well, it's the prefix as I already told you. The speed with -m 20 will be slower of course, because salted.
We actually also have a FAQ and some questions like this https://hashcat.net/faq/morework
just to make sure you understand the problem.
There are of course some other things you could try:
like using maskrocessor mp64.exe instead of the --stdout mode
where my.rule contains the output of:
note: replace the quote '' with "" depending on your operating system and shell escaping rules $xyz normally are variables and need to be quoted
you could also test to a command similar to the --stdout command from above (but without hashcat at the left side and without --stdout) and use maskprocessor on the left side without rules.
you could also test to just use rules for the whole command
where part1.rule contains e.g. the appending of "example" (^e ^l ^p ^m ^a ^x ^e) and some of the random chars e.g
and the dict.txt contains the second part
or use -a 1 with 2 dicts etc etc etc
but my guess is that the mask with -m 20 will be the fastest... and it really is very fast (just like not exactly the same speed as -m 0 of course)
I also noticed that you said -1 but always used ?a, that makes absolutely no sense
We actually also have a FAQ and some questions like this https://hashcat.net/faq/morework
just to make sure you understand the problem.
There are of course some other things you could try:
like using maskrocessor mp64.exe instead of the --stdout mode
Code:
mp64 -1 ?l?d.-_ example?1?1 | hashcat64.exe -m 0 -a 0 -w 3 -O -o output.txt -r my.rule 120b8510fd7091bc12f890433f65036f
where my.rule contains the output of:
Code:
mp64 -1 ?l?d.-_ '$?1$?1$@$g$m$a$i$l$.$c$o$m' > my.rule
you could also test to a command similar to the --stdout command from above (but without hashcat at the left side and without --stdout) and use maskprocessor on the left side without rules.
you could also test to just use rules for the whole command
Code:
hashcat64.exe -m 0 -a 3 -w 3 -O -1 ?l?d.-_ -o output.txt -r part1.rule 120b8510fd7091bc12f890433f65036f dict.txt
where part1.rule contains e.g. the appending of "example" (^e ^l ^p ^m ^a ^x ^e) and some of the random chars e.g
Code:
mp64 -1 ?l?d.-_ '^e^l^p^m^a^x^e$?1$?1' > part1.rule
and the dict.txt contains the second part
Code:
mp64 -1 ?l?d.-_ '?1?1@gmail.com' > dict.txt
or use -a 1 with 2 dicts etc etc etc
but my guess is that the mask with -m 20 will be the fastest... and it really is very fast (just like not exactly the same speed as -m 0 of course)
I also noticed that you said -1 but always used ?a, that makes absolutely no sense