Local Adjustments made in ACR from PS -- How does this work?

[Thlayle]

I am new to PS/LR in the last couple of years and I am using the CC versions. I am getting much more comfortable with PS and turning to it for more and more of my projects. A recent experience with ACR seemed to generate a lot of extra work for me. Maybe I just didn't understand what was happening & how to deal with it.

When I make local adjustments in LR, I can go back and forth all the time with any given photo and redo any of the adjustments (including gradient filters, radial filters & brush adjustments). No matter how many times I do this, I can access those local adjustments I already made, adding or subtracting any amount of the adjustments. Such is, I understand the nature of the parametric editing in LR.

When I do the same in ACR as a filter in Photoshop, working from a PSD file transferred over from Lightroom (fresh image, no adjustments made yet in LR, let's say in this question/example), I find that after I make the same local adjustments that I would have made in LR, the pins for the adjustments are not any longer accessible once I close ACR and then go back in later. They simply don't show up. So...if I made an errors that I want to correct, it seems I would have to reset the ACR filter. I can't just access a particular adjustment and modify it. I can add more local adjustments or reset, but I can't seem to make any changes to the local adjustments I already did.

The image I was working on had some posterization in a small section of the sky. I ended up going back into LR to fix it (essentially starting all over). New local adjustments in ACR wouldn't do the trick.

Is this how it is supposed to work? Does using smart objects avoid this problem. Am I missing something?

[Dave Humphries]

Does using smart objects avoid this problem?

I believe it might, but I've never tried it in anger.

I'm sure someone will be along that knows the definitive answer - I have re-titled the thread to be more descriptive of the issue to attract them.

Cheers, Dave

[Loose Canon]

Hi Randy!

I’m not entirely sure what you are asking here or what you are looking to do. Further I don’t use LR. But what the heck? I’ll give it a shot and see if it helps at all!

If you are working on a file in PS and you want to use the Camera Raw filter, duplicate the layer you are working on, convert that layer to a Smart Object, then use the CR filter. You should then be able to get back to Camera Raw on that layer to tweak your previous adjustments.

Keep in mind that not all options of ACR are available when using it as a filter in PS.

[Manfred M]

Okay Randy - I'm pretty sure I understand what is happening here, so let me try to have a crack at explaining it.

First of all both ACR and the Develop Module in Lightroom ARE IDENTICAL, other than the user interface. They are both Parametric Editors and no actual pixel based data is changed. They rely on storing parameters to formulae in either a sidecar (*.xmp) file or in a Lightroom Catalogue. As long as you only work in one of these two editors, all changes are reversible.

Once you open the data in Photoshop, you are working in a pixel based editor, and at a high level the software makes physical and destructive changes to the data (i.e. other than the undo or stepping back in the history window), the changes are locked in. This information when saved is written to a Photoshop Document (*.psd) file and YOU CANNOT MAKE ANY CHANGES IN EITHER LIGHTROOM OR ACR. Opening a file in Photoshop is a one-way street. Any of the changes you made to the original file in either Lightroom or ACR still there at the point you left them at.

Photoshop has ACR-like functionality in the Camera Raw filter. It looks a lot like camera raw, but isn't. The sliders are in the "neutral" position and any edits done are incremental to whatever has already been done in ACR / Lightroom + Photoshop. This, like any other Photoshop edit, is pixel based and destructive.

Except for one thing. If the layer you are working on is a "Smart Object", the filters act a lot more like what you are used to from ACR / Lightroom and all the filters (not just the Camera Raw filter) remember the settings and you can go back and change them. The downside to the Smart Objects are that they cannot be edited on a pixel basis, so the clone tool, fill tools, brushes, etc. do not work on them. To use these tools you have to rasterize the layer (and lose the reversibility of using Smart Objects), but they you can edit them. What I do is to make a copy of the Smart Object and then will rasterize it. That way I still have all those edits and don't have to go all the way back if I need to make changes to one of the filters. The only downside is that the *.psd file gets larger as you add layers.

Confusing? Yes, until you work this way and understand how this all hangs together.

I hope this makes sense to you.

[Dave Humphries]

Photoshop has ACR-like functionality in the Camera Raw filter. It looks a lot like camera raw, but isn't. The sliders are in the "neutral" position and any edits done are incremental to whatever has already been done in ACR / Lightroom + Photoshop. This, like any other Photoshop edit, is pixel based and destructive.

Except for one thing. If the layer you are working on is a "Smart Object", the filters act a lot more like what you are used to from ACR / Lightroom and all the filters (not just the Camera Raw filter) remember the settings and you can go back and change them. The downside to the Smart Objects are that they cannot be edited on a pixel basis, so the clone tool, fill tools, brushes, etc. do not work on them. To use these tools you have to rasterize the layer (and lose the reversibility of using Smart Objects), but they you can edit them. What I do is to make a copy of the Smart Object and then will rasterize it. That way I still have all those edits and don't have to go all the way back if I need to make changes to one of the filters. The only downside is that the *.psd file gets larger as you add layers.

Thanks Manfred, that's what I had found by experimentation as the best way to work (within the limitations), but wasn't sure enough to say.
Unfortunately it's not quite a case of 'having your cake and being able to eat it'

I have recently started opening in PS from ACR as Smart Objects, but usually the first thing I want to do is pixel based, so I layer copy and rasterize (get the huge file), then do my cloning, etc.

To date; I have yet to find a significant benefit in the use of Open (smart) Object over Open Image. The only times I have needed to re-open an image has (by chance) also required me to go right back to the ACR stage, which I could just as easily achieved by re-opening the .nef file in ACR where the sidecar .xmp file retained the work done in ACR.

Arguably, I could 'Open Image' into PS, do any cloning, then create a new layer and convert that to a Smart Object to get the benefit of further work with ACR filters, if necessary.

This is probably an area I need more training in, something for me to look for in Lynda.com, since I pay for it anyway.

Cheers, Dave

[Loose Canon]

To date; I have yet to find a significant benefit in the use of Open (smart) Object over Open Image. The only times I have needed to re-open an image has (by chance) also required me to go right back to the ACR stage, which I could just as easily achieved by re-opening the .nef file in ACR where the sidecar .xmp file retained the work done in ACR.

For my purposes Dave, opening from ACR to PS as Smart Object is primarily a time-saver if further adjustments are required from RAW conversion. Mostly if I’m looking to work with the photo’s color attributes first thing after conversion, which is fairly rare. But it allows one to work from PS without having to close the PS file and re-open the original RAW file in ACR and starting almost from scratch. Again saving time and steps. If I'm fairly confident with the conversion I won't bother with it.

But…

Arguably, I could 'Open Image' into PS, do any cloning, then create a new layer and convert that to a Smart Object to get the benefit of further work with ACR filters, if necessary.

This is where the CR filter shines, why it was added, and why it was added as a Smart Filter option. I use this quite a bit actually and allows access to adjustments that are either not available in PS or are much easier and accurate to perform in ACR. Say for example one or two colors were pushed too hard in an LCE adjustment, easy fix with a CR filter layer (HSL). Not to mention things like Clarity, and an additional possible B&W conversion technique are available (Grayscale with its Grayscale mix), Lens Corrections (distortion), dehaze, etc. Plus as a Smart Layer (Filter) it comes with a layer mask as well (of course if not used as a Smart Filter you can always add one to the layer anyway).

Typically, the CR filter is something I use in later stage production and that I consider one of the all-tyme best additions of PS! It may not be something that everyone needs all the time but sure is nice that its there. When it came out and I started taking advantage I quickly found I had a lot of use for it and saved a lot of time and effort!

[Thlayle]

Thank you both, Manfred & Dave: very helpful comments. A lot of confirmation of what I believed to be a "learning the hard way" lesson I just recently had in using the ACR filter in Photoshop.

And thanks also, Terry, that also address part of my question.

And Dave, I note from your, "I've never tried it in anger" comment that it was not hard to pick up on my level of frustration with my recent mishap in ACR.

Manfred, yes: all very confusing but I believe I've "got it" now. Certainly I was confused when working with the ACR filter in Photoshop. I was operating under the misguided belief that I could use it and all of the controls within it in exactly the same way I operated in Lightroom. NOT so!

As for the limitations of smart objects and pixel based editing, I believe I am following you when you say the image needs to be rasterized first, although I can't think of where/how that is done. I do know this: I can double-click on my smart object layer and then it opens the original image for me to do things like cloning & so forth. But then before I close the image, I have to save it, and the save is done to the original file, the CR2 file carried over from Lightroom (even though I chose in Lightroom to open such files as PSD by default in Photoshop for editing purposes). This makes me wonder if your comment about "rasterizing" the image points to a different option that might work better than opening/saving the original file in the manner I just described.

I am starting to warm up to the idea of using smart objects. Right now, I am still having trouble being patient whenever I click on a smart object filter, turning it off and then back on for comparison, and then having to wait for what seems like a terribly long time as the filter is re-rendered.

Thanks again. This really helps me understand the ACR filter in Photoshop better. I really had it wrong when I was thinking it would continually re-edit with out the "one way" changes hitting me.

Regards, Randy

[Loose Canon]

As for the limitations of smart objects and pixel based editing, I believe I am following you when you say the image needs to be rasterized first, although I can't think of where/how that is done.

If you try to use, say, the clone stamp on a Smart Layer Randy PS will tell you in a pane that you can’t do it unless you rasterize and will give you the option to do so. Also with the Smart Layer highlighted, Layer>Rasterize>Smart Object. Or, right Click on the Smart Layer to bring up an options pane, then Rasterize Layer.

[Polar01]

Try pulling a new layer above the smart filter layer than clone/heal or other than select from layers below.

Cheers: Allan

[Loose Canon]

While this works as a clone stamping technique it won’t necessarily reflect any changes or carry over any changes you make if you go back and re-adjust the Smart Filter after performing the cloning.

[Thlayle]

Try pulling a new layer above the smart filter layer than clone/heal or other than select from layers below.

Cheers: Allan

Great idea. Seems like I could have thought of that...

[Thlayle]

While this works as a clone stamping technique it won’t necessarily reflect any changes or carry over any changes you make if you go back and re-adjust the Smart Filter after performing the cloning.

Yes, I thought of that too. Still for something simple, maybe a lens flare that isn't too bad, it might be a quick way to go.

[Manfred M]

I have changed my workflow to maximize the speed I can recover any changes when working with Smart Objects.

1. I have ACR set up to import the raw data as a smart object and immediately apply the standard filters like sharpening and noise reduction as well as any of the Nik filters I commonly use right on the bottom layer. As soon as those are done I will duplicate and rasterize the layer, just to speed up edits.

2. As I do much of my editing work with layer masks, blending modes and clipping mask. I will apply all these techniques to the rasterized layer. If I find I don't like the way things are going, I will go back and rework the bottom (smart object) layer and make a new copy and will rasterize it.

I then will use the layer masks and clipping masks and use these on the new rasterized material. It takes a few moments to apply these (and the appropriate blending modes) to the new material, but nowhere near as long as redoing everything.

3. My final step will be to do the destructive edits - cloning, healing, perspective correction on top of everything else. If I have to go back to the beginning, I will lose this work, but at least the amount of rework will be minimal.