Posts: 22
Threads: 1
Joined: Jul 2014
07-23-2014, 01:56 AM
(This post was last modified: 07-23-2014, 07:00 AM by SomeGuy.)
I know pretty well what my password was in trucrypt and i think i know the settings. I used something like "q W e R 1 @ 3 $ " or some iteration of that. I was away for a while and didn't use the TC container for a few months and forgot the exact combo of chars i used.
So i wanted to do something with a rule that stated these chars have to be used only once
qQwWeErRtT1!2@3#4$5%
and space could be used 5-11 times
plus password length set to be 15-19 chars. I don't plan on finishing this any time soon but i really want to start and learn how to set it up. I do plan on getting 4 of the best GPUs next year or whenever they release the next gen of GPUs on the 20/16nm process. (some debate is out if 20nm for GPUs might even be skipped since 16nm is getting closer to release at some facilities.)
Also can this be started and stopped with progressed saved or so i have to break it up like do one at 15 chars let finish and come back when i have time and do a 16 char length.
I know the chars used in my password pretty well and this seems like it'll take forever even with special knowledge -_-
Also how do i tell if i used whirlpool or SHA 512 in my trucypt container. I have used both and not sure about this container. My guess would be whirlpool because i thought that was the best at the time from what i read. So does that matter? If i understand, i do need to know which it is to crack it.
I also used AES twofish serpent as encryption if that matters. I also assume XTU or whatever that is was used. I used the best settings in everything i could click IIRC.
Thanks!
EDIT: i think i want a rule based attack but i dont get how i create a rule that only uses those 20 characters once and uses a spaces X-X times. Along with keeping it in ____ char limit
Posts: 5,185
Threads: 230
Joined: Apr 2010
I also recommend permute.bin in case you know the chars/letters but not the ordering - that also includes the space. Then pipe it to oclHashcat.
Posts: 22
Threads: 1
Joined: Jul 2014
the ordering is qwert12345 but all those chars may not be used and the space can be anywhere. Is there a way to add that in too? Also can you elaborate what your talking about?
Posts: 76
Threads: 8
Joined: Feb 2013
Posts: 22
Threads: 1
Joined: Jul 2014
07-24-2014, 05:01 AM
(This post was last modified: 07-24-2014, 11:36 AM by SomeGuy.)
UPDATE!!!
ok so i put in permute
"qQwWeErRtT1!2@3#4$5% " and it will create a massive dictionary of all combinations? Won't that file be huge? Wouldn't it be easier if i made rules? Also does it have the option to allow chars to only be used once minus the space? That reduces the possible combinations by orders of magnitude.
EDIT: well i think i get what your saying. "qQwWeErRtT1!2@3#4$5% space space space space space space space space" is what i put in the permute.exe or whatever thing and with my 7970m (7870) it appears it'll take an unholy amount of time. Is there a way to force it to only compute the order i need? That reduces time by a huge amount. Also is there a way i can limit permute.exe to 14-19 chars or something?
ummmm...it appears to be adding chars i never enters like H and o??? :/
Posts: 18
Threads: 1
Joined: Jul 2014
07-24-2014, 12:37 PM
(This post was last modified: 07-24-2014, 12:56 PM by mkcon.)
My first idea was to define a lot of user charsets:
1. charset q, Q and the "empty char" (is this possible at all?)
2. charset w, W and the "empty char"
and so on, and use them as a mask ?1?2?3?4....
but this idea will sooner or later overflow oclhashcat's capabilities.
(You need 10 charsets for the chars and one additional for "blank or nothing".)
So i wrote a small script, which stupidly does exactly the same.
Result (for at most one blank between the chars and password length >= 15):
dictionary with about 7.3 million entries.
Thus it should be practical.
Posts: 22
Threads: 1
Joined: Jul 2014
(07-24-2014, 12:37 PM)mkcon Wrote: My first idea was to define a lot of user charsets:
1. charset q, Q and the "empty char" (is this possible at all?)
2. charset w, W and the "empty char"
and so on, and use them as a mask ?1?2?3?4....
but this idea will sooner or later overflow oclhashcat's capabilities.
(You need 10 charsets for the chars and one additional for "blank or nothing".)
So i wrote a small script, which stupidly does exactly the same.
Result (for at most one blank between the chars and password length >= 15):
dictionary with about 7.3 million entries.
Thus it should be practical.
what do you mena you wrote a script?
Posts: 18
Threads: 1
Joined: Jul 2014
(07-24-2014, 01:32 PM)SomeGuy Wrote: what do you mena you wrote a script?
I meant - i wrote a script (i.e. a small program that does the job of wordlist generation).
If you send me the first sector (512 bytes) of your container per pm, i will test my dictionary for you.
As an alternative i could send you the gzipped dictionary - but that's more than 15 MB, so email is probably not an option.
Posts: 22
Threads: 1
Joined: Jul 2014
07-24-2014, 02:33 PM
(This post was last modified: 07-24-2014, 04:25 PM by SomeGuy.)
(07-24-2014, 02:08 PM)mkcon Wrote: (07-24-2014, 01:32 PM)SomeGuy Wrote: what do you mena you wrote a script?
I meant - i wrote a script (i.e. a small program that does the job of wordlist generation).
If you send me the first sector (512 bytes) of your container per pm, i will test my dictionary for you.
As an alternative i could send you the gzipped dictionary - but that's more than 15 MB, so email is probably not an option.
so the dictionary you made used all the chars i stated and with/without spaces after each char? and 14-19 char long?
use drop box? email would work but rather not give it out
Reason why i ask because the password could be like this. It may be clusters like the first one and may or may not have a space at the end.
"q WE r !@ 3 $%"
"q W E r ! @ 3 $ %"
"q W e R 1 @ 3 $ 5 "
So this password i made up from using a pattern on the keyboard that was easy to remember...when i actually used the file regularly. It appears to not be susceptible to any normal dictionary attack. Without knowledge of the pattern i used the only way an attacker could break it would be pure brute force correct? Even with knowing also the entire pattern/password it still seems ridiculous -_-
Also if i used whirlpool wash...was that the best choice because it appears SHA 512 is easier to crack according to the benchmarks listed on site. I have switched to SHA512 in my recent stuff but should i switch back?
Posts: 18
Threads: 1
Joined: Jul 2014
(07-24-2014, 02:33 PM)SomeGuy Wrote: so the dictionary you made used all the chars i stated and with/without spaces after each char? and 14-19 char long?
Oops, i used limit of 15 chars, ok. redone.
(07-24-2014, 02:33 PM)SomeGuy Wrote: Reason why i ask because the password could be like this. It may be clusters like the first one and may or may not have a space at the end.
"WE r !@ 3 $%"
"q W E r ! @ 3 $ %"
"q W e R 1 @ 3 $ 5 "
Should be inside - even with blank at the beginning.
Download from here:
https://www.sendspace.com/file/ihtrzj
The file is a gzipped textfile with unix style end of line convention.
Please tell me, when you downloaded it. I will delete the file then.
It would be nice to tell me if you had success.
Keep in mind that oclHashcat up to now basically supports AES only (no cascades, no other ciphers).
Good luck!