Filters on High Resolution Images

[kingsushi001]

Hi Everyone

Hope this is in the correct place.

Quick question, is there any way to get the same results from filters on high resolution images
than on low resolution images?

For instance, here's a low resolution image with the halftone filter applied:


And here's the same image with a high resolution with the exact same filter settings applied:


As you can see, it's clearly not the same results. Any way to get the same results?
Especially when using smart objects? Not too sure if it matters, but I'm on CC 19.1.4.

Any help will be greatly appreciated.

Arnold

 

[thebestcpu]

HI kingsushi001

Yes there is a way yet not with full flexibility. Note that the filters that are pixel based will only work the same for a given number of pixels for the same setting. At higher resolution images, the same settings wont work (as you discovere)

I don't have time to include pictures so hopefully the verbal description will get you going.

1) Start with a pixel image that is turned into a Smart Object
2) Open that Smart Object
3) Use Image > Image resize to reduce the resolution down to the desired pixels so the filter will work e.g. 10% of original size
4) Turn this into another Smart Ojbect (SO within SO)
5) Apply the pixel based filer for this Smart Ojbect (allows you to come back and make changes to this filter)
6) Save this Smart Object and go back to the main file (the image will be smaller)
5) Not do a transformation on this Smart Ojbect Layer (CMD+T for Mac Ctrl+T for PC) and stretch the image back to the original size (or use the fill in boxes and scale by e.g 1000%) and then accept this transformation

If you go back and open the Smart Object you have the lower resolution image which you can change the filter settings, Save, and go back to the main file and it will be scaled to the original image size still.

Hope that gets you going. Again, not totally flexible yet maybe this approach will help

John Wheeler

 

[kingsushi001]

Hi John

Thanks for the explanation. However, is the results supposed to be very blurry?

 

[thebestcpu]

HI Arnold

I think we are at a point where we need to get down to your specific requirements i.e. the specifics of what you are trying to achieve with your end results. There are usually a number of ways of creating the same result in Photoshop and/or ways to correct for any issues/anomalies that can be created with any particular approach.
Ob
What I was trying to demonstrate is how to have a filter create similar results with the same filter settings for different sizes of images (pixels) and have it all operate in a Smart Object.
So this was more of a "here one way how you can do that" Yes, if you shrink and image in pixels and increase its size back up, you can certainly create different anomalies based on which rendering algorithm is set in General Preferences. Those anomalies can include being blurred, or hard pixelated issues etc.

That there may be other and better ways to do something you desire if you have a specific image or process you want to implement. So I suggest we get down to specifics of the problem you want to solve (as opposed to a generic solution) and see what forum members can provide for you

E.G. - Do you want to just have a B&W resulting dot half tone that works at higher pixel images. Is it really important you do not have to change the settings. Are you looking for hard edges in the result or a smooth result Do you need a generic solution to many filters or to this specific filter etc etc

Just trying to guide you to get the best solution from the forum and hope this helps

John Wheeler

 

[kingsushi001]

Hi John

Thanks for the reply.

I used the halftone filter just as an example. Basically what I wanted was something like this: How to Make a Stylish Pop Art Portrait from a Photo
I just wanted to swap out the image inside the smart object and get he same effect at a high resolution without messing with the filter settings too much.

It's not just for that type of effect, there's quite a few I'd like to make using smart objects, but if I can't achieve the same results at say A1 resolution (for print),
I'll have to do some of the stuff manually. Was kind of looking for a "one size fits all" type of solution.

 

[thebestcpu]

Hi Arnold
I reviewed the link your provided and that was a cool effect. I did notice that he was promoting how you can apply the same effect to many images without other changes without any changes to the filter settings.

How that was achieved was first scaling the image to an exact pixel dimension image. The first image was reduced down to 55% of the original size and fixed pixel dimensions.

I imagine if you used the same approach i.e. a fixed pixel dimension for the starting image, you could have the same leverage across many images with the same filter settings.

So that particular tutorial leveraged pixel based filters for their settings by requiring the starting image to have the same pixel dimensions and you could do the same for your approach as well.

Now, whether it is the images in the tutorial you shared, the example I gave, or for any of your images, if you first reduce the resolution (reduced pixel dimensions) and want to return to a higher resolution image (more pixels per edge), you will lose some image quality.

THE KEY - Start with a fixed pixel dimensions and stick with it for a given effect you desire and then you will be able to leverage the same filters with the same settings.

Hope this helps

John Wheeler

 

[SCTRWD]

I am not sure what you want.. but:
There are two options:

1. You want your filter to work with the size of the image you choose it to have in your document.
Then just create Smart Object with the source of your image, add filter to it, and transform your Smart Object to any size you want it in your document.

2. You want your filter to work with hi-res image but the results in your final document have to be low-res
This is one of the many good uses of Smart Objects.
Many tools in PS(including Filters) work good only on hi-res images. But you need the result in low-res...
Then you just create Smart Object with hi-res source, make all the editing in hi-res, save and get the results in low-res in your document. Yes! They will be scaled down... But:
Some effects are just not even possible or give really UGLY results(snow, hail, rain... You name it... ) in low-res . But the same effects made on hi-res Smart Object source and then scaled down to low-res may look MUCH better in your final document than the same effects made on low-res version in the first place.