except I dont want to use the transform tool
good luck.
Daniel,
You need to scale, angle and distort the image to fit each screen naturally. Each screen and image is different. How else would you do this if you don’t use the transform tool?
And unless you use your eye and retouching skills to mask up to the right point on screen bezels, I don’t see how this can be done, even if you can devise an action.
Neil
Daniel,
You generally need to scale, angle and distort the image to fit each screen naturally. Each screen and image is generally a bit different.
Unless you use your eye and retouching skills to mask up to the right point on screen bezels, and crop each image in an exact 16:9 aspect ratio in perspective (with minor adjustments to fit the screen), I don’t see how this can be automated. And unless everything is facing dead ahead and is properly sized, cropped and scaled, how do you do this without transforming the images?
Neil
Crop to the aspect ratio of the display (16 x 9?). Then use free transform distort and pull the four corners of the image to the four corners of the display, whatever the angle of the screen.
If the display image is the same for every screen pic placed, you could crop to proper aspect ratio, run through image processor to get them the same pixel size, and then use an action to place them. All actionable. If the display images are different, but the paths are already cut, I’m sure you could script it, but I wouldn’t know where to begin…
J,
The problem as I see it is that none of the screen openings is going to be exactly the same, even if the photographer tries hard to match the angle and size. And there is enough variation in screens that even a "standard" 16:9 opening inside the bezel is going to vary slightly with different screens. There is enough manual prep work here that manually sticking a picture into an opening and quickly tweaking it is the easy part.
Of course you, can easily repeat a transformation multiple times with the "Again" command in the dropdown menu. But that assumes that you are starting off with a matched set of original images — and that you placing them into a matched set of openings.
Well, one other thought comes to mind…I was assuming different images placed inside different screens. Is this the case? Or is it just one specific screen (at the same size and angle) with a variety of images placed in it? And how many of these are there going to be? If it is just a few, it doesn’t pay to spend much time on a script.
Neil
The problem as I see it is that none of the screen openings is going to be exactly the same, even if the photographer tries hard to match the angle and size. And there is enough variation in screens that even a "standard" 16:9 opening inside the bezel is going to vary slightly with different screens. There is enough manual prep work here that manually sticking a picture into an opening and quickly tweaking it is the easy part.
Of course you, can easily repeat a transformation multiple times with the "Again" command in the dropdown menu. But that assumes that you are starting off with a matched set of original images — and that you placing them into a matched set of openings.
Well, one other thought comes to mind…I was assuming different images placed inside different screens. Is this the case? Or is it just one specific screen (at the same size and angle) with a variety of images placed in it? And how many of these are there going to be? If it is just a few, it doesn’t pay to spend much time on a script.
Neil
Daniel, if the vertical sides are to remain parallel I have a Script that seems to performs that based on the topmost path (only if it has exactly four corner-points); and as a Script it could be recorded into an Action.
Im not sure how good it works, but mail me at pfaffi(at)viennapaint.com so I can send it You if You are interested, its to big to post here Im afraid.
Sorry, Daniel, the equation in the Script breaks down pretty quickly at differing sizes of the image thats to be placed in the screen, so to maintain constancy one would have to transform the image at the start before running the perspectival transformations (and further deteriorate the image quality) or possess mathematical skills beyond mine.
I guess the other contributors were right to advise You to do it manually.
Thanks all for input, looks like Illustrator is the answer, was originaly hoping for photoshop solution but thats fine. Found this:
www.layersmagazine.com/illustrator-and-photoshop-warping.htm l
I don’t see why you would need to go through the extra steps of doing this in illustrator. Photoshop can accomplish the same thing. Use transform. On a smart object if you must, then simply paste your image into the smart object. If you’re doing it to something that’s the same size every time, it’s as easy as copying and pasting your image into the smart object. If you have to do a different perspective every time, obviously you cannot automate something where the parameters change every time, but you can’t do that in illustrator either. In illustrator, using the method you’ve linked, you’d have to draw a new path for every new shape you want to paste into. Besides, it’s just more convoluted and takes more time than simply doing a four-point transform in photoshop. If you’re using the same image perspective for more than one image, converting it to a smart object and pasting your image into it couldn’t be simpler.
The matter bothered me and I looked into it further:
It is possible to automatically transform a layer perspectively according to a four-corner-path via JavaScript and its less difficult than I originally thought because the functionality is already part of the Scripts that are included with Photoshop.
So it would for example be possible to automate that process for a whole folder of tv-screen-images if the paths conformed to some conventions.