I think may be that all smart object layer have pixels rendered for the layer pixels by Photoshop for the layers content these pixels can not be changed by Photoshop tools only use by the tools to change the layers composite view. Every smart object layer has a associated transform to position and size the layers composite. You can have several Smart object the share a common object. Each layers transform position and size the object over a different area of the document canvas to make a Picture Package, It may be if you reopen the raw object in ACR and change the size of the object that layers associated trans may need to be adjusted to work with the changed object size. That the transform associated with original object no longer works the way you want. The reopened RAW object Crop in ACR has no effect on the current document canvas size. That Is What you are seeing.
Thanks for the reply JJ,
I've read and re-read several times though and dont quite understand what you are saying.
Looking at your screengrabs though, you appear to have 3 copies of the one smart object? I understand if you crop one they will all be cropped.
I have a document with several layers, all different images/smart objects but all the same dimensions (6016x4016 px). If I open one of these images and crop in to a small portion of that canvas size no matter where it lies within the frame it is by default placed within the centre of the existing Photoshop document. I'm hoping to have it "Paste in Place" where it should lie within those initial dimensions (6016x4016 px).
I'm using this process to swap elements of images out and understand that masks would give the desired effect but the issue comes in when I am wanting to dynamically rotate the smart object at that zoom range. Without cropping the smart object I have no anchor points to work with and am forced to zoom back out to the PSD canvas boundry, making it difficult to judge my rotations.
Hope this makes sense
It is like I wrote. You started a smart with a smart object layer in your composite. The Smart object you state is a file. that open into Photoshop through ACR, So You placed in the image into your document. During the place operation ACR converted files into a RGB Image with the ACR Adjustment you made in ACR.. If you did not crop the image ACR would have passes the full size image to Photoshop for the document. As part of the Place operation Photoshop creates a Transform for the layers pixels to fit the Image within the documents canvas size it it is to be resized during Place. If it is larger than the canvas the transform scales the image. If it is not larger than the canvas the transform does not scale the image it is just positioned on the canvas full size over the canvas which may cover the entire canvas. When you reopen the object in ACR you are not redoing the Place operation you are just redoing the ACR part of the operation. ACR is replacing the contents or the Current Object. ACR need to replace the object with an object the same size as the object it is replacing for the layer transform to work like it did with the object being replaced. I show where I place in a RAW file into a large canvas the duplicated the smart object layer twice and changed the the associate layer transforms to position and size the images to fill the large canvas. I then replace the object image with a smaller cropped image The transforms no longer work to fill the canvas for there were fewer pixels in the object.