The specifics of the landscape image Automatic translate

The examples discussed in previous chapters related mainly to the task of depicting the interior. It would seem that the analysis of showing open landscape will add little, but this is not so. Landscape painting has its own specifics - there are almost no problems with the image of walls and ceilings, but the depth of space is not limited. That is why, when discussing methods of depicting space to the horizon, it is appropriate to discuss the question of how it is most natural to convey the appearance of objects filling it (houses, trees, etc.) located on different planes. The results obtained will help to understand some features of the image of objects inside the interior.

The main problem, as before, is the discussion of the methods of the most accurate transfer on the two-dimensional plane of the picture created by the brain’s work of the subjective three-dimensional space of visual perception. Here, the general law will continue to operate: an adequate transfer of subjective space in the picture is in principle impossible (we will discuss further specific exceptions to this general rule). Therefore, the problem of choice continues to exist here - which image elements to shift the inevitable distortions, and therefore, the choice of the most natural option for the solvable artistic problem of a single system of perceptual perspective.

For a comparative assessment of possible options for promising constructions, we turn to schematic images of a certain conditional landscape. In fig. used the option in which the depth and width are perfectly transmitted. In other words, the main thing here is the error-free transmission of the earth’s surface, and this main thing is realized. At the same time, mountains on the horizon are flawlessly transmitted. Very often, artists face just such a task. To appreciate the features of this method of transmitting open space, it is useful to compare it with the same conventional landscape, but executed according to the rules of the renaissance version of the perspective system (or, what is the same thing, compare it with a photograph). Even a cursory comparison shows that the photographic perspective greatly distorts the transmission of natural visual perception. A noticeable decrease in the mountains on the horizon, a slight decrease in the horizon, as if the compression of the distant part of the space shown in the diagram (the distance from the distant tree to the mountains) are striking. This compression and reduction of the displayed objects does not apply to the entire image. In the foreground, an extended object (garden) is visible, the dimensions of which are increased. Consequently, the renaissance version of the perspective system reduces the size of distant ones and increases the size of nearby areas of the earth’s surface. Today, some artists use all sorts of photographic materials instead of nature; they must understand that photography often does not bring the image closer to the correct one (that is, the corresponding visual perception), but, on the contrary, moves it away from it.

Any sufficiently saturated image should, as you know, contain perspective errors. As in the example image of a conventional interior, they are shifted by the size of the verticals. The height of the trees shown in the diagram is increased (front - by 17%, and back - by 10%). The height of the house is also increased, at about the same scale. Since the width of the house is conveyed correctly - after all, the used version of promising constructions correctly conveys depth and width - its visible appearance is distorted (returning to the terminology used earlier, “image resemblance”). The configuration (silhouette) of the house was transferred with an error. In fig. 19, which is based on a Renaissance perspective system, the configuration of the house is conveyed quite correctly. There are no similarity transfer errors. This was already mentioned in the discussion of the drawings of the conditional interior. But this does not at all follow that the verticals are correctly shown. The front tree is increased by 12%, and the rear tree is reduced by as much as 30%. As can be seen, in this case, the vertical transmission is distorted, but if in Fig. 18 they are everywhere enlarged, then in fig. 19 - enlarged in the foregrounds and greatly reduced in the far planes.

Possible options for a single perceptual system are not limited to the perspectives, of course, of the two shown in the figures. In fig. 20 a variant of perspective constructions is given, without distortion conveying similarity and depth on all planes and, therefore, with an error - width and height. Correctly transferred mountains on the horizon. The image will not be commented - everyone is able to do it himself. I would just like to note that in many ways it is better than drawn according to the Renaissance rules. In fig. another variant of perspective constructions is given, correctly conveying the width and height (and, consequently, the similarity), but erroneously - the depth. Here, the depth is transmitted so weakly, and the space is so compressed that most likely this option will not attract the attention of artists. Perhaps the only positive quality of this image is the correct vertical transmission: the trees here are shown flawlessly of the correct height, and the mountains are shown correctly.

The four options discussed do not at all exhaust all possible schemes of perspective constructions in landscape painting that follow the laws of the perceptual perspective system. In relation to this topic, it is appropriate to say a few words about the renaissance version of the scientific perspective system. As we have already shown, it cannot be a standard of absolute correctness of a promising pattern (without taking into account the work of the brain). This allows us to consider the Renaissance version, capable of only partially conveying natural visual perception, as one of the options for a unified scientific system of perspective along with others. But it turned out that a mathematically accurate account of the work of the brain is also not able to create an absolute standard of drawing. Moreover, it was established that the sum of errors in transmitting the appearance of the interior of these systems is the same. Therefore, we agree further to consider the Renaissance version as an equal version of the perspective system, which differs from others in the special nature of the distribution of errors, which often forces artists to abandon it.

The last statement can be illustrated by the example of Cezanne landscape painting. As you know, many of Cezanne’s landscapes were compared with photographs taken later, and this made it possible to document that Cezanne did not consider the Renaissance perspective system worthy of use. In this regard, F. Novotny called his monograph as follows: “Cezanne and the end of the scientific perspective” [Novotny, 1938]. E. Laurent came to similar conclusions [Loran, 1943]. To consider this issue from the perspective of the perceptual system of perspective, we turn to the corresponding visual material. To simplify the analysis, not reproductions of paintings and photographs of motives will be used, but their lines, in which everything that is not of interest for evaluating promising systems is omitted.

In fig. the eruption of Cézanne’s painting “Provencal Farm near Gardan” is shown, and under it is the eruption of the corresponding photograph. The calculations needed to compare them can be found in my book on the general theory of perspective (Rauschenbach, 1986); here they are omitted. If we compare the cuts with the images of the conditional landscape, then immediately their kinship becomes apparent. They show that Cezanne sought to flawlessly transmit the surface of the earth, in other words, to an adequate image of depth and width, and precisely followed his visual perception. Therefore, the size of the hills in the background is larger than in the photograph, and the horizon (compared to the photographic one) is increased. We should talk about reducing hills and lowering the horizon in the photo, since Cezanne conveyed this correctly, but the usual terminology is preserved here. The numerical evaluations convinced that this is not artistically justified liberties of the master, but the desire to convey the main thing in his picture - the surface of the earth - such as it is seen by a person. In conventional landscapes, the distance from the house to the foot of the mountains in the figure, using the Renaissance version of the prospective construction, is 80% of the same distance in the figure, correctly conveying the surface of the earth. For Cezanne, this ratio, as everyone can see, has the same value - which means that he transmitted the surface of the earth correctly. The differences between the Cezanne landscape and the conditional landscape are due to the fact that the foreground on the conditional landscape begins closer to the artist than Cezanne.

As has been repeatedly emphasized, any scientific system of perspective contains errors. In that version that Cezanne used and which corresponds to fig. 18, the errors, as already mentioned, are shifted to the vertical image (they are enlarged). It is extremely interesting and unexpected that Cezanne, who did not know anything about the perceptual system of perspective, nevertheless strictly (not only qualitatively, but, which may seem surprising, and quantitatively) follows its rules. The main verticals are the heights of the houses, and they, compared with the photograph, are clearly enlarged, and by the value predicted by the theory of perceptual perspective. At the same time, he is forced to introduce a similarity transmission error into the picture: the configuration of buildings (the ratio between height and width), which any photograph conveys completely correctly, is incorrectly transmitted to him. This indicates that the rules of the perceptual system of perspective are not arbitrary constructions, but reflect some essential aspects of artistic creation. In this case, Cezanne, but in all likelihood, considered it important to keep on the picture the ratio of the distance from the horizon to the house and his height that he observed in kind. Simple measurements show that in the picture and in the photograph, the distance from the roof of the house to the horizon is about twice the height of the house. It can be said differently: the distance from the house to the horizon in the image of the conditional landscape using the Renaissance version of promising constructions, as already mentioned, is 80% of the same distance in the figure following the “brain picture”, but the corresponding ratio of the heights of the houses is also equal to 80%.

The considered example is interesting in that it shows how an artist who sincerely seeks to accurately convey his visual perception begins to build space in his picture, choosing the appropriate version of a single perceptual perspective system. It turns out that for this it is not at all necessary to know its mathematical basis, it is enough to follow your visual perception and put up with the fact that somewhere, as if by itself, an image error will occur. Previously, this was shown in the analysis of a promising scheme that determined the construction of the interior of the church in the painting of V. D. Polenov, now it is based on the example of the Cezanne landscape. It will be shown later that both ancient and medieval art are not exceptions in this sense: even then the artists intuitively followed their visual perception and, accordingly, came quite clearly to a suitable version of a unified system of scientific perspective.

Returning to the work of Cezanne, it should be noted that on his other canvases one can observe a similar desire for the correct transfer of the surface of the earth. This is evident even from an analysis of his paintings “View of the Estacus” and “The Chestnut Alley in Jade de Buffon”, cited in the already mentioned book [Rauschenbach, 1986]. In many cases, it strikingly accurately follows (of course, intuitively) the version of the scientific perspective shown in Fig. 18. Perhaps the book of Novotny should be called “Cezanne and the beginning of a scientific perspective,” because Cezanne’s work does not at all mean its end. Attempts to understand the promising constructions of Cezanne, relying on the rsnaissance system of perspective, are completely meaningless. He did not pay the slightest attention to it, working in an equally scientific system of perceptual perspective, in that version that was more in line with his artistic plan.

Relying on the renaissance version of the perspective system as the only scientific one, many researchers argued that Cezanne increases long-range plans. Now, as already mentioned, the opposite can be said: Cezanne conveys long-range plans correctly, photography makes mistakes - it reduces them. However, it would be wrong to say that Cezanne never distorts the geometry of visual perception. If this is required by the artistic task that he is solving, he calmly admits distortions. In fig. given the breach of the painting "Mount St. Victoria from the quarry of Bibemus." The silhouette corresponding to the photograph is shaded, and the dashed silhouette reflects natural visual perception (that is, the shaded one corresponds to the Renaissance version, and the dashed silhouette to the perceptual versions of the general scientific perspective system). The solid line shows the contour of the mountain chosen by Cezanne. As can be seen from a comparison of the three contours, Cezanne emphasized increased (in comparison with the natural perception) the size of the mountain, based on the artistic problem that he solved.

The general consideration of promising constructions given above should be supplemented by the study of images of individual objects on different planes. The findings will be true for interiors. I must say that a single subject, if it is shown on different planes, will have a different set of errors. To evaluate each of its images, one can again consider the errors of transmitting depth (image of the length of an object deep into space), scale (the correct ratio of the dimensions shown for its front and remote parts) and similarity - but not for the entire extended space, but for a small image object.

Without delving into a detailed study of the observed patterns (this is done in the already mentioned book), we restrict ourselves to the most general provisions. In all variants of the perceptual perspective system, the largest image errors are concentrated in the foreground. On the middle plane, they are moderate in nature and on the long-term plan they practically disappear. Thus, without exception, all variants of the perceptual perspective system give a flawless image of very distant objects, such as mountains on the horizon. Unfortunately, at the same time to show what is in the foreground, too, as we see it, is almost impossible. An artist striving for an impeccable transfer of objects located in the foreground of the picture will experience enormous difficulties and will not be able to master such a task. It is not surprising that masters try to avoid this by placing objects in the foreground whose perspective distortions are insignificant (grass, shrubs), or starting the image from sufficiently remote areas of space, so that the close foreground is below the lower edge of the picture. The difficulties discussed here are clearly visible, for example, on the portrait of Mendeleev by Repin. The writing desk that separates the artist from the portrait is conveyed with severe distortions of natural visual perception, although the professional skill of the painter is undoubted.

As already mentioned, the general patterns described above do not correspond in all respects to the Renaissance perspective system, in which the depiction of foreground objects is just as difficult as in the variants of the perceptual perspective system, but there is another difficulty: long-range objects are transmitted with severe distortions. The main characteristic feature of distortions in the renaissance version of the scientific perspective system is that the objects of the foreground are greatly increased, while the objects of the foreground are reduced as much. Mountains on the horizon, which the perceptual system will transmit perfectly, in the Renaissance system become like inexpressive mounds, only the middle plan is transmitted almost undistorted. Therefore, the Renaissance version is very suitable for depicting objects without a foreground and a long plan, such as a ship on the high seas. The situation will change dramatically if in the background the mountainous coast is visible, and in the very front - a boat. In the renaissance system of perspective, the boat will turn out to be unnaturally huge, and the mountainous coast will turn, as already mentioned, into a totality of miserable hills.

The features of various systems described here are valid not only for deep spaces (ship at sea), but also for relatively shallow ones. This can be illustrated by referring to the picture of Cezanne "Spring and millstones in the forest of Chateau Noir." In fig. 26 are drawn in the outline of both the picture and the corresponding photograph. Cezanne painted his picture in his usual version of the perceptual perspective system, correctly conveying the earth’s surface. Therefore, it can be argued that the vertically standing cylinder shown in the middle plan (probably a spring fence), the millstones on the left in the background and the group of trees in the foreground are depicted by the artist with the correct ratio of their sizes.

If we turn to the drawing of photography and take our mind off of the fact that the trees grew a little from painting to painting the landscape, then the distortions inherent in photography (the Renaissance perspective system) are striking. The cylinders located on the middle plane of both images are exactly the same, while in the background the millstone in the background is noticeably reduced, and the group of trees in the foreground has “spread” in breadth. Both, as already mentioned, are the main type of distortion that is characteristic of the Renaissance perspective system.

It was previously shown and confirmed in the analysis of the canvas "Provencal farm near Garda-pa" that Cézanne’s version of the scientific perspective is characterized by increased vertical transmission. Consequently, the cylinder in the picture of Cezanne should have been higher than in the photograph, but this is not. Here it is once again appropriate to emphasize that the artist is not a slave to scientific systems and has every right to deviate from their laws. In the case under consideration, this is quite reasonable. There is no horizon image in the picture (which would be much higher than in the photo), and therefore, Cezanne, without increasing the height of the cylinder and thereby formally violating the relationship between the height of the cylinder and the distance from the cylinder to the horizon line (which is not in the picture), made an inaccuracy that is not visible to anyone.But then he used the opportunity to give a visible outline of the cylinder without distortion: the ratio between the height and width of the cylinder in the picture and in the photograph is the same. Thus, a formal violation of the rules (which are justified when transferring an object located in an open space, for example, a house in the painting "Provencal Farm near Gardan") only improved the image.

From the examples examined, it is seen how important it is to analyze the artist’s works, based on what we called the “brain picture”. Evaluation of the applied techniques for building prospects in this case gets an objective character. Thus, the work of Cezanne gives rise to reflection on the need for the artist to know the theory of perspective. After all, Cezanne followed his visual perception directly. In the same way, Chinese artists worked in antiquity when they painted amazing landscapes, conveying in them their admiration for the power of nature. It is interesting to note that medieval Chinese artists perfectly felt the difficulties that they might face when trying to depict a deep space, including its close areas. In their landscapes, the foreground is always hundreds of meters away from the artist.The theory of the perceptual system of perspective suggests that for all its variants with such a removal, all image errors become negligible and, therefore, conflicts between visual perception and the picture cannot occur. In those cases when there was a need to show closer areas of space, the Chinese resorted to an isolated display of medium or very close plans. Thus, for example, paintings appeared depicting animals (horses), in which neither the fore nor the distant plans are given, or the favorite plots “flowers and birds”, where there is no middle or long plan. The theory of perceptual perspective suggests that with an isolated image of plans, errors are practically absent. They arise when trying to convey on the plane of the picture a deep holistic space.The theory explains these features of Chinese art, but knowledge of the theory for the artist at that far time was not at all mandatory. If we talk about contemporary artists, the theory of perceptual perspective makes clear the difficulties that they encounter when they want to convey nature on their canvas undistorted, but they must follow their visual perception, and not strict rules. They must understand when errors can be avoided, and when not, and in the latter case, freely choose those elements of the image on which it is most reasonable to shift them.but they must follow their visual perception, not strict rules. They must understand when errors can be avoided, and when not, and in the latter case, freely choose those elements of the image on which it is most reasonable to shift them.but they must follow their visual perception, not strict rules. They must understand when errors can be avoided, and when not, and in the latter case, freely choose those elements of the image on which it is most reasonable to shift them.

In those cases when the artist is attracted by the display of a holistic deep space, the problem of errors becomes the most urgent. The approaches developed in this book allow one to study the questions of correspondence of the image depicted to visual perception that have never been touched upon before. On this path, one can analyze the works of those artists who strove to follow nature, without resorting to the doctrine of perspective, and those that can be attributed to the "prospectists."

In conclusion, it is appropriate to draw attention to another circumstance that makes the absolutely accurate transmission of natural visual perception almost impossible. It concerns mainly the peripheral parts of the picture. The fact is that, transforming the reflection of external space that has arisen on the retina of the eye, the brain transforms it in different ways: long-range plans are very stretched, and close ones are not. As a result, mutual displacements of the depicted objects occurring on different planes of the picture occur. This can be seen, for example, from a comparison of Fig. 18 and fig. 19. The described effect is clearly visible if you look at how much of the top of the tree shown in the foreground is projected into the mountains of the horizon. It should be remembered that Fig. 19 follows the image arising on the retina, and in fig. 18 the work of the brain is taken into account. Effect,the subject in question is almost irrelevant for the artist and the viewer, especially since the person is accustomed to such displacements, for example, due to binocular vision. Everyone knows from his experience that, looking alternately with his left or right eye, he observes the displacement of nearby objects relative to distant ones.