblast 2.1 have broken ecm passthrough
so use at least version 2.2.

Releases
========
Official releases can be downloaded from tsdecrypt home page which is

   http://georgi.unixsol.org/programs/tsdecrypt/

Contact
=======
For patches, bug reports, complaints and so on send e-mail to

   Georgi Chorbadzhiyski <georgi@unixsol.org>
Something went wrong with that request. Please try again.