How can you achieve realistic closeup perspective in Photoshop using 3D Postcard feature?

Question

I'm trying to achieve this look in Photoshop 3D, but I am struggling to get the perspective to look realistic. I create 3D Postcards from screenshots of software I am trying to showcase. I have experimented with the depth of field blur for realism, but the problem I'm having is mostly just getting the proper angles in XYZ space, by rotating the card, in combination with the camera views. In the 2nd example, there looks to be another card floating above the other 3D card, with the dropshadow landing on top. This makes me think that this was done in Photoshop 3D.

I've seen an example posted in this forum on a similar topic called "3D Card Effect" See example 3. There was no explanation in the comment section where the user showed his final example where they somehow achieved a realistic camera view. I can never seem to get the perspectives to look correct, they always seem to look off.

Also, to get this closeup on a screenshot, the image tends to get very blurry and lo-res. So, how would something in Example 3 have such hi-resolution if the user is using Photoshop screenshots for their mockup? I'll attach my best attempt as example 4. Clearly it's too closeup for nice resolution

The second issue I have, is if I am using 2 cards and I move one in Z Space, I then select both cards and choose 3d/Merge 3D Layers, this way I can spin the camera to get an interesting angle. But when I merge the layers sometimes the top layer will rasterize strangely and turn very blurry, or any effects on the layer will act strange. Like an outerglow will become dark noisey solid blue. Was wondering if there is a step in between that I may be missing, like converting the layers to smart objects before converting to 3D cards.

Any thoughts would be much appreciated.

Answer 1

This is possible without 3D. Learn to distort layers with Edit > Transform. There's even perspective tool, but as well you can scale, skew, rotate and use distort (=drag corners). Increase canvas size to have room. An example:

Be sure you have transform controls crossed to be visible with the layer move tool:

There's no need to make selections, the current layer is adjustable as soon as you have clicked the move tool.

You can readjust the distortion infinitely with no quality loss if you have converted the layers to Smart Objects. Learn the right click options in the layers panel for speedy work.

The shadow is a copy of a layer (not a smart one, use a rasterized copy) which is blackened with curves, blurred and finally made partially transparent. You can use as well Layer Style Drop shadow to get adjustable shadow, but the blurred copy method allows local edits. Here a part of the shadow is deleted.

Layer style is infinitely readjustable, but the blurred copy isn't.

Closeups and extreme perspectives have a hiccup: Screenshots have easily too low resolution, because parts of them need enlargening. The next has got fake +100% resolution increase with On1 Resize. Without it the near part looked blurry and pixelated. There's also lens blur and vignetting. The latter is is subtle and made with masked adjustment layer:

There's also a good free program to add resolution without blurring the image. It's SmillaEnlarger. Those enlargeners differ radically from Photoshop's resampling. They try to guess sharp edges and thin lines. Often they succeed to keep them sharp and thin.

If it must be in 3D for finding easily good viewing angles and perspectives, you can use Illustrator's Effects > 3D > Rotate. The result is possible to take to Photoshop with copy and paste.

The need of resolution boost is the same as in Photoshop for closeups and extreme perspectives. See an example:

Illustrator has a trap: Distant areas get easily pixelated (like in my screenshots) if in the beginning of the job raster effect resolution is not set high enough. When I started the new Illustrator drawing I had raster effect resolution only 72 dpi and that's too low (Photoshop's 3D has the same problem if you start with too low image resolution). Having 300dpi raster effect resolution makes much finer results, but rendering the image isn't instant.

Illustrator has benefits: simple, mathematically right result and it's easy to use the same wiewing angle and perspective repeatedly. In Illustrator you can have more than one image with the same or different 3D views. You can readjust non-destrucively all of them via the Appearance panel. Shadows, decorative graphics and texts can be made in Illustrator, too.

Use 3D effect Extrude & Bevel instead of rotate for thickness and edges:

ADD: Screenshot is a planar rectangular piece. Every possible perspective is achievable with geometric distortions, where straight lines stay straight. So, Photoshop's transforms should be enough. Bigger problem is to keep the perspective and the apparent light consistent, when several pieces are placed to the same image.

Your examples have also texts on the images. Keeping them readable needs contrast reduction and recoloring in the images. The lens blur effect is surely useful. Now the real difficulties start. Anyone can grab together something like the following in ten minutes:

But finding a new and interesting composition, keeping images still natural looking and texts readable + fitting all to the wanted style needs extreme care. The problem is larger than a pure technical question - definitely a good reason to get a pro designer.