Hacking TS File using C/C++

What are you doing?

Imagine a Baseball game. It has its rules and when players follow them they have a nice game and have fun.

Now you been for the past few days trying to play. But you have not follow the rules nor even try to understand what are they. So you find your self lost. You have try all sort of twisted ways.

Now this question seems funny. As all the tools people have offered for the most part where written in C or C++? So what is the question then?

Stop all what you are doing. Start from fresh. Say who you really are? What is your interest? What are you trying to accomplish?

Then once people REALLY know what you are trying to do. You see you will get better help.

Now In general forget on trying to open the best packages. This game is not what we want to see. Instead is what we are allow to see. It will depends on what have been broken. In general, BISS, Power Vu and some tandbergs. Then there are very few system crack that are left from the old school. For that you just use the emu like Vplug or E2 boxes running Osemu or Oscam. That with trying to orient you.

Now lets go back to you question. The answer is yes. Many here on a daly basic use C/C++ to try new Ideas as to who to attack CSA or DES. I do it every day. Is that is what interest you, let me tell you you have a lot to learn. But do not get confuse. It can be done. I did it from 0. I got unemployed and decided to have fun learning crypt theory. But it is not easy. Required lot of time and reading. So if that is what you are looking then relax and start reading. You have a few years of learning. But it can be done as I and many has done it.

dale_para_bajo is right!!

you are making cwl file, where you think that you can make one cwl file with all cw keys. TSDEC format is "0 00 00 00 xx 00 00 00 xx" where are checsums and backspaces, which means a lot of bytes more that only useful key like CWFinder format "112233445566" and wordlist with all keys in cwfinder format will be reeeaally big file... so imagine how big will be cwl file with all keys... and with cw1 keys, because without there keys, the tsdec will show errors... you can only modify source code sor tsdec and maybe your cwl project will be useful with edited version of tsdec.

sorry but making wordlist with all keys is waste of time...

srijan20 said:

Yes, It can be the Waste of Time, But After Finding BISS Key You can Decrypt the TS File using TSDEC.

Click to expand...

i edited my reply btw... you aren't weird guy but it's a question what you really want to do.

if you are using tsdec for FINDING biss keys... it's waste of time
if you are using tsdec for DECODING ts record... then it's not waste of time

srijan20 said:

I Think,
BISS Key Can Be Found Fast By Using Latest Updated CSA Algorithm & Latest Cracking Software.

Click to expand...

CSA algorithm is the same for almost 20 years,there is no latest,no updates.
Biss key can be found in 1~2 minutes using rainbow tables.
What is your idea of faster than this?

sirijan20 do you have any relation with strong5000?

You both act the same way. Ask for advise, then never really read what we answer to you guys. You guys need to start listening.

kebien is correct about no updates. The difference from now and 20-30 years ago is that we have GPU for fast calculations. But CSA is the same.

I did offer to create a program for Fun & Learning. But you guys keep thinking you can make a Novel Prize program. So I guess I quit on that. I will give you my best advise.

Go and read on aycwabtu

Nice project. There you have the latest. HEHEHEHEHE

But listen that is a nice project designed to BF BISS just like the finders we have. That is my best suggestion for you. The program is no finish it run on CPU not GPU and in single core. But it does about 3.6x10^6 keys/s. Just set it up to do calc in SSE x64. Use compiler option

There you have it guys a nice program to have fun. Enjoy it.

kebien & dale_para_bajo

i very much appreciate your work and your knowledge on this forum also i wish you much success and happiness in your work

now, my comment:

faster way is to calculate the key - than "searching mode" to the key!
my idea, i like to think of it, why is the need to constantly "search" key using BF? i'm looking for the ability and shorter way to calculate the key !!!

one example as reason:

when this we be know and have learned, we can easily "find" cw key (without classic CSA RBT tables and without BF). this solution is not simply but i believe that this is possible, because mathematics is the essence of life. this is not a fantasy
i am a mathematician, but i do not know any programming language to create an algorithm and other application - that's my weakness and lack of knowledge.
i only use my knowledge of mathematics and MS Excel as a tool to decrypt i have also make many applications in the .xls extension. this is perhaps the hard way, but for me - it is not difficult
i'm still trying to find a solution to calculate the key. maybe there's also still someone else who's trying get same result on some other method ...

I know for one a funny story.

(maybe this short joke is the only association to the current topic)
... but today, microbiology is capable that of DNA from cells procreate duplicate every living organism .

Therefore, anything is possible

sorry but my knowledge of the English language is still modest

regards :thum:
shishmish

Why different timing Tanberg(DES) vs CSA
You can not compare oranges with bottles.

HEHEHE another one without explanation. I guess at one time in the history of my country people exchanged bottles for Oranges. Weird NO! It is like the word Salary. Well Spanish Government use to pay to all his employees with SALT!. And you say SALT worth nothing. well at the time there was no refrigeration SALT was use to keep meat and other things from spoiling. At that time SALT was a high commodity.

So it is not only that DES is less complicated it is in fact that it can be change into a GPU program more easily using less resources. See a GPU may start with 2000 cores but due to high resources maybe only 100 really perform work. Or due to Wave Front we can hide the time waist in transfer from different types of memory better in one than the other. It is complicated to explained. In fact only way I find to get the fastest is to TEST, TEST & TEST.

"faster way is to calculate the key - than "searching mode" to the key!"
wofff! how I explained this. A Crypt is build with the Intention that no one can "CALCULATE" the key back. Well it is analysed with the available tools at the time the protocol is proposed. Now NSA, USA nati0nal s3curity Ag3ncy do that. BUTTTT you know their current state of technology in the agency is at least 10 years ahead. So the proposed a Crypto Protocol that they know NO ONE in the world could break BUTTTT with their level of technology they can break. Same happens in Europe. So after proposing lets say CSA they realize it is still to strong for their like and put LAWs to prevent 64bit use and rendless it to 48bit.

So you should NOT have the resources to CALCULATE a Key. That is the reason for Stream Cypher & Block Cypher. Now even when I say you should NOT be able to Calculate, well there are some approaches that you me reach but never to the key. That I had ever heard no one had calculated this keys.

Now I went and tested CWFinder09-Beta as I said I have not seen it for more that 10 years. The one I have is Beta. Still nice. It does said it uses 128bit SSE instructions. That does not have to mean that it in fact do BitSlice!. Now after setting up to highest I get about 2x10^6. It uses 13% of my CPU, I can see at least 4 cores get work.

So as you see 10 years later it is considered as god as it can be.
Now the program aycwabtu I gave you as reference do the same 12% of usage on 4 cores but it do a litle more about 3.6x10^6.

NOW THE BIG WARNING. A GPU is designed with the knoledge that a GEEK will over use and Over Everything. So it has GOOD ventilation. And High End USers have Double the FAN output atleast.
You should never use a Laptop to BruteForce anything. Laptops are design with the minimun ventilation in mind. A general use Desktop may also not have the best ventilation. ONLY GAME PCs are build for constant use and abuse.

Well if you have a Good CPU, you can modify the aycwabtu program to run in multi core, this will quadruple the output initialy. But SOON CPU will GET HOT and the system will trotle down the SPEED of CPU to at least 1/4 of the normal speed. So ONLY on GAME PC You could get atleast 4 time the ouput on CPU MultiCore.

Now I do not even know why I am doing this as I said ALL the Good tools have been Made for long time. Your next be4st option is going into GPU. ANd for that you guys have Cudabiss.

Now lets make some numbers. CSA is 48bits. 2^48 = 281474976710656 posible keys

If I want to do all keys in 1 day for example I need to do

60 sec * 60 min * 24 hours = 86400 second in 1 day

So I need to do 281474976710656 / 86400 = 3257812230 or about

3.26x10^9 keys per second.

Now lest see where we are and how long will it take for CWFinder09-Beta to do all.

3.26x10^9 Speed needed / 2x10^6 speed of CWFinder09-Beta = 1630 days needed.

Now if you use Cudabiss with a good GPU

K2TSET said:

Depend on your numbers on cudacores and the version of cudabiss (some support more cores than other)

On a GTX970 with 1664 cuda cores: Crunching along at 305790944 keys/s Time to go: 255hours 35minutes

So about 10+ days for a full search in mode 1

Click to expand...

About 10 days. See the improvement. Here is where you should put your effort.

I forgot and here the BADD news.

After 10 days of Over Heating you High Performance PC you Home is not only warmer but you may had spent $50, $100, or $150 dollars in Electric Bill depending your energy cost rate!!!!

shishmish said:

Code:

how was it possible "found" or calculate 8-byte DES key decryption of ciphertext (nice number) "0000000000000000" which can gives the result - plaintext (also likable number) "8787878787878787"?
whether this is happiness? I do not think so

Click to expand...

i'm sorry, i forgot to write "magic" (found ???) DES Key: 0E329232EA6D0D73
... but the procedure of obtaining this result (key) is still hiding a big secret

srijan20 said:

@ dale_para_bajo

You Told That CSA is 20 Years Old & It is Not Updated. And You Give A Link of A Project.

I Saw Many CSA Projects, But the Similarity between all Projects are all of Them use libdvbcsa for their Program. Why?

So, My Questions Are:

What is libdvbcsa?
Why We use libdvbcsa?
And Where We Get The 20 Years Old CSA?
What the format of Old CSA? [C/ CPP/ DLL]

Click to expand...

This is how you want to reinvent the wheel

CSA was not written by any hacker.
CSA as we use today is the software implementation of a dedicated hardware in all boxes and CAM modules (where it was ripped out from).
It was accepted by ETSI as to protect the providers content.
This first implementation was done by 2000,17 years ago.
Still in use today,is not really about to be fast or the best,it was intended to be a secret inside the hardware.
Freedec was the first software that emulated it.
There is no old or new format,you surely know very little about it,no need to update,since has not been changed by the providers.

Crack yourself up if you do not want to use the libraries available,just do not ask others to do it for you,since really big minds worked together to produce the libary.
Feel free to do your own library,nobody stop you.

Now,you are confusing between what CSA is ,and how is implemented, with finding a CW that CSA needs as a seed in order to decrypt.

We all know you are not friend of similarities,so,please make your own library and projects.
My bet goes that rainbow tables beats you to the ground until time ends.
basically,at asking all this question,it shows that you can never acquire the necessary knowledge before they phase out all to CSAv3.

kebien said:

CSA algorithm is the same for almost 20 years,there is no latest,no updates.
Biss key can be found in 1~2 minutes using rainbow tables.
What is your idea of faster than this?

Click to expand...

That dosent take into account how long it takes to get crypt8, it can take a lot longer to get a valid c8 and sometimes you cant get.

@shishmish

tandgerg uses Americas DES as its core. NSA did the same as CSA. Even when the original designed was a 64bit they got smart and put International Commercial Laws to manipulate what the US Users or companies could export from the US. So they limit Crypt exportation to 56bits.

So it was decided to use the last consecutive 8th bit as a general parity bit. Loosing the last 8th bit makes then 8x7bit word or 56 bits. There there are other short cuts due to tandberg own implementation.

How Old is CSA? I do not know for sur I acepts kebiens word about 200o or so. To me all started at the end of the 70s beginning of 80's just at concept level on a desktop with pencil and paper. So to me has past 40 years. But it is unimportant. It was a hardware thing that for some miscalculation a Software Version got release. They did not thought that hackers could reverse engineer the code. And 1rst leak was release.

What are the libs,
I like the one with the header

libdvbcsa, ffdecsa, DES cscrypt Copyright (C) 1995-1996 Eric Young ([email protected])

There are other FREE libraries build for Linux crypt. And I bet you the officials release by NSA on Linux or the kernels.

As to why some people use libdvbcsa I guess because BitSlice has become popular. And libdvbcsa has that capability. But who know what is in the mind of the users. I test them ALLLLL. ALL I can get my hand on. As Gpu programing is like been Lost in Space. You never know what will be good and what will be slow like a turtle.

You so my comment on how your reee-question of same thing makes the people that try to help you unconfortable. You do seems to be a nice guy. Be smart try to stop that bad behavious and you will end up getting more help.

I guess you need to go and read for your self a bit. That will make you smarter and litle less unbeliever.

There is a GOOD friend here that post from time to time. That guy does the oposite. he reads a lots. So his post here are just links to where he find his good reading. His name is Tigilito.. he eill do like this


h__ps://en.wikipedia.org/wiki/DVB-CSA
h__ps://en.wikipedia.org/wiki/Common_Scrambling_Algorithm
h__ps://web.archive.org/web/20040903151642/http://csa.irde.to/
h__ps://www.cdc.informatik.tu-darmstadt.de/~jwaelde/breaking-dvbcsa.pdf

There you can start your self-learning experience

You see how you get people confuse. I go into the Fight and forget my main porpoise.

Lets forget about a project to have fun.

My Main objective is to show you and the other guys to stop thinking on creating a Faster Update for a library on CPU. If you are interested on building your own faster version CPU is not the Hardware.

You need to jump into the GPU wagon. And in GPU you use Cuda or OpenCL. Both use limited version of C/C++. In general in GPU you get atleast a 10X factor but with a Goof GPU it is common to get 100X factore if you get smart and limit the use of memory access.

So write code that do not or limit sending or receiving data from GPU. Try Not to use Arrays so that you only uses the private registers. I know almost impossible. Uses the smallest of program as possible.

In general there are 3 types of memory. private, local and global. Each level give you a delay. In general it been use as reference that each level of memory give you a 10x delay. So if you use private you are in 1x no delay. When you use Local you get 10X. Sadly Arrays are save in Local memory. If you use Global you get 100x delay.

And this delays are getting worst by the minute by the clashes of different cores trying to read the same memory at Local or Global Memory!!! They end up having to wait fro the previous to get a chance to read them self.

Ups I was about to start a GPU class. So Forget about DLLS, C++ or C. well C will always be the smart way. You can do anything with C.

To go back yo Objective see the sample provided. 1000 days in CPU vs 10 days in GPU. 100X factor. Good luck.

Martin.Wigston said:

That dosent take into account how long it takes to get crypt8, it can take a lot longer to get a valid c8 and sometimes you cant get.

Click to expand...

Ok,valid point,but will also be valid for the guy trying to make it faster,right? you go and tell him.
He will need a crypt8 too,or he must find a very new ingenious way to find the key.
Meaning he will need data to compare to also.
Basically,all the time spent in searching is by comparing some packet data to a known data until it decrypts correctly and matches,the faster the data is compared,the faster you find the key.
So far,nothing beats rainbow tables at this task.