1) To process a RAW ﬁle into RGB for further editing or printing in Photoshop
2) To change the processing of a RAW Smart Object as a layer in Photoshop
3) As a filter on any layer (Smart Object or otherwise) or a selection in Photoshop
Let’s start with Case #1: processing a RAW ﬁle into RGB. This is the base case, and it’s the most important because the RAW ﬁle contains all the original information from the capture, with the data stored in its most pure form. To view or use the RAW image in any way, it must be transformed into an entirely different format, usually RGB (Fig. 1).
There can be many interpretations of any given RAW photo into RGB, and there’s no "right" or "perfect" rendering. It’s always an interpretation that’s pretty much a one-way street, meaning it’s a destructive process, which is why we have Smart Objects.
When you change your mind about how you want your photo to look, going back to the RAW data always gives you the most ﬂexibility. But if you’ve made transformations to the image in Photoshop or applied filters to it, going back to the RAW data can be difficult, or impossible if you’re the designer and don’t have access to the original RAW ﬁle.
Which brings us to Case #2: ACR can be used to change the processing of a RAW Smart Object in Photoshop. Once you’ve created a Smart Object layer, you can reopen ACR and edit it at anytime, starting with the original source data again, to change your interpretation (Fig. 2).