Hi @Stream_biofilm,
Yes, that is one route for doing this. The other would be to use taxa-based filtering if there's a short list of taxa that you'd like to exclude. For filtering all features found in a specific sample, generating a list of features as a metadata file is still what you need to do.
That might not be advisable, as many features (particularly if they are abundant) could represent cross-contamination from real samples; check out the notes in this thread.
I hope that helps!