Smart objects

[unfocused]

I thought I would kick this off with a tribute to one of my favorite features: "smart objects."

Since first reading about this technique in a Scott Kelby book a few years back, I've incorporated this in my standard workflow.

Basically, I make one set of adjustments to each image using Adobe Camera Raw to optimize whatever portion of the image I want to begin with. For example, I may adjust the image to maximize a subject's face. Then, I bring it in to Photoshop as a smart object.

Make a copy as a smart object and then I have two layers that can now be adjusted independent of one another in RAW. I'll double click on the smart object to return to RAW and work on another section of the image. Let's say I want the subject's clothing a half-stop darker than their face, or I want to bring the background down a little, or I want to give the sky a different exposure than the foreground.

I make those adjustments on the second smart object and then when finished send it back to photoshop.

Now, I have two (sometimes more) layers, each optimized for a certain area of the image. It's now simply a matter of adding a mask to the top layer/layers and painting out or in the areas I want in order to create a single image.

Because I'm a real paranoid about saving steps, I'll usually preserve all these smart object layers by doing a copy and merge-layers. Ending up with a single layer that has all my adjustments (no longer a smart object) while still having all the various smart object layers in the file in case I need to go back and re-do something.

I'm wondering: Do others use smart objects and how do you use them?

 

[Mt Spokane Photography]

I'll have to admit that I do not know what a smart object is.

What is the advantage? I usually select that option when going from LR to PS, but I don't know what it does.

 

[unfocused]

Honestly, I'm not sure I know exactly what it is. Basically though, a smart object retains all the underlying data, so that when you manipulate it (change sizes, etc.) you retain the original information. I use them almost exclusively in connection with RAW editing.

Because I can make several duplicate layers that are each smart objects, I'm able to return each layer to RAW simply by double-clicking on it and then adjust that layer.

People on this forum often talk about dynamic range. Well, a smart object is one way to expand your dynamic range without using HDR. Let's say I have an image where the main subject is properly exposed, but I feel like I just need an extra stop of exposure on, say, a rock. I can take one smart object layer and in RAW reduce the exposure by a stop and then, when I return to photoshop, I use a layer mask to paint in the rock.

I would say that what I like about the process is that it allows a lot of fine-tuning of exposures and by using a soft brush and a layer mask, it's fairly easy to make local adjustments to portions of an image. And, because I am able to return each layer to RAW, the adjustments are non-destructive and easily fine-tuned.

 

[privatebydesign]

No, Smart Objects don't give you any more DR, they just give you more control over the DR you have.

They get very complicated when you start nesting Smart Objects and there are many limitations to using them in PS, but if you are not sure how you want to work a file, or if you want to play with multiple RAW edits it is very effective.

PS CC has ACR as a filter, and the truth is it works better and faster than multiple Smart Objects layers most of the time.

 

[Tabor Warren Photography]

I use smart objects almost daily.

Though, when I use them, it is typically for nothing more than resizing the image for a certain appeal. I simply open the photo, right click the background layer, select convert to smart object, and bam! resize away. What is nice about using them this way is that I can shrink it down immensely, click okay, change my mind, blow it back up, and it doesn't look all pixelated. It's great when designing layouts for albums as well.

Cheers,
-Tabor

 

[Mt Spokane Photography]

Tell me More!

I actually took a Photoshop course a couple of years ago, (Maybe Longer) but we were at ver 5 then. I used LR, but the instructor did not. If he talked about smart objects, I missed it. Of course, we did a lot with layers, so I might have just missed the term. He is a local Photographer who worked for one of the Local Theaters, and since I photograph Stage and Theater, we had a lot in common to talk about.

 

[climber]

I usually open image from LR to PS using Edit In --> Open as smart object in photoshop. Then, when it is already in PS, if I double click on the layer (smart object) it opens in Camera Raw with all sliders in the same place as when I leave Lightroom. Actually, double click brings you back to Lightroom adjustments with all previously adjustments made already there. Here you can fix something (let say noise reduction) and click OK. It brings you back to PS with that adjustment, but all of your work in PS is also there. Let say, if you have several adjustment layers above.

There is option, if you bring image from LR to PS using Edit -> Edit in Adobe Photoshop. That way, the image will open as a normal layer. You can also convert it to Smart Object in photoshop by right click on the layer _-> convert to smart object. And now if you double click to that smart object, it also brings you to Camera Raw. But in that case, all sliders will be set at zero (0). In the first option (when opening as a smart object from LR), sliders will be at same position as in LR.

Smart Objects are also very useful when using (Smart) Filters. Let say Gaussian Blur or something like that. If you apply it to the normal layer, it is applied and that's it. You can't go back and readjust it. If using filter on a Smart Object, you will get a small icon below layer. Now you can double click on that icon and readjust your filter amount. Let say blurring. In that way you can add as many filters as you want. And all filters will have a mask already there. Now, if you don't want that blurring in some area, you simply mask it out. The only drawback is, that if you apply more than one filter to smart object, all filters will have the same mask. So, if you mask out something, all the applied filters will be masked.

I also use Smart Object in the next way. Let say, I have several adjustment layers (or any other layers) above background layer. At some point I decided to go to LAB mode. To preserve all my work non-destructively, I simply mark all layers -> right click somewhere on them and click to Convert to smart object. Now, it will appear as a single layer. And now I can turn it to LAB mode. If I would turn it without merging it into one smart object, the adjustment layers would be discard. And when I have that merged Smart Object already converted to LAB, I can also double click on it. It opens as a new document, with all the layers there and in RGB mode. Here I can do some more adjustment if I want and save it. The document in LAB mode will automatically update.

Smart Object are also useful, if you have some kind of a template. Let say, you have in one image several smaller images and a text. If all images are smart objects and you want to change an image with another one, you can simply right click on it and choose Replace Contents. It opens your browser, where you can find that new image and replace the older one. All the filters and layer styles like drop shadow, stroke, etc. will be remained. Only the image will be replaced.

Hope I help a bit. There is a lot of tutorials on the web, how to use it.

 

[climber]

Here are few short tutorials about smart objects:

http://tv.adobe.com/watch/learn-photoshop-cc/making-nondestructive-transformations-with-smart-objects/

http://tv.adobe.com/watch/learn-photoshop-cc/opening-an-image-as-a-smart-object/

http://tv.adobe.com/watch/learn-photoshop-cc/linked-smart-objects-in-photoshop/

http://tv.adobe.com/watch/learn-photoshop-cc/smart-filters-and-the-filter-gallery/

 

[plam_1980]

I have watched several videos from Joel Grimes and also a course of his on Creative Live, he is using Smart Object more or less as you descrive your workflow. Here is his blog

http://www.joelgrimesworkshops.com/Joel_Grimes_Photography/Welcome.html

there are some videos from him on YouTube too and you can also see this one, which is inspired by him as well

http://tv.adobe.com/watch/russell-brown-at-photoshop-world-2011-orlando/smart-object-blend-mode-portrait-techniques/

 

[Besisika]

I'm wondering: Do others use smart objects and how do you use them?

Yes, I do. Actually, on every single photo I work on. You have an option in ACR that will open all photos as smart object automatically.
I don't use multiple copies to change the "brightness" as described. I use an ajustment brush for that. Smaller size I guess.
But I do use it when correcting multiple lighting temperatures (white balance). Example; you are in door using a flash and part of the picture has a window lighting and part has tungsteen in the back.
You focus on the main subject first, bring it to PS, right-click to create a smart object, double click and ajdust for other part.
Once you have all layers, you use a soft brush to bring each layer up.

 

[unfocused]

No, Smart Objects don't give you any more DR, they just give you more control over the DR you have.

Yes, that's exactly correct and I probably didn't explain myself very well.

A better example: earlier this week I shot a group of nine people for a company's website. It was a diverse group of people and by using multiple smart object layers I could do some fine tuning of individual skin tones and exposures. For example, one woman was very tan, so I might lighten her up a bit. A couple people were very fair-skinned, so I brought them down a hair.

Using layer masks I can blend the various layers and achieve a final group image where exposure has been optimized for everyone, but it is subtle and not apparent.

As with most things Photoshop, there are lots of ways to arrive at the same place. Being a bit of a fanatic about non-destructive editing, I find the Smart Object route works for me.

 

[dppaskewitz]

Does anyone know: do smart objects work the same way as editing in LR, where the edits are commands stored in a separate XMP file so that your underlying RAW file is never changed? Sounds like Adobe took that from LR and built it into PS?

I don't edit in PS very often (I have the CC version but really haven't started to learn how to use it). My workflow (if I can dignify what I do with such a term) is to import everything into LR, figure out which photos I want to work on, edit in LR, then if I think PS would be useful (mostly so far only for content aware fill to get rid of stuff) I use "edit in PS" and create a Tiff from the photo (with the LR edits) for that purpose. Sound like I should edit the LR adjusted photo as a smart object in PS. If I understand correctly, that would just be adding further commands to the XMP file and I wouldn't end up with so many versions of the same photo. Does this make sense to anyone?

OK, I have some tutorials to watch. Thanks for the links.

 

[pdirestajr]

Does anyone know: do smart objects work the same way as editing in LR, where the edits are commands stored in a separate XMP file so that your underlying RAW file is never changed? Sounds like Adobe took that from LR and built it into PS?

I don't edit in PS very often (I have the CC version but really haven't started to learn how to use it). My workflow (if I can dignify what I do with such a term) is to import everything into LR, figure out which photos I want to work on, edit in LR, then if I think PS would be useful (mostly so far only for content aware fill to get rid of stuff) I use "edit in PS" and create a Tiff from the photo (with the LR edits) for that purpose. Sound like I should edit the LR adjusted photo as a smart object in PS. If I understand correctly, that would just be adding further commands to the XMP file and I wouldn't end up with so many versions of the same photo. Does this make sense to anyone?

OK, I have some tutorials to watch. Thanks for the links.

Not really. It's more like putting "containers" around the file to protect the pixels and original file. Then filters and adjustments are applied to the container file. They are very helpful when you want to do destructive edits.

But the problem is that the Photoshop file will get really big really fast, as you are basically nesting files within files.

 

[distant.star]

.
Tim Grey answered this question in his "Ask Tim Grey" newsletter today. He told me it was okay to post his answer here in its entirety. From:

AskTimGrey.com


Today's Question

Could you explain on a more technical level exactly what a Smart Object is and how it differs from a "dumb" object? This question is raised because of your remark about "de-smarting" by using Layer > Rasterize > Smart Object. So I guess it's not rasterized. But what does that mean?



Tim's Answer:

A Smart Object provides a way to reference a source image to provide a non-destructive workflow approach for various tasks within Photoshop. In most typical workflows you would have the Smart Object embedded within your master Photoshop PSD file, for example, but Photoshop also now supports linked Smart Objects, where the source image for that object is not embedded within the master image.

This differs from simple pixel values in that it is possible to return to the Smart Object and apply non-destructive changes to your original adjustments. In other words, when you're using a Smart Object you're getting benefits that are similar to the workflow benefits of using adjustment layers rather than applying adjustments directly to pixel values in your images.

However, it is important to understand the workflow ramifications of using Smart Objects, which includes both benefits and potential pitfalls. Let's address the benefits first.

Because a Smart Object is essentially a "container" for image data, Photoshop is able to perform tasks in a non-destructive way that simply references the source data in that Smart Object. Let's consider a few examples.

If the Smart Object you're using is a RAW capture, the original capture data for that RAW capture will be embedded within the document you're working with in Photoshop, but adjustments via Adobe Camera Raw do not actually affect the original RAW capture data. As a result, if you open a RAW capture as a Smart Object, you can double-click on that Smart Object at any time to return to Adobe Camera Raw and refine the adjustments you originally applied.

If you are using a Smart Object to apply Smart Filters, you'll see the same effect you would otherwise achieve by applying one or more filters, but you can return to the filter settings later and refine those settings. In other words, you are effectively applying a filter effect in a non-destructive way, as though that effect were being applied using an adjustment layer.

You can also use Smart Objects for more complex scenarios, such as when creating a composite image. By embedded another image into a document as a Smart Object, you can, for example, resize that image as many times as you want within the composite without having a cumulative negative impact on image quality, because each time the Smart Object is referencing the original image data, not modified data.

Of course, as I've mentioned in the past when addressing Smart Objects, they aren't without their pitfalls. Most significantly, using Smart Objects can create problems in the context of a layer-based workflow. For example, let's assume you open a RAW capture as a Smart Object, then perform some image cleanup work on a separate layer using the Spot Healing Brush tool. Later you decide you want to change the appearance of the image, so you double-click the Smart Object to bring up Adobe Camera Raw, and apply changes to the photo. However, you are only affecting the Smart Object, not the layer with your image cleanup work, so the pixels on your image cleanup layer no longer match the underlying photo. It is this sort of issue that causes me to not use Smart Objects for most tasks, despite the potentially significant benefits offered by Smart Objects.

As for rasterizing, that refers to the process of converting an image to actual pixel values. So, for example, in the case of a Smart Object, rasterizing means converting the embedded image data and all of the saved adjustments related to that data, and process all of the information into actual pixel values. You end up with a smaller file (in most cases) but without the benefits of the Smart Object.

 

[privatebydesign]

Tim's answer is, of course, spot on, but you can work around the limitations, you just need to have a clear idea of where everything is and why you are doing it.

For instance I print a fair bit for other photographers and am often asked to print the same picture in multiple aspect ratios to fit "standard" mats. To maintain maximum flexibility in the file I will nest smart objects, here is a typical workflow.
Do ACR adjustments to TIF file and then open as Smart Object, create an empty layer and then do necessary cloning of dust etc onto that, then make a print style curves layer, saturation etc and anything else the file needs to get printed well. Then select all of the layers including the original Smart Object, then convert all that to a second Smart Object, then I will do the conversions to the various aspect ratios. Doing this enables me to push and pull the file multiple times with no loss in IQ, it maintains maximum editing latitude because if I notice another dust spot I can open the second Smart Object and edit the clone layer, or I can adjust the opacity of the adjustment layers to make them editable.

Finally, I might convert all of that to another Smart Object and do Smart Sharpening to that, this gives me the ability to print to different sizes all with a completely editable output sharpening layer.

It sounds more complicated than it is, but if you edit in stages and then convert all of it to another Smart object you can overcome many of the "limitations" they present. Of course this isn't appropriate for many, or even most situations, but for my specific cases it saves me a lot of time, though file sizes that start out at 25mb can easily be pushed outside of .PSD limits and have to go to .PSB files instead, and Lightroom can't handle .PSB's which is very annoying.