thanks :)

There are a couple of things which will suggest if you can have success removing unwanted sounds.

One - Is the unwanted sound concentrated or spread across the spectrum? Wind noise is usually low, hiss is high, 60 Hz is concentrated at 60, 120, etc. Hz. You can usually remove these and still have usable sound. Buzz from a fan or other machinery is usually wide spectrum and if you remove it you destroy your soundtrack. Voice covers a pretty wide range in the heart of your spectrum, so removing voice doesn't leave much.

Two - What percentage of your spectrum are you trying to remove? If you have a rock song where there is someone wailing unintelligible mumble under thundering music, you can remove the vocal and hardly tell is is missing. If you have a strong singer belting out a Ballard with a sweet accompaniment behind and you remove the singer, you won't have anything left.

Some people believe in the 25 percent rule. If you try to remove more than 25 percent of your spectrum you are likely to end up with a mess unless the noise is highly concentrated... such as wind noise or hiss.

You can only find out be trying.

I would 'sample' the background noise from the original audio, and then cut out the voice and replace it with copies of the sample. If your recording is long, there will go some substantial handwork in this (and time), but you'll get the best result.

Is the camera recording 5.1 audio?
If not, the best you can do is sample-and-loop or simulate the ambience with a processing fx.
If you have a long section of background and loop judiciously, you may have pretty good luck.