hybrid attack Netntlm too slow

[Marsupilami]

this sounds very strange and I'm pretty sure that it doesn't behave like you think it behaves.

How would the second process even know when the "second" word from possibilities.txt begins. It has no information about the words. The second command only runs everything it get's as input.

I think there are many ways to troubleshoot this. e.g. using a "cat" or "type" command as the first part of your pipe, instead of the stdout command.

you should also test with the newest beta version of hashcat from https://hashcat.net/beta just to make sure and probably also test older versions from https://hashcat.net/hashcat

you should also look carefully at your status prompt (hashcat output). maybe you have a lot of rejected password candidates (there is a line for that too)?

I think it has something to do with the stdout command. When i remove it and just pipe on cat i get the error "Token length exeption". On beta version i have the same behaviour. Its difficult for me to split the commands.. cause the a6 (hybrid attack) expects a hashfile I get the Token length expetion. But I want to start the cracking process in the second part after the pipe for creating more work and the first pipe just should just create the candidates. 
I did tries with prince and maskprocessor but unfortunately it dont work.
My thought was that the princeprocessor reads in the dictionary and the maskprocessor adds the remaining characters and then hashcat uses the result as input for cracking.
pp64.exe possibilities.txt | mp64.exe ?a?a?a?a?a?a?a | cat
I would expect something like
M%RFOFtafDK!%A but when i cat the result i just get the output of mp64.exe tafDK!%A