In the field of interpreting (not just sign language), interpreters usually work from their SECOND language into their NATIVE language. The reason being that one has a larger vocabulary file in their native language than their second, making it easier to produce an equivalent message.
There are a lot of factors that go into ASL-to-English interpreting ("voicing" is interpreter jargon that doesn't really mean anything to non-interpreters).
One difficult piece is that ASL is a spatial language and English is linear. Meaning that ASL can take advantage of space and convey more than one concept at a time. English can only express one concept at a time, thus it takes more time to say something in English than it does ASL. On the other hand, ASL is a high-context langauge and English is a low-context language. This means that sometimes an interpreter has to be more narrative in ASL than English and sometimes THAT takes longer.
Also I think many interpreters are not TRULY bilingual. Some, I think, are even more like parrots. They can mimic ASL and copy what they've seen, but don't have an adequate understanding of ASL to really be able to interpret bidrectionally.
