Smart Object Scaling Produces Artifacts

Please look at this picture:
When I scale down an icon using a smart object (original size is 1024x1024, the scaled down object shown here is 48x48 px), the borders of the vector shapes are added a weird sharpening effect that is not desired. When I rasterize the smart object before and scale down the resulting bitmap, the scaling is much more precise. What is wrong here?
The original graphic contains layer styles and my scaling is set to Bicubic Automatic.
I use Photoshop CC in the latest version.

Vector rasterization is a different process than resampling pixel data. That may result in a different distribution of the semi-transparent pixels used for the antialaising and presumably that also messes with your layer styles. There may simply be some very faint pixels that trigger the style to be rendered. Also of course some layer styles can never be smaller than 1 or 2 pixels, which at icon sizes may extend content. Therefore flattening the stuff beforehand may be unavoidable.
Mylenium

Similar Messages

Photoshop CS6 - Pasting Smart Objects from InDesign CS6 look awful

Since upgrading from CS5 I'm having a huge problem using Smart Objects.
What:
Smart Objects pasted from InDesign into Photoshop are pixelated and corny beyond belief. Originating material placed in InDesign is hi-res (e.g. 600ppi) but when pasted in PS it looks crappy.
How:
Selecting an image in InDesign and pasting it in PS in a 72ppi file.
Typically I need to copy elements from previous art in ID and re-compose and use for web.
Why:
Can't seem to fix it. Have tried with Anti-Aliasing both on and off to no avail. None of the Image Interpolation settings affect this either.
Only solution is to scale the placed Smart Object to something like 400% where it suddenly looks nice again, rasterize and then scale it back down.
Theory:
Smart Object scaling is flawed in PS CS6.
Note: All scaled images in InDesign already look gritty and awful. Even with Display Performance: High Quality set (using default settings). Seems to me like CS6 scaling is b a d.
Anyone else seeing this?
Thanks
Image 1: Smart Object rasterized at same size as placed. (ouch. awful.)
Image 2: Smart Object Scaled to 400%, rasterized and then scaled back to original size (yum yum, like it used to be without up-rasterize-down-scaling)

That does not resolve the anti-aliasing issue in CS6 - it still looks rather bad.
The anti-aliasing sampling method applied to smart vector objects is just extremely bad looking. And it is not merely an issue in Photoshop: exporting artwork from Illustrator at web/screen resolution using the web export is absolutely abysmal looking (not other word applies here, unfortunately).
I explain more about this issue here: http://forums.adobe.com/message/6002543#6002543
The best option is to blow up the artwork, rasterize, and then scale down with bicubic or bicubic automatic, depending. But even then you will not achieve the same quality that can be attained with other tools that, for example, support catmul rom and mitchellnetravali for downsampling.
Anyway, Adobe applications are very weak when it comes to proper downsampling and file size optimization for web and screen graphics. You're better off looking someplace else.

Smart Object Anti-Aliasing issues in PS CS5

I am having a heck of a time working with smart objects that contain transparency in CS5. Here is an example of what I'm experiencing:
1,700px smart object scaled down to 156px. No transparency in SO. Good anti-aliasing.
Same smart object, but with a layer mask or transparency in the smart object. But alas: horrible anti-aliasing, throughout the entire image.
I frequently embed my transparency within the SOs I work with, due to the limitations of linking embedded SOs with layer masks etc. I experienced no problems in CS3. I'm thinking this is a bug? Does anyone have any insight? Thanks!
UPDATE: after performing many tests, and noticing that having transparency in a smart object does not necessarily lead to poor anti-aliasing in the parent image, I have narrowed-down the issue. The issue will occur only if at least an antire column or row of pixels (of any height/width) along the border of the image are transparent. It does not matter if the transparency is at the top, bottom, sides, or surrounding the image, but it must take up some amount of an entire side. This is looking more and more like a bug... Is anyone able to replicate the issue? (Adobe engineers?) Thanks.

Here is the file I'm seeing this in. Incidentally, if I drag the problematic smart object to another open image, the appearance is the same (Photoshop is, after all, the epitome of wysiwyg software). But this works both ways: I can also re-create the smart object in another document without the poor anti-aliasing, then drag it into the original image and it will still look fine. (in this file there are two side-by-side, so that you can compare the two. you can also play with the smart object on the left to see the anti-aliasing go from good to bad; just go into the SO and delete a row or column of pixels from the border and re-save).
http://dl.dropbox.com/u/8162462/SO-aa-issue%3F.tif
(i used Dropbox; lmk if there's a better way to post a file in this forum)
Since I just started using CS5 (from CS3), I have suspicions that it may be a version translation issue (either in the TIFF or smart object). Either way, I'm glad that this is not a common or easily-reproduced issue. Let me know if you learn anything from the file
Thanks,
Cory

Smart Object Irregularity

This is a bit of a "deep dive"...
I'm trying to develop a retina workflow in photoshop for web projects. In doing this I found a strange irregularity in photoshop using smart objects and Photoshop CC 2014's new asset extraction.
Essentially what is happening is that when I import an image into an already open photoshop document and make that raster image a smart object (either by importing it using the "place embedded..." option or by converting the image to a smart object after placing it) I am getting different results when I try to extract those images at 2x using the new extractor. Of course you will have to perform a transform on the image to scale it down in the photoshop document after you make them smart objects (i.e. I know that 2x of the original smart object size will still be blurry — that's not what I'm doing). The image that is converted to a smart object after it's placed saves as 2x perfectly clearly and the one that was "place embedded..." comes out blurry as if it wasn't a smart object at all.
What should I do about this? I thought these two options were different ways of accomplishing the same thing. Honestly, I like to use the "place embedded..." option because it saves me steps, but it is the one that isn't working.
Thanks,
Ethan

You may want to add your +1 and your observation on different behaviour for placed and converted SOs to this thread.
Photoshop generator not working with smart object & scaling.

After scaling a smart object do I need to remake it as a smart object again?

Using Photoshop CS5 - after scaling a smart object do I need to remake it as a smart object again?

No! Just keep working with it you can adjust the original smart object file and retiuch it and all will update just click on the smart object in the layers panel everything will work as you expect it to.

DISGUSTED! ImageReady functionality lost for producing smart objects in GoLive!

It is ABSURD that the rollover/layer-based slice capabilities feature of ImageReady was eliminated and instead replaced by way of mediocrity using manual copy and paste functionality within Dreamweaver. What kind of step-backward mentality is this? I am beyond disappointed and feel betrayed by Adobe, not to mention unbelievably, inconsiderately inconvenienced. Does the title "industry standard" require bowing to irrational development or is this mere laziness on Adobe's part?
Previously I had been able to develop multiple series of rollovers within Photoshop using layers and then switch to ImageReady in order to automatically produce rollover code for a set of images...AS ONE IMAGE FILE. These images could then be placed into GoLive (another hot-button issue!) as SMART OBJECTS - gee, what a concept! - where the targets, if necessary and not already entered within ImageReady, could be linked to any page within the the site folders. Now I can't access the secondary image anywhere at all. It appears that the only solution for us lowly GoLive CS2 users is now to be required to TRIPLE THE AMOUNT OF WORK AND place multiple objects within a page layout, leaving (no pun intended) room for pixel space errors and requiring multiple manual page adjustments when the work was ALREADY DONE IN PHOTOSHOP! Layer-based slices are a work of genius, but NOT without the added capability of automatic rollover detection. How am I supposed to work this out in GoLive?
NOW, only NOW am I discovering that this "merged capability" within Photoshop is missing one of THE MOST INGENIOUS FEATURES EVER and now I am scrambling to find some garbage workaround in order to capacitate the need for a string of visual interface objects. This is the kind of move I EXPECT from Microsoft, but NEVER have I been this put out with Adobe. NEVER.
Is this the only solution? Bend to the will of the Dreamweaver cabal and join the ranks of the Microsuck copy-and-paste zombies? Not only is the new CS3 series "upgrade" cost-prohibitive for most anyone that isn't within the realm of the design elite or corporate ladder-climbing ranks of internal design group budgets, the tools appear to force anyone without the time necessary to learn the intricacies of manual HTML tagging to create nothing but the same garbage website in an infinite spectrum of ineptitude.
Is the new GoLive 9 supposed to accommodate this ridiculous shot-in-the-foot move or am I REALLY going to have to resort to a complete reload of CS2?

Dave and Buko, that's my entire point...I shouldn't have to be using CS2 after purchasing CS3. I have full versions for both thankfully and for this very reason but I don't know if I can install CS2 in addition to CS3 without causing some system file issues. I did briefly read about someone else trying that in another forum.
IR does indeed work great with GoLive CS2. My issue is that the rollovers in the extended Photoshop CS3 do NOT work and don't export code at all. It's damn frustrating that after making the purchase, I feel duped.
And Buko, no, I never even bothered with Dreamweaver. I attempted to learn it several times and was just put off completely. I previously owned both Macrodmedia Studio Professional and the complete Adobe CS2, but the upgrades for the combined product sets which made sense for me and didn't break the bank like purchasing the "Master Suite" didn't even include things like Fireworks, which supposedly has some sort of workaround for rollovers when setting multiple targets in an image file.

Why do vector shapes in scaled smart objects blur?

This remains a point of frustration.
Why does converting a vector shape to a smart object in Photoshop mean that it's then treated as a bitmap and blurs when scaled up?
If i paste the same vector shape into Illustrator and then back into Photoshop as a smart object it then behaves correctly, and remains sharp.
See the image for example of what i mean. Left is the master shape, centre is it converted into a smart object in Photoshop, right pasted to and from Illustrator. These were then scaled 300% resulting in the centre one blurring.
I don't really want to have to invovle Illustrator in this way and keep all my vector work in Photoshop now it has the tools to do so, but this issue creates me a lot of extra work.
Or am i doing something wrong?
Rob

Smart Object layers are transform to other sizes through interpolating the pixels rendered for the embedded object not be using vector graphics on the embedded object.. If your resize resizes the pixel count way un the image will become soft. Yoy can try seting your Photoshop Default interpolation setting to other setting then Adobe default Bicubic Automatic to something like plain bicubic which IMO is a better general purpose option then Bicubic Automatic.
You can also open the embedded object by double clicking the smart layers smart object icon in the layers palette. Once opened in Photoshop resize the vector graphic using Image size and save the embedded vector object larger. Photoshop will update the smart object layers content then.

Photoshop CS6 Scaling issues: Vector/Raster smart objects

My team and I have noticed some strange issues when affecting the overall interpolation of a .psd via scaling, cropping, or canvas resize.
Test:
Original .psd is even x even overall pixel dimensions.
Original vector and raster smart object assets are all proportionally scaled with width x height percentages equal.
Preferences: "Snap vector tools and transforms to pixel grid" is turned off
Resize .psd via Image > Image Size palette.
Scale styles, constrain proportions selected.
Every interpolation style tested.
Resize the .psd to 50%
Result: All smart objects (both types) no longer are proportionately scaled. Usually off by a minor amount under 1/10th of a percent.
Some objects' center point shift a half pixel to the right or left.
This also occurs if we use the crop tool or resize the canvas.
Is this a known bug and if so, is there:
1) any setting to truly lock proportions?
2) any resolution being pursued currently?
Though the shift and the disproproptions seem minor, the overall quality of items such as icons, hairlines, and other things that should be more crisp become far more degraded in CS6 vs CS5.
We have over 100 users and need to evaluate if we should drop back down to CS5, where we weren't having any of these issues, as our work demands a lot of precision, so you can see why this is worrisome.
Thanks!

I'm not going to say anything about the abysmal rendering of Vector Smart Objects in CS6 because it'll be beating a dead horse but I can explain the issues you see with measurements after scaling.
The slightly disproportionate scaling and half-pixel shift of the centre is unavoidable when it happens because the resulting Smart Object/Vector Smart Object must measure an integer number of pixels in height and width and be perfectly aligned to the document pixel grid. That's true in CS5 and CS6.
Say a SO/VSO originally measures 102 x 103 pixels and you scale it by 50%. The Options bar may say the result is 51 x 51.5 px but the result will really be 51 x 52 px, and therefore the new height will really be 50.49% of the old height.
The original SO/VSO is an even number of pixels wide and an odd number of pixels high, therefore its centre will be halfway down a pixel. The 50%-scaled SO/VSO is an odd number of pixels wide and an even number of pixels high, therefore its centre will be halfway across a pixel.

Vector shape to smart object, is blurry when scaled

I've recently been upgraded to Photoshop CS5.5 and am having issues with smart objects.
Quite often I would:
make a shape layer, say a circle
apply a layer style, say a stroke and drop-shadow
Convert this to a smart-object
Resize this object to whatever I needed and it would scale perfectly, the layer styles and crisp edge would scale.
Now, when I use this same workflow, the whole object becomes blurry - like it's become one giant bitmap instead of retaining its original nature.
The blurred smart object loses it's vector nature and looks horrible, this is surely an error.

It depends on how you scale the smart object layer. If you just use Ctrl+T free transform Photoshop will matieilize the pixels for the layer then resize the materilazed pixel layer like normal raster layer not text not a shape layer. If on the otherhand you double click on the smart object icon in the layer in the layers palette Photoshop will open the smart object in photoshop. If it a RAW file it wil open in ACR. If is is a Photoshop Document like a PSD, Tif, Jpeg etc it will open in Photosgop. If it is some Photoshop object like a layer group a collection of layers an vector shape layer it will open in Phpotoshop. You can then scale the smart object with Photoshop tools like ACR or Image size set resample constrain scale layer style etc to scale the smart object when you use save the change Photoshop will update the smaer object layer in the document. You can do the save by clicking the Close icon and respond yes to save. Scaling the smart object layr this way will scale text and vectors the way you want them to be scaled.
http://forums.adobe.com/message/3498406#3498406

Black Edges Around Scaled Smart Objects

Hey folks,
I've checked around and haven't really found a good solution for this, hoping somebody can help me out. I have a smart object with transparency that looks great at 100%, as soon as I scale it down I start getting black lines around the edges and a "bevel and emboss" type effect around the edges. I DON'T want to scale down the contents of the smart object, that's the whole point of the smart object, to retain resolution. The "anti-alias" checkbox in the transform panel is disabled and can't be checked. I'm using Photoshop CC 2015. Please help as this has been driving me nuts for months. Thanks!

You are correct. I tried this in CS6, and the drop shadow causes the downscale algorithm to add an ugly dark border. I tried changing the image interpolation in the General options, but that has seemingly no effect at all.
Under circumstances it looks quite terrible, and the result becomes unusable for serious work. The only way to circumvent this is to remove the drop shadow in the SO, and apply the effect instead on the SO itself.
*edit* Out of curiosity, I also tested this in Photoline with cloned instanced layers (comparable to a SO in Photoshop), and there are no issues.

Smart Objects rendered incorrectly in CS6

Here is a logo, pasted as a smart object from Illustrator into Photoshop CS5.1 :
Here is the same logo pasted as a smart object from Illustrator to Photoshop CS6:
Note the gaps in the orange gradient at the top of the graphic, and the transparent areas in the border of the shield. This discrepancy appears (to me) to be a result of using a "sharper" antialiasing mode in CS6 which is leaving 1 pixel wide gaps between objects that are actually completely aligned to the same points in Illustrator. Scaling the object will result in the gaps showing up in different areas of the graphic. This is incorrect behavior on the part of Photoshop CS6. Will there be a fix forthcoming? I'll have to keep working in Photoshop CS5.1 for the time being until issues like this are resolved.

I think the discussion got sidetracked here on talking about anti-aliasing of "raster" smart objects. THIS IS NOT WHAT THIS COMPLAINT IS ABOUT.
The original complaint, which still stands, is that vector artwork is now rendering with 1 pixel wide gaps between elements that are, in Illustrator, exactly aligned. If you go back and look at my original two examples, the logo in the second screen shot has gaps in the orange banner in the logo, and is missing elements in the border around the shield. These are gaps or holes in the logo caused by a change in the way vector objects are rendered in Photoshop CS6. This can NOT be corrected by changing the antialiasing style - there is no choice of rendering method used. You get the new one, that is broken.
In this particular example, I opened up an old CS5.5 Photoshop file in CS6, and edited it. During that process I resized the logo a little bit, and didn't notice the gaps and lines that now appeared in the logo. I ended up sending out artwork that was affected by this bug. Luckily I did finally notice and in the nick of time was able to fix it by using the .PNG export work-around.
I was going to post a very simple example where I had a grid of filled boxes in Illustrator that were exactly adjacent to each other, which should result in a single larger solid box in Photoshop, but instead results in 1 pixel wide gaps between the elements. You can do this experiment for yourself. The reason I have not posted such an example is that, to my dismay, this example also now exhibits the same thin gaps in Adobe Illustrator CS6. So the rendering of vector images is broken there, as well.
Dragging and dropping elements out of InDesign into Photoshop is also a workflow that I have developed when creating web pages from previously produced print designs. This bug also affects this workflow. Every seperate vector shape in a logo or other compound artwork is now being rendered in such a way that small gaps might appear between elements that should be touching. I imagine that they are being pixel aligned or otherwise sharply antiased so that two shapes that are touching in the vector artwork are now being drawn not-touching. In the original example, if I resize the logo to a different size, the places the gaps happen change - some places that are gapped now touch, others that didn't have gaps now do.
THIS IS A BUG which will affect legacy artwork, and will affect workflows. This has degraded the usability of the Creative Suite as a whole, and should be corrected as soon as possible.

After loading a JPEG file into adobe CC as a smart object and trying to resize image I get big jagged edges around the picture.

Also, the anti-aliasing box isn't checked but is greyed out so I can't check it. How can I fix this issue? Thanks for any help!

Smart object layers transform like pixel layers. The resize may be using your default interpolation method. If you left it at Adobe default setting Bicubic Automativ and you are down sizing a sharpen image it may be using Bicubic Sharper. When Bicubic Sharper is used on sharpened image jagged artifacts may happen. Try changing your Photoshop Interpolation preference to plain bicubic. I'm not a fan of Bicubcr Automatic or Bicubic Sharper. I will use Bicubic Sharper when I re-size my soft unsharpen images for the web.

Ps CS6 Problem: Transform applied to Smart Object fails to transform an attached Smart Filters Mask

Ps CS6
OSX 10.6.8
Problem: Transform applied to Smart Object fails to transform an attached Smart Filters Mask.
I mean a Transform, including Free Transform, as found in the Edit menu. A simple move by the Move Tool is OK.
A workaround until this bug is squashed is to encapsulate the Smart Object + Smart Filters + Filter Mask inside another Smart Object and transform that.
However, that will not be a satisfactory solution in some cases. If a filter has size parameter(s), e.g. Gaussian Blur radius, a scaling or warping/distorting transform applied after the filter will obviously differ from the filter applied after the transform.
In any case, the workaround is inconvenient to subsequent editing and experimenting with filters and masks.

R_Kelly wrote:
I don't think that's a bug since the implementation seems to be purposely done.
It's been that way since photoshop cs3.
If its been like that since CS3 then I think it's a bug which remains because nobody (or not enough people) has complained before.

Photoshop CS6: Pros and Cons of Using Smart Objects

I haven't had Photoshop CS6 for that long, and have only just got past feeling uncomfortable with using Curves, now I've learnt how to use them properly.
My concern is - I am currently learning about Smart Objects. The concept, at first, seemed like 'the best thing since sliced bread', being able to non-destructively use filters, Shadows/Highlights command, Unsharp Mask, endlessly scale using Free-Transform etc etc, without harming pixels at all.
However, the more articles I read about their use in Photoshop, the more I am afraid to start using them in my workflow.
I understand that when you convert to a Smart Object, this process is non-destructive, i.e. I can perform as many readjustments to a filter, for example, and Photoshop will always work from the embedded container file (which has had no filter adjustment made to it) to adjust the filter to your most recently adjusted settings. If you later decide you don't want to use a filter at all, and rasterize the Smart Object back into a regular layer again, is this process non-destructive as well?
Then there is this article, which I struggle to understand properly:
http://bjango.com/articles/smartobjects/
Please see the part 'Smart Objects Created in Photoshop'. It seems to say I can't scale with a Smart Object without causing interpolation and blurry edges. Please can somebody clarify what the writer of this article is trying to get across, because it is well documented that Smart Objects can be endlessly rescaled non-destructively.
Please understand I use Photoshop primarily for editing photographs.

There is much modern focus on "non-destructive" editing, but keep in mind if you don't overwrite or destroy the original file there is no destruction at the highest level. Put in layman's terms, you could always start over with the raw file.
That thought segways into my next one: Non-destructive editing makes sense if you need to use the same information for a variety of somewhat related purposes, or if the work product may need to change (e.g., to suit the whims of a fickle client).
But at another extreme, if you're editing for a particular purpose - say creating the best possible print from an exposure - sprinting right for the finish line by changing pixel values directly and being done with it can be an extremely effective approach. This requires that you get things right the first time, and that takes practice.
Some folks do their Photoshop work by building up layer after layer and using smart objects, smart filters, etc., and this can be effective but no computer has yet been built that can composite all that stuff in real time with a big image. So there IS a cost to doing it. What you might gain by being able to re-do things, you might not have needed to gain if your control responses were instantaneous and you could tweak the intermediate result at every step very easily. Note the number of posts about how slow Photoshop CS6 is/was at editing deep documents, some by people using 2012 computers.
As with most things, it's horses for courses. It's good that Photoshop gives us rich tools and choices for how to work.
Regarding your specific question, bear in mind that what's communicated to the parent document from each of its embedded Smart Objects is a flat, rasterized image. Think of the embedded smart object kind of like going off and opening another document, making the changes you want, saving the document, then flattening it and pasting the pixels into your parent document.
In the very first example in the linked article, they show how the smart-object-rasterized image of a vector circle, subsequently scaled by resampling the parent document in which the Smart Object is used, becomes fuzzy as it is scaled up. Once you understand this you realize that of course you could scale up the smart object itself, e.g., to a size equal to or larger than what's ultimately needed by the parent document, and then it could be crisp in the parent document where it's used.
Of course, having all your smart objects at a size larger than you need takes up even more resources.
-Noel

Smart Object Transformation in CS5

Has anyone noticed that when you scale down a smart object in Photoshop CS5, the transformation tool resets the dimensions to 100%? In CS4, you are able to see the transformation properties once you scale down (by option-Shift clicking on the bounding box). It would be helpful to retain the scaled dimensions of the transformed smart object in case you need to import other smart objects to rescale.

Thanks. Here's a link to the video of me transforming in CS5:
http://www.notchill.com/tmp/CS5_SmartObject_Transform.mov
Please note: The smart object I imported was already scaled down from the original size, but the transform controls show it as 100%.
In contrast, here is a screen cap of me doing the exact same process in CS4:
http://www.notchill.com/tmp/CS4_SmartObject_Transform.mov
As you can see, once I get into transform mode, it shows that the smart object has been scaled from its original. This data is very valuable.
Hope this helps to clarify. Please note this applies to rotation as well.

Smart Object Scaling Produces Artifacts

Similar Messages

Maybe you are looking for