Aligning perspective in Photoshop. Photoshop CS6 - Perspective Correction Tool

To be honest, I really love photographing architecture. At the same time, without a special lens, I often encounter an unpleasant effect - parallel verticals in a photograph become converging. This is expressed in the fact that the walls of buildings or columns, etc. They look like they are leaning towards each other. These perspective distortions occur because when shooting tall, vertical buildings, you need to tilt the camera to get the entire object in the frame. You can bypass this effect directly when shooting, but, alas, it is not always possible, so you have to apply perspective correction in Photoshop.

First, I want to tell you how you can slightly reduce perspective distortion when shooting. This is important because when corrected in Photoshop, part of the image will be cut off. As I already said, the effect of converging lines occurs when the matrix deviates from the plane of the object, for example, when shooting from below a tall object.

To reduce the angle of the camera, you can try:

• move away from the subject,
• try to stand on some elevation and raise the camera above your head,
• change focal length to trim the bottom and/or top part photo,
• use a special tilt-shift lens.

This is worth remembering when shooting, but sometimes the shot has already been taken, and the problem only became noticeable after some time. Well, this is where the ability to correct perspective in Photoshop comes to the rescue. I would like to add that with such correction there are no special digital effects; the method was carried over from film photography. In essence, it consists in deflecting photo paper from the plane of the frame when printing, thus compensating for the distortion of the verticals.

As always, there are several ways to solve a problem in Photoshop. I will tell you about two, in my opinion, the simplest and most convenient. The first one is suitable for correcting perspective in automatic mode and does not always produce the desired result, but in terms of speed it only takes a few seconds. Let's take a photograph of an Armenian church as a basis:

Original image with perspective distortion

It is clearly visible that the verticals are piled towards the center. Select the tool " Crop Tool» (« Trimming"), and be sure to check the box at the top " Perspective» (« Perspective"). Next, we move the upper edges of the frame towards the middle so that they are parallel to the verticals in the photograph.

Correcting perspective with the "Crop Tool"

Click " Enter" and we get the corrected picture. The building has become a little flattened, to get rid of this you need to stretch the picture vertically. This effect also appears in the second method, so I will talk about the correction towards the end of the article.

The perspective has been corrected, but the picture is slightly flattened

The second method is to use the command " Perspective» (« Perspective") from the menu " Edit» (« Edit«) — « Transform» (« Transformation"). It gives you more options and, in addition to perspective, allows you to correct the tilt of the image to the left or right if the camera was tilted to the side when shooting.

Using the Perspective Command

When using this command, you need to pull any top edge to the side and visually select the desired level of perspective change. In addition, if you move the middle top point to the left or right, you can “tilt” the image in the desired direction to compensate for the tilt when shooting.

The result obtained is somewhat different from the previous one. Now we need to stretch the image vertically to compensate for its “flattening”. To do this, go to the menu again “ Edit» (« Edit«) — « Transform» (« Transformation") and select the command " Scale» (« Scale"). Stretch the image upward beyond the center point.

Final image with corrected perspective

Using any of these methods you can easily and quickly correct perspective in Photoshop. However, the second method provides slightly more possibilities. In some cases, you should not strive for absolute parallelism of vertical lines because When corrected, the picture becomes deformed, which is especially noticeable if there are round objects in the frame, such as domes.

Now we release the mouse button and Photoshop adds a frame around the image, inside the frame there is a grid, and in the corners and in the middle of the sides there are movement handles:

If you don't see the grid, make sure you have the Show Grid option checked in the options bar at the top of the screen:

Now we need to drag the top left frame handle to the right until the interior grid lines are parallel to the slanted left wall of the building. In order for the marker to move strictly in the horizontal direction, you must first hold down the Shift key:

Most likely, changing the right side of the frame grid will also affect the left, already finished, side, so you will have to correct it again. You can also drag the handles located in the bottom corners, but in my case this is not necessary. Please note that part of the image on the left and right sides is currently outside the crop box. These areas will be trimmed after the final application of the tool:

Once you have adjusted the angle of the grid, you can also change the size of the cropping frame by dragging the handles located in the middle of the sides of the frame. Here I'm dragging the right-hand handle to the left to crop out the unwanted area on the right side of the photo:

When you're done with the settings, click on the checkmark in the options bar or press Enter to finally apply the tool:

Photoshop will instantly trim off the excess and apply perspective correction; the previously tilted houses are now vertical:

And now about the sad thing. One of the problems with this tool is that it is not that precise algorithms for its application do not exist. After using the tool, it may turn out that the angles of inclination of objects are not the same as we wanted, that is, in our case, the inclination is not completely corrected or, on the contrary, they are tilted in the other direction. If this happens, simply press Ctrl+Z to undo the changes you made using the Perspective Correction tool, and then try again. It may take several attempts to achieve the desired result.

Another problem you may encounter is that after you've corrected the angle, all the objects in the resulting image may look a little flattened vertically. In my case, the hotel building looks lower than it was originally, and people walking towards foreground, looks shorter than before. We can easily solve this problem by stretching the image using Free Transform.

Before we get started, let's look at the layers panel, where you can see that my photo is currently a background layer (in English Photoshop, instead of the name of the layer "Background" it will be "Background"):

Photoshop doesn't allow us to use Free Transform on the Background layer, but luckily all we have to do is rename the layer. To do this, hold down the Alt key and double-click with the left mouse button on the layer in the layers palette. This action will rename the layer to “Layer 0” and remove the lock icon:

Now we can apply free transformation, to do this press Ctrl+T. An overall frame will appear, and to stretch the image, drag the handle of this frame, located on the top edge, slightly up:

Once you are satisfied with the result, press Enter to apply the transformation.

Well, that seems to be all, look at the finished result:

In this lesson we will learn how to build different kinds perspectives, having looked at examples of perspective in photographs and drawn compositions, we will become familiar with the basic principles of constructing perspective and try to apply it in practice.

1 Step

Perspective is used to achieve a three-dimensional effect in images on a two-dimensional plane. The lack of perspective in the image will be visible even to the naked eye. Using basic techniques building perspective, you can give a sense of realism even to some of your absolutely fantastic drawings. So, I created a new document and placed a human figurine on it for order.

2 Step

Let's start with a perspective with one point of convergence. Any drawing on planet Earth will be made with the presence of a “horizon line”, regardless of whether it is located directly on the canvas or not. In the example below, we have a very ordinary horizon line in the middle of the image. In addition to the horizon line, we will also need a vanishing point. When working with perspective, at one vanishing point we have a point at which all the lines along the Z axis converge, designated VP (Vanishing Point) in our drawing, and a horizon line designated HL (Horizon Line). In this case, all parallel lines going to the horizon will be directed at an angle leading to the vanishing point.

3 Step

If we need to draw a cube, or parallelepiped, the lines along the X and Y axes will look like in normal constructions (black squares in the example shown), but the lines along the Z axis will go towards the vanishing point VP.

4 Step

Shifting the horizon line can give your composition more space for viewing the area and can be useful in cases in which it is necessary to display the landscape, or to perform a view from a higher angle. The vanishing point remains in the same position along the X axis as it was, but moves higher vertically, as shown in the figure.

5 Step

If, on the contrary, the horizon line is lowered down, this will give us the opportunity to depict details in the sky: birds, skyscrapers, aviation - in general, everything that comes to mind in this direction. This way we focus less on the ground and horizontal surface and get a more familiar point of view for a person, in which the horizon line is slightly below, and the eyes unhinderedly examine the street or the surrounding space.

6 Step

Now can you guess where the horizon line and vanishing point are?

7 Step

Yes, the previous step was probably unnecessary. Even without the obvious presence of these objects, we can always understand where the lines ultimately converge - at the vanishing point. This makes it clear that without a horizon line in the drawing, you can get a beautiful view from above.

8 Step

9 Step

This is what will happen in the end, if you move the vanishing point infinitely far away - we get a full-fledged top view. In practice, you are unlikely to need to draw from this perspective, but knowing how views from different angles are created will allow you to better handle the use of perspective when designing your own work.

10 Step

You're probably already wondering how I draw all these lines. Now, using perspective with two vanishing points as an example, let's look at the technique of drawing guide lines in Photoshop. Previously, we looked at perspective with one vanishing point. Now we will draw another vanishing point and get the new kind prospects. As for the first point, create a new layer for the second (Ctrl + Shift + N). Next, I took a regular 4px round brush and, while holding Shift, drew a vertical line of red color.

11 Step

Here I would need to draw vanishing lines, but as a right-hander, it’s not very convenient for me to draw from left to top to right to bottom, and vice versa. Therefore, to speed up the process, you can display the canvas horizontally, perform all the necessary actions, and then return it back. Go to Image > Rotate Canvas > Flip Canvas Horizontal. You may have noticed that I have this command tied to the F1 hotkey (which by default brings up the help window in Photoshop). You can set hotkeys in the Window > Workspace > Keyboard Shortcuts & Menus menu.

12 Step

If you get curved lines by hand, you can use other tools to draw them, besides a brush. This could be using a pen followed by aiming (Path + Stroke Path), Line Tool and so on.

13 Step

Here is the result of the freehand drawing.

14 Step

To correct the lines, use the Free Transform tool (Ctrl + T). Any line you create can be rotated with its help by pulling the corresponding corner of the object in the desired direction.

15 Step

Draw a few more lines until you have enough to work with. Now let's flip the canvas horizontally in the opposite direction by going to Image > Rotate Canvas > Flip Canvas Horizontally. I also lowered the Opacity of both vanishing points to make it easier to paint over them.

16 Step

Here the lines of the object tend to the first vanishing point (indicated by red arrows) as if it were constructing a perspective with one vanishing point (step 3).

17 Step

Look at the dotted red lines in the example below to guess where the end points of the second surface (circled) will be. This is where the power of projection comes into play. A little imagination will help you figure out where this surface should be.

18 Step

If you now draw parallel lines (red arrows) heading towards the second vanishing point (VP2) you will get a nice second face of our box.

Step 19

This technique also works for complex symmetrical shapes. If you look at the example below, the point circled in red will be along the path of the perspective line from the first vanishing point passing through the corner of our future figure (indicated by the mark “collide” - the intersection of two perspective lines). Then this line leaves the corner point in the direction of the second vanishing point (red lines), stopping at the point obtained by intersecting the parallel to the horizon drawn through the initial one specified in the red circle. In general, this is easier to see in a picture than to describe in words, so let's look at an example.

20 Step

Having carried out similar constructions for each of the points, we obtain very specific results.

21 Step

Finish the connections process and notice that the vertical lines ( pink color) continue to remain strictly vertical regardless of their position on the canvas.

22 Step

If we add a few more lines along the X or Z axis, they should all also go to their vanishing points, as shown in the figure below.

23 Step

Often using two vanishing points on the same canvas gives a very noticeable result.

24 Step

To complicate the task, you can move one of the vanishing points far beyond the boundaries of the canvas so that we cannot see the point itself. In this case, drawing lines correctly becomes a rather difficult task. The best solution would be to first draw a set of parallel horizontal lines, and then transform them using the Transform Tool. To begin, draw a series of parallel lines on the canvas, remembering to use the Shift key.

25 Step

Now use Free Transform (Ctrl + T) to rotate the lines in the direction you want. Keep in mind that at least one line (red) must coincide with the horizon line (green), otherwise the perspective will fail.

26 Step

Now we will repeat the same figure but without the effect of wide-angle lenses, which brings the vanishing points closer to each other. By moving the vanishing points further apart, you will get a picture that is more familiar to the eye, which will be perceived much more realistic. Remember that the Z axis lines (red) should be directed towards one vanishing point, and the X axis lines (blue) should be directed towards another, in this case an imaginary point located outside the canvas.

27 Step

Not bad. With such a low horizon line, we obtain a viewing point from a height characteristic of the level of the human eye.

28 Step

Now let's try to raise the horizon line and see how this affects the entire composition.

29 Step

The verticals of our figure remained practically unchanged in length, but their position changed slightly in accordance with the current lines of perspective.

30 Step

And again draw the horizontal edges of the figure along the Z and X axes as required by the perspective construction procedure.

31 Step

With this position of the horizon line, we get a top view of the object, as if we were viewing the area from a balcony in some hotel.

32 Step

Now consider perspective with three vanishing points. Everything will be the same here, only with the addition of an additional vanishing point that does not lie on the horizon line. Let's start by creating a standard perspective box with two vanishing points.

33 Step

Using the techniques from step 24, create lines for the intended vanishing point, located quite high outside the canvas, as shown below.

34 Step

This vanishing point is responsible for the direction of the verticals present in our drawing.

If the cubes in your composition are in different planes, you can create several combinations of vanishing points along the X and Z axes for each cube. But keep in mind that the verticals must go in exactly the same direction, so the Y vanishing point for each cube must be the same. This is true for anything: when drawing buildings, walls, cups, etc.

35 Step

Returning to our cube, we need to remove its vertical faces and redraw them according to the new perspective lines going into the sky to the third vanishing point. When working with perspective at three vanishing points, a great sense of scale appears, a kind of macro view of the object, because in real life everything is perceived more straightforwardly.

36 Step

As you can see, raising the horizon line in combination with the third vanishing point below works just fine. Imagine what would happen if you moved the third vanishing point even higher - towards the horizon line, giving an even more breathtaking vertical perspective. By using this technique sparingly, you can get very quality result with a subtle hint of perspective.

37 Step

But now let's look! Cool, yeah?? Now try to create something similar yourself.

38 Step

Now let's look at how perspective is present in photographs. Fortunately, in these photographs we have a clearly checkered floor, which makes it easier for us to determine the perspective lines along the X and Z axes. Like all normal photographs, this one also has a vertical perspective with a third vanishing point, but in this case it is almost impossible to determine it due to the absence of obvious vertical objects.

39 Step

This image shows a great example of Z-axis perspective, and by looking at the window lines on the building on the left (highlighted in green), you can easily spot the perspective lines going down the street. The wall on the right with its serifs also helps us establish the vanishing point.

40 Step

Since our building is obviously rectangular, and we'll assume the architects did a fantastic job of building it at exactly 90 degrees, the windows on its left plane (highlighted in yellow) and the second wall plane on the right show us the direction of the lines along the X axis. Also here you can observe a clear vanishing point somewhere in the sky - the verticals in the photo perfectly demonstrate its position. The perspective here is similar to the one we talked about in step 35.

41 Step

In this photo, the most obvious vanishing point is right above the top border of the canvas. The road and long buildings (colored green) make this vanishing point easy to spot.

42 Step

The perspective along the X axis is a little more difficult to grasp, but a trio of parallel buildings (highlighted in yellow) will help us with this - we can determine how far away the vanishing point is by catching a barely noticeable tendency towards convergence. Also in the picture there is a hint of a third vanishing point (pink lines) - if we look at the verticals of some high-rise buildings.

43 Step

Here, many of the buildings are not parallel, like our cubes in step 34, but the street planes (green and yellow highlighting) are located at their own angles, which allows us to determine their perspective along the Z axis.

44 Step

However, notice that the vertical lines, no matter what, all converge to one vanishing point. After all, even if the buildings are deployed in different ways, they are all directed strictly vertically.

45 Step

Now let's create something of our own based on the knowledge gained in this lesson. Here the X perspective lines are red and the Y lines are blue. We will not draw the Z axis into depth for now, since now we only have a two-dimensional image. Each set of lines was created on a new layer (Ctrl + Shift +N) to provide greater flexibility in the work.

46 Step

Well, now let's add a little depth by drawing perspective lines along the Z axis to the third vanishing point.

47 Step

Now let's draw the remaining lines in X and Y to provide depth to the letters and get rid of the extra length of the green perspective lines in Z.

48 Step

Because the green lines are on a separate layer, you can turn their brightness down to zero to paint them black. This can be done by changing the corresponding Lightness parameter in the Hue/Saturation window.

49 Step

Now let's paint our creation with some nice color.

50 Step

Now all we have to do is add some polishing touches to the background and voila! Fast and efficient construction using perspective is fully integrated into the appropriate setting.

Conclusion

Determining perspective in the presence of parallel structures is quite easy. Also, having studied the basics of perspective construction, you can now easily apply these methods to draw simple compositions. People who look at your work should see it from a perspective they are familiar with, something they have seen everywhere around them their entire lives. If the construction turns out to be sloppy, this can sometimes even lead a person into momentary confusion. If there is none of this, the eye will lack something to complete the picture. Therefore these basic principles should be carefully studied and then practiced drawing buildings. Just do everything so that all the lines converge at the desired vanishing points. You can even start with rectangular chairs or tables.

Well. I hope you have received material for reflection, so I leave the lesson for you to work through, and I ask you to post all questions and comments in the comments below on the merits.

Share the lesson

legal information

Translated from the site psd.tutsplus.com, the author of the translation is indicated at the beginning of the lesson.

Note.

In versions earlier than Photoshop CC, some of the functionality described in this article may only be available with Photoshop Extended. Photoshop CC does not have special version Extended. All the features of Photoshop Extended are available in Photoshop CC.

The Perspective Correction feature makes it easy to correct perspective in images that contain perspective planes, such as the sides of buildings, walls, floors, and any other rectangular objects. In this filter, the user specifies planes in the image to which editing (drawing, cloning, copying or pasting, and transforming) is then applied. All editing actions are performed taking into account the perspective of the work plane. When retouching, adding elements to an image, or removing parts, the results look more realistic because the adjustments are correctly oriented and scaled according to the perspective of the plane. Once you've finished correcting the perspective, you can continue editing the image in Photoshop. To preserve perspective plane information in an image, save the document in PSD, TIFF, or JPEG format.

Editing perspective image planes

You can also measure image elements and export these 3D characteristics and measurements to DXF and 3DS formats for later use in 3D graphics applications.

The Perspective Correction dialog box (Filter > Perspective Correction) contains tools for defining perspective planes, image editing tools, and the Ruler tool, as well as a preview area. The capabilities of the tools in the Perspective Correction function (Area, Stamp, Brush, etc.) are the same as the capabilities of the corresponding tools in the Photoshop Tools palette. Tool options can be adjusted using the same keyboard shortcuts. Opening the Perspective Correction menu displays additional tool options and commands.

Correct Perspective Dialog Box

Perspective Correction Tools

The perspective correction tools work similarly to their equivalents in Photoshop's main tool palette. Tool options can be adjusted using the same keyboard shortcuts. The tool you select affects the options available in the Perspective Correction dialog box.

Edit Plane Tool

Selecting, editing, moving and resizing a plane.

Create Plane Tool

Defines the four corners of a plane, adjusts its size and shape, and creates a new plane based on this data.

Region Tool

Select square or rectangular areas. You can also use this tool to move or clone selected areas.

Double-clicking the Marquee tool on a plane selects the entire plane.

Stamp tool

Drawing using a sample image. Unlike the Stamp tool in Photoshop's Tools palette, the Perspective Correction filter's Stamp tool can't copy elements of another image. See also Painting with pixels from a pattern in Correcting Perspective and Retouching with the Stamp Tool from the Photoshop Tools panel.

Brush Tool

Coloring the plane with the selected color.

Transform Tool

Eyedropper tool

Select the color to draw when you click in the preview area.

Zoom tool

Zoom in or out on the preview window.

Hand Tool

Allows you to move the image in the preview window.

Enlarge or reduce the viewing image

Move the image in the preview window

• In the Perspective Correction dialog box, select the Hand tool and drag the pointer in the preview area.

  Select any tool and, while holding down the Spacebar, drag the pointer in the preview area.

1.(Optional) Prepare your image for the Perspective Correction feature.

Before choosing the Correct Perspective command, do one of the following:

In order for the results of work with the Perspective Correction function to be placed in a separate layer, this layer must be created in advance. Saving the results of a perspective correction in a separate layer allows you to preserve the original image and adjust the opacity, styles, and blending modes for that layer.

If you plan to paste an element from the Photoshop clipboard, you must copy the element before selecting the Correct Perspective command. The element you're copying may be in another Photoshop document. If you are copying text, you must rasterize the text layer before copying it to the clipboard.

To ensure that perspective correction is applied only to specified areas of the image, you must select those areas or create an image mask before selecting the Perspective Correction command. See also Selections with the Marquee Tool and About Masks and Alpha Channels.

For a perspective object to be copied from one Photoshop document to another, it must be copied into the document in perspective correction mode. When you paste this element in Perspective Correction mode into another document, the object's perspective will be preserved.

2. Choose Filter > Perspective Correction.

3. Mark four corner nodes on a flat surface.

By default, the Create Plane tool is selected. To specify corner nodes, click the image in the preview area. When creating a plane, it is recommended to use a rectangular object as guides.

To create additional planes based on the specified parameters, use the Create Plane tool and Ctrl-drag (Windows) or Command-drag (Mac OS) a corner node. For more information, see Define and adjust perspective planes in the Perspective Correction window.

Using the Create Plane tool, specify four corner nodes.

To construct a plane based on the specified parameters, drag a corner node while holding down Ctrl (Windows) or Command (Mac OS).

4. Edit the image.

Perform one of the following actions.

Select an area. Once a selection has been created, you can clone, move, rotate, scale, transform, and also apply a fill to this area. For more information, see .

Paste an element from the clipboard. The inserted element becomes floating area, the perspective of which corresponds to the perspective of the plane into which the element is moved. For more information, see Inserting Elements in the Perspective Window.

Drawing with color or pixels according to a sample. For more information, see Paint with color in the Perspective Correction window or Paint with pixels using a pattern in the Perspective Correction window.

Scaling, rotation, mirror reflection and flipping the floating area. For more information, see About the selections in the Perspective Correction window.

Measuring an element on a plane. To display measurements in Photoshop, there is a Display Measurements in Photoshop command in the Perspective Correction menu. For more information, see Measurements when working with the Perspective Correction window.

5. Export 3D graphics characteristics and measurements to DXF or 3DS format.

6. Click the button OK

Before you click OK, you can display the grids by using the Show Grids in Photoshop command in the Perspective menu. For more information, see Displaying Grids in Photoshop.

3D graphics properties (planes), textures, and dimensions created by working with the Perspective Correction filter can be exported to formats used by CAD, modeling, animation, and special effects applications. Exporting to DXF format creates a file with 3D graphics characteristics and all measurements. Along with geometry characteristics, exported 3DS files contain rendered textures.

Open the Perspective Correction menu and select Export to DXF or Export to 3DS.

In the Export to DXF or Export to 3DS dialog box, select a location for the saved file and click Save.

Before you begin editing with the Perspective Correction filter, you need to define rectangular planes that match the perspective of the image. The accuracy of the plane determines how correctly all adjustments and parameters of a given image will be scaled and oriented.

Once the four corner nodes are defined, the perspective plane becomes active and the bounding box and mesh are displayed. Scale, offset, and reshaping are used to fine-tune the perspective plane. Using the grid options, you can align it with the elements of the image. Sometimes aligning the bounding box and mesh with the texture or pattern of the image helps fine-tune the perspective of the image. Adjusting the grid cell size makes it easier to count image elements.

In addition to its auxiliary role in aligning perspective planes with image elements, the grid allows you to visualize measurements when working with the Ruler tool. There is a special option to associate grid dimensions with measurements taken with the Ruler tool.

Creating Linked Perspective Planes

After creating a plane in the Perspective Correction window, you can create additional planes with the same perspective. After creating a second plane from the initial perspective plane, you can create additional planes from the second, and so on. You can create any number of planes. New planes are created at an angle of 90°, but they can be rotated to any angle. This is useful for precisely editing planes to repeat geometry complex scene. For example, corner cabinets in the kitchen can be part of the same plane. In addition to tilting planes, you can always change their size using the Edit Plane tool.

Bounding box and grid warnings in the Perspective Correction filter

The color of the bounding box and grid lines changes according to current state plane. If the plane is invalid, then the corner node should be moved until the bounding box and grid lines turn blue again.

Blue

Acceptable plane. It should be remembered that acceptable planes do not guarantee that the desired perspective correction results will be obtained. Make sure that the bounding box and grid are accurately aligned with the geometric elements or flat area of ​​the image.

Red

Invalid plane. The Perspective Correction filter is unable to calculate plane proportions.

Yellow

Note.

Although invalid red or yellow planes can be edited (such as tearing off perpendicular planes), it is difficult to get the results oriented correctly.

Show or hide the grid, active selections, and perspective plane boundaries

From the Perspective Correction menu, choose Show Edges.

Note.

While resizing or repositioning, the borders of the selected areas are temporarily displayed, even if the Show Edges option is turned off.

Setting perspective plane grid cell sizes

By default, when you view an image in a Photoshop document window, the Perspective Correction filter grids are not visible, although they are saved in the image and appear each time you run the Perspective Correction filter. There is a grid display option that will appear in the Photoshop document window after you've finished working with the Perspective Correction filter. The displayed meshes are raster and not vector.

Open the Perspective Correction menu and choose Show Grids in Photoshop.

The Display Grids in Photoshop command should be selected in each session of the Perspective filter.

If you plan to display grids in Photoshop, you will need to create a new layer for the results of the Perspective Correction filter. This way the meshes will be stored in a layer separate from the main image.

When painting or retouching, highlighted areas allow you to correct imperfections, add elements, or enhance an image. In perspective correction mode, creating selections allows you to paint or fill specified areas of the image while maintaining the perspective defined by the image planes. Using selections, you can clone and move specific elements of an image in perspective.

The Marquee tool creates a selection within a perspective plane. If the selected area extends over several planes, then it is divided in such a way as to match the perspective of each of them.

The generated selection can be moved to any place in the image, respecting the perspective established by the plane. If the image contains several planes, then the selected area is adjusted to the perspective of the plane into which it is moved.

The Perspective tool allows you to clone the pixels of a selection as you move that area around the image. When working with the Perspective Correction filter, a selected area whose pixels can be moved to any part of the image is called floating area. Even though the floating area pixels are not on a separate layer, they appear to be a separate layer hanging over the main image. While the floating area is active, you can move, rotate, or resize it.

Note.

When you paste an element while using the Perspective Correction filter, the pasted pixels are in a floating area.

Clicking outside the floating area deselects the selection. When you deselect, the contents of the floating area are placed into the image, replacing the pixels that were previously under the selection. Making an exact copy of the floating area also cancels the original selection.

Inserted element in Perspective Correction.

The Perspective Correction filter has another option to move selections. The selected area can be filled with pixels from the area where the pointer is moved.

Copy and move a selection from one perspective plane to another

Selecting areas in the Correct Perspective filter

Move selections in the Correct Perspective filter

To determine how the selection behaves when moved, from the Move Mode menu, choose one of the following:

• To select the area where the selection area moves, choose Destination.

  To fill the selection with the pixels of the area where you drag the Selection tool pointer (similar to dragging a selection while holding down Ctrl or Command), select Source.

Drag the selection. Hold down the Shift key while constraining the movement so that it is aligned with the perspective plane grid.

Move, rotate, and scale floating areas

Select an area in the perspective plane.

Alt-drag (Windows) or Option-drag (Mac OS) a selection using the Marquee tool to create a copy of the selection and its pixels.

The copy becomes floating area, which appears to hover over the main image. You can move the floating area and use the Transform tool to scale or rotate it.

Perform one of the following actions.

• Click outside the floating area to deselect it. The contents of the floating region are placed into the image, replacing the pixels that were previously under the selected region.

  Click inside a floating area with the Marquee or Transform tools and Alt-drag (Windows) or Option-drag (Mac OS) to create a copy of the selection. Immediately after the copy is created, the original floating area is deselected, and that area replaces the pixels that were previously underneath it.

To repeat the move of the last copy operation, press Ctrl+Shift+T (Windows) or Control+Shift+T (Mac OS). This way you can easily create exact copies content.

When working with the Perspective Correction filter, you can paste elements from the clipboard. The element you're copying can be in the same or a different Photoshop document. As soon as you paste it into the Perspective Correction window, the element becomes a floating area that you can scale, rotate, move, or clone. When you move to a selected plane, the floating region adjusts to the perspective of that plane.

Inserting an element when working with the Perspective Correction filter

Copying an element to the clipboard. The element you're copying can be in the same or a different Photoshop document. Please remember that insertion is only possible for raster (non-vector) images.

Note.

If text is copied, it must first be rasterized. Right-click on the text layer and select Rasterize. Then choose Select > All and copy the layer to the clipboard.

(Optional) Create a new layer.

Choose Filter > Perspective Correction.

If necessary, create one or more planes in the image.

To paste an element, press Ctrl+V (Windows) or Command+V (Mac OS).

The inserted element is now a floating area in the top left corner of the previewed image. By default, the Marquee tool is selected.

Using the Marquee tool, drag the inserted image onto the plane.

The image is adjusted to match the perspective of the plane.

Note.

After you paste an image into the Perspective Correction window, do not use the Marquee tool on the image other than to drag the pasted image into the perspective plane. Clicking anywhere in the image deselects the floating area. In this case, pixels are permanently inserted into the image.

Select the Brush tool.

Set the brush color. To do this, do one of the following:

• Select the Eyedropper tool and click any color in the image you're viewing.

  Click the Brush Color area to open the color picker and select a color.

In the Tool Options area, adjust the settings for Diameter (the size of the brush), Hardness (the smoothness of the edge), and Opacity (how much the paint layer obscures the image underneath).

Select "Recovery" mode:

• To draw without overlaying color, light, or shadowing surrounding pixels, choose Off.

  To paint with a light overlay of surrounding pixels, leaving the selected brush color, choose Luminosity mode.

  To paint by overlaying the color, light, and shadow of surrounding pixels, choose On.

Specify drawing options (optional).

• For continuous drawing with automatic tuning To match the perspective of one plane to the perspective of another, open the “Perspective Correction” menu and select the “Allow multi-surface operations” command. If this option is turned off, you can only draw one plane in perspective at a time. To switch perspective, you need to stop drawing and continue it in another plane.

  To make the drawing consistent with the perspective of only the current plane, open the Perspective Correction menu and choose Perform Clipping Operations to Edges of Surface. If this option is disabled, you can draw in perspective outside the current plane.

Drawing is done by dragging the pointer. As you paint on a plane, the size and shape of the brush scale and change orientation to match the perspective of the plane. Moving the pointer while holding down the Shift key limits the brush strokes to a straight line corresponding to the perspective of the plane. You can use the Brush tool to specify a point on an image and Shift-click another point to draw a straight line connecting those points using perspective.

04/19/2011 A. F. Afanasyev Updated 08/11/12

Building a Perspective

Perspective refers to the image of the real objective world on a plane as it is perceived by the human eye. It is divided into two types: geometric and physical, which artists call color or air.

Geometric perspective a section of descriptive geometry, where the laws of image on a surface are studied using the lines of volumetric objects, the sizes of which decrease with increasing distance to the viewer as it is perceived by the eye.

Color perspective studies the change in tone (color) of an object depending on distance and influence environment: lighting, weather, neighboring tones, etc.

Geometric perspective is divided into linear perspective, when the image is built on a plane, panoramic, if it is made on a cylindrical surface, and dome, obtained on inner surface domes, for example spheres, ellipsoids.

We will only consider linear perspective. It has its own strict geometric rules, without knowledge of which constructing a picture “in depth” is impossible.

Linear perspective. Let's draw a line for the base of the painting and a horizon line (Fig. 198), which is taken at the level of the artist's eye (which means that in a sitting position the horizon line will be lower). All parallel lines perpendicular to the base of the picture are depicted converging at point P, located on the horizon line. If parallel lines are inclined to the line of the base of the picture, then their vanishing point F will be shifted to the left or right from point P, i.e. from the middle of the horizon (Fig. 198, b). Point P is called the main point of the picture.

If we draw parallel lines in plan, perpendicular to the base of the picture, and parallel lines inclined to it at an angle of 45°, then the inclined lines will cut off identical segments on the base of the picture and on the lines perpendicular to it (Fig. 199, a). We will have to recognize the same rule (Fig. 199, b) if we depict these lines in perspective (an example of segments of equal length is shown in both drawings with a thickened line).

It remains to explain how to find in perspective the vanishing point D of parallel lines inclined to the base of the picture at an angle of 45°. Point D is called point range or dot distance, it is laid off from the main point P to the left or right along the horizon at a distance equal to the distance of the point of view (S) from the picture. The distance of this point is chosen arbitrarily by the artist within the range from 1.5 to 2-2.5 diagonals of the picture and does not change during further construction. Thus, points P and D are special points in perspective. With their help, a number of constructions are made.

So, for example, if parallel lines converging at the main point P divide the base of the picture into equal segments (marked with numbers 1, 2, 3...) (Fig. 200), then parallel lines drawn through these points of the base and converging at range point D, will cut off on the first straight lines the same segments equal to them, but depicted in perspective. By drawing straight lines through the ends of these segments, parallel to the base of the picture, we obtain a perspective image of a horizontal surface dissected into squares.

Having divided the horizontal plane of the picture into proportional, i.e., perspective, dimensions, we can construct a series of vertical segments located at equal distances (in space) from each other, taking, for example, the value AB as the natural size of this segment in the plane of the picture (see Fig. 200). This construction can be done anywhere in the picture plane. It is clear that the perspective value of the vertical segment will not change if it is moved along a line parallel to the base of the picture.

As we can conclude from the drawing, the degree of contraction of horizontal segments perpendicular to the base of the picture depends on the level of the horizon line and on the location of point D, i.e., on the distance of the eye to the picture. The degree of reduction of vertical segments also depends on this. Since the range point on the field of the picture does not always fit, it is necessary to temporarily increase the width of the picture with additional sheets to the left and right. But you can do without this, if you take into account the construction of perspective shown in Fig. 201. Using straight lines drawn through points P and D and points 1, 2, 3..., we cut the horizontal plane in the picture into 16 squares. Let us set aside half and a quarter of the distance from point P to range point D. By connecting points 1/2D and 1/4D with point 1, we notice that straight lines pass through points B and E. In this way we can obtain the perspective of the square OBF2. Let's draw its diagonals and get the perspective vertex of the original square (one of 16), which we took as the standard at the beginning of construction.

In turn, the perspective of the square OBF2 can be obtained using the diagonals of the square OEC4. From here we conclude that the perspective of a checkerboard floor in the form of squares or rectangles (double squares) can also be constructed using half or a quarter distance from the main point P to the range point.

Rice. 202. Incorrect way of constructing the perspective of rectangles using their parallel diagonals (for example, the perspective of rectangles on both sides of the sun does not match)

Rice. 203. The correct way to construct perspective using the diagonals of rectangles converging at a point on the vanishing line

Rice. 204. Constructing the perspective of points and any figure in the horizontal plane using a plan and a perspective grid

Let us pay attention to the fact that the diagonals of squares on a horizontal field are lines inclined to the base of the picture at an angle of 45°; in perspective, they converge at the point of distance D, i.e. in perspective, the diagonals of squares or rectangles (whose sides are parallel to the plane of the picture ) cannot be parallel. Therefore, it would be incorrect to construct a perspective using diagonals of rectangles parallel to each other, as shown in Fig. 202 (this construction is sometimes found in the practice of amateur artists). Only those parallel lines that are parallel to the plane of the picture remain parallel in perspective.

In Fig. 203 shows the correct, simplified construction of the perspective of rectangles, if one of them is taken as a standard, taken by eye. In this case, there is a vanishing point of one of the diagonals of the rectangle, located on the vanishing line of the figure’s plane (in the case under consideration, both vanishing lines, for both the vertical and horizontal planes, pass through point F; it does not have to be the main point P). All diagonals of the remaining rectangles of this plane will intersect at this point.

Each of the rectangles can be divided in half by a line passing through the point of intersection of its diagonals (line EO is outlined in Fig. 203); the diagonals of the new rectangles will have their point of intersection on the same vanishing line.

It is clear that the perspective of rectangles of arbitrary size can be obtained by dividing a large rectangle in half using its diagonals, and then further dividing the resulting halves, but their number will be: 2, 4, 8, 16, 32... Division by any number Equal parts of each of the rectangles can be made by dividing the side parallel to the plane of the picture. The resulting points are connected to the vanishing point, and the lines are intersected by the diagonal of this rectangle. A grid of equal rectangles similar to their common rectangle is formed. If the diagonals of the rectangles have a vanishing point at the range point D, then they are squares. This means that when the point of view (distance to the picture) changes within the accepted conditions (from 1.5 to 2.5 diagonals of the picture), each rectangle can become a square for the plane in question.

In order to construct any point (and therefore a figure) in perspective, you can use a perspective grid. In Fig. 204 shows a plan showing a square grid and triangle ABC lying in its plane. Given points P and D in perspective, we will construct a perspective square grid. To do this, from point D it is enough to draw one common diagonal of squares and, through the points of its intersection with the lines converging at point P, draw straight lines parallel to the base of the picture.

Points A, B, C in perspective are constructed at the intersection of the corresponding grid lines or between them. If necessary, the mesh is made finer in these places. You can clarify the position of points A, B and C (as well as any other point if the figure is complex) as the intersection of a line perpendicular to the base of the picture and an auxiliary line drawn through this point at an angle of 45° (see construction of points A and K). In perspective, the auxiliary line will pass through point D.

Rice. 205. Constructing a perspective of a wall using its facade of any scale

Rice. 206. Simplified construction of an ellipse using the points of tangency of the sides of a square

Rice. 207. Constructing an ellipse in frontal perspective using auxiliary points, for example 5 and 6

Rice. 208. Construction of an ellipse in angular perspective. Example unsuccessful use linear perspective to build a picture

Rice. 209. Construction of an ellipse in the vertical plane

Rice. 210. Method of “wrapping” a surface for constructing complex volumetric figures in perspective

Let us pay attention to the fact that straight line AC intersects with the base of the picture at point M - the same for plan and perspective (as well as other straight lines).

Let's construct a perspective of a figure located in a vertical plane. In Fig. 205 the wall of a room is taken as an example. Let the perspective of the wall be determined in the painting (the height is given, the width is determined by the construction of the floor). Let's draw the façade of the wall on any scale. Having set aside points 1, 2, 3... on the basis of the picture, reflecting the proportional distances between the elements of the wall, we connect the extreme point 9 with extreme point walls (point B) to the intersection with the horizon line (point F 1). Using point F 1, we divide the base of the wall into the perspective proportions of its elements.

In terms of height, the direct proportionality of dividing the wall into given segments (levels of windows, doors) is preserved, so here you can use any straight line inclined to the sun. For this purpose, straight line CF was used, on which points 9, 10, 11, 12 are marked. Straight lines parallel to 9B will determine the level of windows and the height of the door on the wall.

In Fig. 206 shows a simplified construction of an ellipse, which is a perspective of a circle lying in a horizontal plane. To do this, we first construct the perspective of a square in which a circle is inscribed. Having drawn the diagonal of the square, we find point K, which will determine midline square 3-4 and points 3 and 4 touching the circle of its sides. Knowing the minor axis of the ellipse 1-2, the direction of the major axis (in the middle of the minor axis) and at least one of the points of the ellipse (3 or 4), you can find the size of the major axis and construct the entire ellipse (see p. 221 of this chapter).

Constructing an ellipse using required quantity additional points are shown in Fig. 207. Here we use a plan of half of a circle depicted in perspective. Point 5 was obtained as belonging to the diagonal of the square; for point 6, an auxiliary straight line 1-7 was constructed (further construction is shown in the figure). Similarly, you can obtain others necessary for constructing a point.

In Fig. 208 shows the construction of an ellipse in the horizontal plane, offset relative to the central axis of the picture, i.e. in angular perspective. In drawing and painting, ellipses are not made from this perspective (explanations follow). We use the above drawing to practice constructing additional points of the ellipse. It is better to avoid crossing construction lines at an acute angle, which gives inaccuracy (example with point 7). Therefore, for point 8 we carried out auxiliary line, passing through one of the already found points of the ellipse - point 5. Its construction is shown in the drawing. For point 7 it would be more convenient to use a straight line passing through point 4.

The construction of an ellipse in the vertical plane (Fig. 209) is, in principle, no different from that described above. In addition to the main points lying at the midpoints of the sides of the square (1, 2, 3, 4), and points belonging to its diagonals (5, 6 and two points paired with them), an additional point 7 was found using a straight line drawn through point 5. In The construction uses a semicircle located in the frontal plane.

To build a complex perspective volumetric figure you can use the so-called wrapping surface (Fig. 210), when the figure is limited by vertical and horizontal planes with a perspective grid so that a parallelepiped is obtained, which is then constructed in perspective.

In conclusion, it should be noted that the application of perspective rules must be balanced with the perception of the eye and some constructions must be avoided. In Fig. 208 shows how unnatural an ellipse will look if the distance of the eye from the picture is too small. But even with the accepted distance ranging from 1.5 to 2.5 diagonals of the picture (or according to other data: the viewing angle should be in the range of 28-37°), sometimes the construction of perspective is not consistent with visual perception. According to the data cited by M. F. Fedorov in his perspective analysis of many classic paintings, artists who knew perspective very well resorted to deliberate violation of its rules. It was expressed in its main features in the following: in the smooth curvature of straight lines towards the vanishing point on the horizon, in the use of several vanishing points for objectively parallel straight lines, in the exaggeration of the size of objects in the background. This is explained by the fact that in nature we do not perceive reality like a camera lens. Due to the so-called “relative constancy of perception” of the human eye, we psychologically equalize the sizes of distant objects and nearby ones. Therefore, the artist depicts the outstretched hand of a person in the foreground of the picture as large as it would appear in a photograph, just as the head of a horse, when viewed from behind, will not be as small.

When examining objects, we rotate the axis of the eye, i.e., the eye perceives each object as if in a frontal rather than an angular perspective. This explains some leveling slanted lines along the edges of the picture towards the horizontal. The same is true with ellipses, into which the bases of bodies of rotation turn in perspective: we do not notice the deviation of the major axis of the ellipse from the direction perpendicular to the axis of the body. Artists depict all ellipses in horizontal planes straight, that is, with the major axis parallel to the base of the picture. Likewise, a ball in perspective always remains a ball, and is not projected in the form of an ellipse, which is supposed to be done according to the rules or can be seen in a photograph.

The artist also does not limit the angle of view to the limits of the most visual perspective, but sometimes increases it to 70-90°, correcting the perspective with real perception.

From this we can conclude that you need to know the perspective and use it to build and verify your visual perception, and ultimately “the object must be depicted as it appears to our eyes and as it really is.”

As an example of the practical use of the rules outlined in Fig. 211 the construction of the perspective of a five-pointed star is given. Having a plan of the star, let’s define points P and D (or construct the desired contour of the square in perspective and use it to determine points P and D). Having found point O in perspective, we will construct the perspective of line 1-2 and on it points A and B using straight lines perpendicular to the base of the picture and their perspectives. Point B in perspective will determine the position of a straight line parallel to the base of the picture, which will make it possible to find point C. It remains to construct other points, symmetrical to A and C, and connect them into an outline of the desired perspective of the star.