Yeah, thanks, that is another option I also thought about, but it would be clunky to automate (unless I'm just using a few symbols, as you suggest), and speed of processing large ASCII sets definitely suffers vs. using native WL binary data. Thanks for throwing that out as a possible solution.