What are the advantages of the different methods to Edit In Photoshop?

[clee01l]

Operating System

1. macOS 26 Tahoe

Instead of editing directly in Topaz, I find myself using Photoshop (and its layer ability) to invoke Topaz PhotoAI.
My choices are to :

1. "Edit In Photoshop". Lightroom creates a derivative TIFF image and passes that to PhotoShop.
2. "Open As A Smart Object Photoshop". Lightroom creates a Smart Object and passes that to PhotoShop. The derivative image file is not created until Photoshop saves it.

Is one method preferred over the other? If so, why? I've been using the Smart Object method mostly because I need PS to do other things to my image after I invoke Topaz Photo AI. And I can end up with a layered TIFF preserving the PS edits and the Topaz Edits separately if I desire

As a Side note, I find the Topaz Wonder process and SuperFocus process to be very good at enhancing low and medium resolution images

 

[Johan Elzenga]

There are advantages and disadvantages. The advantage of opening as a smart object is that the raw edits are still non-destructive. Double-click the smart object and it opens in Camera Raw with the Lightroom adjustments set, but adjustable. That means you can apply filters like the Adaptive Wide Angle filter while keeping the raw file adjustable.

The disadvantage of a smart object is that you cannot change its pixels directly. If you want to use remove, or clone, then you can only do that by adding a layer on top of the smart object that contains the new pixels. But that implies that making adjustments to the smart object in Camera Raw after you added those pixels, means you may have to re-do this, because they no longer match with the rest of the image.

 

[johnbeardy]

Flexibility is the smart object method's big advantage. As Johan says, you can go back and change the raw adjustments in the ACR dialog, and you can apply PS and PS plugin adjustments as smart filters which remain non-destructive. With a TIF, you can get near to the latter by right clicking the background layer and converting it to a smart object, but you can't go back into the raw data. I'm not sure it's a disadvantage to be forced to use extra layers - I might say it's a reward for good practice. TIF is more quick and dirty, smart object files more advanced.

There is an interesting and more advanced variation that has been around for years but never seems to gain traction - the linked smart object.

To try it out, create a document in PS and use File > Place Linked to point to a raw file that you know is in LR, then drag it into position and commit it. You now have a PS document linked to the raw file and its xmp file which record the LR adjustments (I'll attach a layer so you know what to look for). There are a couple of tricks, but that's all you need to know.

What I like is that when I adjust the raw file in Lightroom and save its xmp, PS will update the linked smart object. For example, I might notice a dust spot I had missed, so I fix it in LR and save the xmp. When I open the file containing the linked smart object in PS, the updated xmp is read and updates the image. If the document is already open in PS when I make the LR edit and save the xmp, all I need to do is tab over to PS.

While there are some disadvantages, I feel as if this is the way Adobe should have done LR->PS smart objects, and I've asked a few times over the years for it to be added. It's high on my LR wish list - almost as high as Edit With being able to call OS actions!