When devising what became Smart Objects in Photoshop, we had several goals:
- Enable non-destructive transformation & filtering
- Tighten integration with Illustrator & other apps
- Facilitate “edit one, update many” parent-child links among layers (a la “symbols” in Illustrator, Flash, etc.).
The last of these is far & away the most obscure and, frankly, confusing. Photoshop doesn’t present any indication about which layers are linked to which others, and it doesn’t offer a Library panel for managing SO’s. (There’s so, so much potential yet to be unlocked here.)
In any case, Julieanne Kost uses a real-world example to demonstrate how to link Smart Objects together (simply duplicate them on the Layers panel) and how to duplicate & unlink them.
5 thoughts on “Demo: "The Difference Between Duplicating a Smart Object and Creating New Smart Object via Copy"”
I use that feature all the time. It’s very helpful when building web comps and producing similar elements, but have some different. New Smart Object via Copy works perfectly.
Agreed! So much potential.
It seems like the implementation has come along way over the past 5-6 years. Originally free transforms weren’t supported, then they were, but it was slow. With CS6, I have dozens of D800 RAW files in a 3.5GB multi-viewpoint pano and the smart object perspective transforms are happening in real time…and that on a 13″ MacBook Pro (i.e. no discrete graphics card).
It does appear that there is some smart caching going on, as it was taking 15 seconds to see the free transform changes after reopening the file and readjusting smart object layers. If, however, the smart object was opened and closed (i.e. bringing up Adobe Camera RAW and simply accepting the settings), the transform updates happened with no lag whatsoever.
It’d be great if this caching allowed free transformations on smart objects with smart filters applied (e.g. the amazing Adaptive Wide Angle filter is dropped temporarily when entering free transform mode).
And on the topic of free transform, is there anyone at Adobe I could make some feature requests on this? It’d be great to be able to pick the 4 handles defining the transformation homography? The 4 corners of a layers bounding box are typically not one wants to align with other layers or correct for key-stoning. Also, the perspective transformation mode doesn’t correctly maintain horizontal to vertical ratios (e.g. key-stoning a subject that should be a square can make it into a rectangle of uneven proportions ); I don’t know if this would require the FOV of the image to be transformed, but that information is easy to grab from the EXIF metadata.
I suppose you’re alluding to this with the “there’s so, so much potential here”, but for consistency and better understanding of the features, I’d suggest relabeling some or all of this functionality as Symbols and introduce a Library panel.
Good stuff though, I’m a big fan of Smart Objects and people tend to make fun of my, “Did you make it a Smart Object?” comments all the time.
I think you featured http://www.canlinkit.com/ on your blog a short while back, I’m using that a lot in my daily workflow.
I’d love to see a more robust way to template things using smart objects in Photoshop itself, more visible ways to link to external files would be a super sweet start!
I’m using Slicy to export graphics from my PSD’s – but Slicy does not scale smart objects the way Photoshop does, which makes them unusable in my particular workflow.
For shortcut fans – Cntrl/Cmd+Shift+j=copy of smart layer, rather than duplicate smart layer [Cntrl/Cmd+j].
I just wish you could open several smart objects into a single file in PS from LR as I use them a lot and this is the main bottleneck for myself with regard to smart objects.