Antialiasing & Color mapping in V-Ray 1.5

Nov 2, 2020 Update
Illustration to the tutorial about V-Ray renderer universal settings, focused on anti-aliasing & color mapping concepts.
Advertisements

Versatility Myths

Hi everybody.

In this tutorial, we will look into the one of the fundamental and, probably, the most controversial and frequently raised subjects in the 3d rendering. It is the rendering engine setup. We will setup the V-Ray renderer for 3ds Max.

Among the many of 3d artists, especially the beginners, the two contrast myths about universal V-Ray setups exist.

One believes there are some simple setups, which are carefully hided from them by professionals by keeping the tuning secrets in a strict confidence. Only the chosen people know the secret checkboxes and values that allow creating realistic 3d renderings, which hit the awards and admiration of the all observers. That there supposedly are some settings, just like the "magic wand", which by checking some ticks in covert places lets to have the amazing results despite the actual quality of the 3d models, correct lighting setup or beauty and realism of the present scene composition.

Other think there is no any magic and all you need is the long and difficult work with the renderer parameters for each new scene. That firstly you need to run the rendering in a test mode, with the visible samples of the *V-Ray sampler *on and, by a sample map, long and boringly try the optimal settings to soften and eliminate aliasing; to loose the much time for trying than to start the competently tuned rendering. Then to begin the Irradiance map calculation with the enabled Show calc. phase option and to look intently at the Irradiance map sample passes, supposedly controlling the calculation process and many other shamanistic superstitions :)

Curiously enough, but there are no really universal one-size-fits-all setups. Like, actually, the separate settings for interior and exterior as well. But, there are some hints and tricks, on how to improve the quality of visualization in general. You will learn about them in this tutorial. There is nothing complicated here, everything will be explained in simple language. So, let's go!

Base V-Ray settings

At first, we have to make clear that there are starter settings, that may be beginning of the any rendering, for not to draw our attention each time to any possibly incorrect options. We got to fully concentrate on the actual scene setup. Only in some situations, when the rendering has some parasitic artifacts, we need some slight tuning to get rid of them. This is true not for V-Ray only, but is an absolute concept for any rendering application. To tell the truth, the correctly set lighting with a good scene geometry almost never give the unwanted surprises on the rendering, and in most cases you don't need to set anything but final resolution.

Just this like starter settings, that are really universal in 80% of situations, that truly simplify your work and embellish your renderings, will be discussed in this tutorial.

Getting started

So... Press the F10 button to bring the Render Scene dialog on screen.

Similarly, we can go to the Rendering option of the main 3ds Max menu, and choose the first Render item from the appeared there list.

Now we have the one of the most important tools and interfaces of the any 3d rendering artist opened.

Before beginning the work on the setting up V-Ray, we surely have to set it as a default rendering engine.

Just go to the tab named Common and open the Assign Renderer rollout. The items of this rollout allow setting the rendering engine for rendering, displaying the materials in Material Editor, and the renderer for the specific real-time or pseudo real-time engine for rendering the scene right in the viewport.

Click the ellipsis (...) button near the Production and near the Material Editor (if the lock button is released) and choose V-Ray from the showed list.

To set the V-Ray as a default rendering engine and to not do this each time 3ds Max starts, surely we need to press the lower big Save as Defaults button.

V-Ray Frame Buffer Rollout

The first thing that should be done to begin the actual V-Ray renderer setup is the activation of the specialized V-Ray Frame Buffer, which completely replaces and excels by functionality the integrated standard 3ds max frame buffer. It may be found at the same named rollout of the V-Ray Render scene window. To enable V-Ray Frame Buffer, the option Enable built-in Frame Buffer must be checked.

The using of the V-Ray Frame Buffer greatly eases the work with V-Ray by its functionality and convenience, as well as eliminates the possible misunderstandings with gamma correction.

In addition, it is important to leave active the Get resolution from MAX function. This step can free us from the need to constantly go to the V-Ray tab, V-Ray:: Frame buffer rollout and set the desired rendering resolution.

The V-Ray raw image file and Split render channels are responsible for forming the formats, which needed for post-production using the compositing software, and they will be learned from the separate tutorial on rendering the render elements passes and compositing the cool realistic final image from them.

Anyway, the post-production using the Render elements is an unusual way to obtain the final rendering and it has nothing to do with the universal V-Ray settings. That is why we may leave these functions this time.

V-Ray Global switches Rollout

The next in line is a V-Ray:: Global switches rollout, it contains the parameters of the global scene influence. They are: presence or disabling the displacement, reflections, refractions, old cameras, light sources, Sky map compatibility, global material override (see the tricks of Wireframe Rendering in V-Ray), and so on.

This rollout is correctly configured by default and there is nothing needs to be changed. The one exiting parameter that may get one's attention is the Reflection/refraction parameter. This option is responsible for the global activation or deactivation the reflections and refractions in the scene materials. Even if the materials have the tuned reflections or refractions, the deactivation of this parameter will lead to complete absence of those.

Most interesting here is the Max depth parameter. This function globally sets the number of the secondary reflections or refractions in the scene. It is similar to the so-named parameter in the *VRayMtl *material. Using this function, we can lower the maximum number of the repeated reflections and refractions of the ray that is reflected of refracted by the scene objects. Such a measure may reduce the rendering time of the visualization of the reflecting surfaces, practically not affecting their quality, especially on the high glossiness level materials.

Yet it is worth to remember that in the complex scenes with the number of the transparent objects this may lead to appearing of the artifacts, such as rendering the material's Exit color instead of transparency. In such cases, the secondary reflections may be set locally on the each exact material.

In addition to Reflection/refraction, in this rollout is another key parameter that often misleads the beginner 3d artists. This parameter named Default lights. Many newcomers, by the "good" advice from other users, that not know much about the purpose of this parameter, always try to remove the check mark from the Default lights checkbox. Thereby they try to disable the standard light sources of the 3ds Max, which by default are always in the scene for the presence of nominal illumination in an empty scene or a scene that does not contain light sources.

Default lights are needed for rendering without light sources, set manually by the user, to look not solid black, but with the geometry of objects, already lit by default auxiliary light sources. However, the property of the default secondary lights of 3ds Max is such that they automatically shut off when the any source of light are in the scene. If we create at least one light source in the scene, the default lights will be disconnected immediately. A special case is when the scene is really having no sources of direct light and it is fully illuminated by the secondary illumination. But this case is not even a rare, it is probably unique, and this may be an exterior visualization, in which the diffuse glow of the sky is simulated by the special parametric HDRI map named VRaySky. VRaySun, which is a source of direct light itself, manages it and its presence in the scene shut off the default 3ds Max lights again. The deactivation of Default lights option in the V-Ray:: Global switches rollout is needed in very rare cases only, when you really plan to illuminate the scene by a secondary illumination only, without any sources of direct light in it. This is a very unusual situation and to worry about this option is not worth our attention, we will leave it as it is set by default.

V-Ray Image Sampler (Antialiasing) Rollout

The next parameters rollout is a V-Ray:: Image sampler (Antialiasing).

For making clear what kind of a rollout is it and what is it responsible for, we certainly need to introduce the concept of aliasing. Only after understanding what aliasing is, we can understand what the antialiasing is needed for.

In the computer graphics, aliasing is a phenomenon that shows up in a form of jaggedness when converting the vector shape to raster.

On the above figure, the each white transparent line contoured square represents the single pixel. To the left we can see the ideal vector figure combined with the pixel grid for better clearness. To the right is the same figure, but rasterized, that is described as the pixels. Pixel is a square of the one color. Notice that the ideal left shape on the border with the background occupy only the parts of a pixels, while the usual raster image can hold the pixels of a single color only, without any gradients or polychromy within one.

As mentioned above, a pixel is a square of one colour. During rasterization of vector shape, the colour of the obtained square is determined by the colour of its geometrical center. Returning to our example, when the vector shape intersects the pixel at its center (1), it will be fully painted in yellow. If the geometrical center of the pixel will be background's black, it will be painted completely black (2).

As you may know, the all three dimensional objects are vector in some sense, as in fact they do not have pixel resolution and can be rendered to any size raster images, and their shape is described by mathematical formulas rather than by the individual pixel's colors array.

You may try to lay a circle on floor with the baby bricks at home. Then you certainly would able to see that this like circle can be observed as a relatively smooth figure only from the far distance, while being jagged from the closer view. The same thing happens when we try to describe curvilinear vector shape with the square pixels. Pixels are the same bricks, and even with the best will in the world, the pixel-described curves will surely have jags. It is the presence of these jags is called aliasing. Certainly, the higher resolution, the greater the number of pixels we have to describe this shape and the smaller jags are, because the size of one jag will decrease relatively to the complete shape size. However, even much higher resolution cannot fully prevent from the aliasing effect. Though, in this way we can carry the relative pixel size to the so, that it will not be noticeable for the eye, and thus eliminate the unpleasant effect.

On the left, for clarity, is the same rasterized shape as on the previous figure, and on the right is also a rasterized shape, but with four times more pixels. It is easy to notice that on the right, the jaggedness less distinct, because of more and smaller pixels. Describing the vector shape with even greater number of pixels, the jaggedness could become even less visible.

Unfortunately, it is not always rational to produce rasterization of vector shapes in ultra-high resolution. That is why, to eliminate aliasing, the anti-aliasing mechanism was developed.

In short, the essence of anti-aliasing is that the pixel-jags painted not in the exact color of shape or background behind it, but averaged by coloring into a mixed middle color of both figure and background. Thus, this visually blurs the sharp edges of the figure. At the same time, the pixels of the shape itself and the adjacent background pixels are also averaged.

On the left is the same rasterized shape, and on the right is the rasterized shape with antialiasing. It is easy to see that figure on the right, in consequence of the apparent blurring of pixels, looks much smoother and more curvilinear. Although, in fact it consists of the same pixels, but averaged slightly in color, thus creating the effect of transparency and little blur.

Rasterization process, i.e. the gathering the samples of colors from the vector shape to bitmap pixels is called sampling.

Antialiasing process uses the so-called supersampling method. Its working principle is that to determine the pixel color, not one color sample from the center is taken, but the pixel is divided into several parts, named sub-pixels, and color samples are taken from them all. Thus, pixel becomes not of just the shape or background color, but of the current mixed color of all subpixels.

The reverse process to supersampling, when to determine the color of few pixels just one point of vector shape in their averaged geometric center is taken, called undersampling.

V-Ray has several algorithms for the averaging pixel color at the boundaries of curved shapes. They are the Fixed rate sampler, Adaptive DMC sampler and Adaptive subdivision sampler.

That is to select the one of those algorithms of antialiasing, the V-Ray:: Image sampler (Antialiasing) rollout needed for. In addition, it is used for a choosing the antialiasing filter method.

Fixed, Adaptive DMC and Adaptive subdivision Image Samplers

When one of the antialiasing algorithm has been chosen, an additional rollout with its parameters appears, they are V-Ray:: Fixed image sampler, V-Ray:: Adaptive DMC image sampler and V-Ray:: Adaptive subdivision image sampler, accordingly.

Fixed is the simplest algorithm, which is roughly selects color samples relating on the sub-pixels colors, without any adaptivity. Its only option is Subdivs and it determines the number of pixels subdivisions, which will determine the final pixel color. The number of used sub-pixels is equal to the square of the Subdivs value. Fixed rate sampler is a brute force that has no adaptability in the calculation of the number of subpixels needed to form pixel colors. This algorithm calculates each case of AA with the same care, regardless of conditions. Therefore, this algorithm should be applied only in special cases, when the scene have number of DOF or blur effects, highly detailed textures and fine details, on which artifacts appear, from which you can not get rid of in any other way. Sometimes it happens that the calculation of adaptability of other algorithms may take longer than calculation with a fixed value. However, in other cases, for the lack of adaptation to the conditions, we have to pay by big time rendering when using the Fixed image sampler.

Adaptive DMC in essence is the same as Fixed, just with adaptability. The essence of his adaptability is the automatic determination of the required number of subpixels for each pixel separately, rather than use a fixed value, as in the case of the Fixed rate sampler.

The required number of sub-pixels, by default is selected based on parameters set in the V-Ray:: DMC Sampler rollout. Alternatively, based on the Clr thresh: parameter, if checkbox Use DMC sampler thresh is unchecked. The lower the value of Clr thresh: parameter, the lower is the adaptability, higher the quality of anti-aliasing and longer the render time. The parameters Min subdivs and Max subdivs determine the minimum and maximum number of each pixel subdivisions, depending on the required amount of adaptability.

Adaptive DMC sampler may be applied in the same cases as Fixed, but when the scene has less problematic areas. This will save time rendering at the same level of quality, as that is with a fixed number of subpixels. In some cases, the Adaptive DMC sampler turns out to be even faster than the more technologically sophisticated Adaptive subdivision sampler, which will be discussed a bit further. For example, that is in some scenes with Motion blur. However, is not true for all scenes and finding of its benefits can only be done experimentally. Such experiments on changing the filter and selection of best option would be justified only by the multiframe animation setup, where each additional minute of rendering of each frame can result hours or even days in result.

Adaptive subdivision is the most advanced adaptive anti-aliasing algorithm in V-Ray. Its main difference from the Adaptive DMC algorithm is a possible to use not only the supersampling, but also the undersampling on unimportant parts of the generated images, thus saving the lion's share of antialiasing computing time. Min. rate and Max. rate parameters determine the minimum and maximum values of undersampling and supersampling. These values determine the number of samples, i.e. color samples used to determine the color of produced pixels. In particular, if the rate setting is -1, it means that to calculate the color of the four (!) neighboring pixels the one sample is used. A value of 0 means that each pixel uses one sample, and the value of rate, equal to 1, means that for determining the color of one pixel, the four samples are used, etc. Mathematically speaking, the rate value is the power of the number 4 in the final number of subpixel samples of a particular pixel color of the rasterized vector shape. The Clr thresh: parameter is responsible for automatic selection of the desired Min. rate and Max. rate, as in the Adaptive DMC sampler.

Another feature of the Adaptive subdivision sampler is the presence of additional parameters that control the automatic selection of the needed Min. rate and Max. rate. These are the Object outline and Nrm thresh:. Where Object outline controls the quality of antialiasing on the borders of the object - namely, using the maximum Max. rate supersampling value, and Nrm thresh: enhances the quality of sampling within objects, depending on variations in the direction of adjacent normals, that is, small and sharp details on the object.

Randomize samples function allows you to randomly change the location of anti-aliasing samples. Nominally they are on a strict grid-like pattern. While smoothing the horizontal or nearly horizontal, and vertical or nearly vertical lines on the visualization, the strict periodic arrangement of samples may lead to parasitic straight or step-like bands. To avoid these artifacts, it is sometimes useful to do color sampling slightly chaotic; it is for this purpose Randomize samples setting is needed for.

The Randomize samples option in universal V-Ray settings should be left active.

Adaptive subdivision sampler is the most flexible, most technologically advanced algorithm of antialiasing used in V-Ray. It gives the greatest computing speed with the best quality of the image. That is it should be used as AA (AntiAliasing) default algorithm in universal settings of V-Ray. Only if there are artifacts on the fine details, and they are not removable by other ways, the less adaptive and less flexible Adaptive DMC or even a rough *Fixed *algorithm should be used.

Despite the existence of special algorithms that reduce the effect of aliasing, in practice, the most preferred and rational choice is NOT to increase the number of subpixels by banal raising high the AA settings. But to decrease an aliasing by increasing the resolution of the final image while using the slight AA. The essence of this trick is that when the resolution is decreasing, the relative aliasing jags are also getting smaller, and a slight AA removes them. High resolution and a small amount of AA give excellent results, far surpassing the ones, which can be obtained by simply overstating the parameters of AA. But the most interesting thing is that the calculation in high resolution, at least a half times greater than required, would take less time than rendering the image of the same quality, but initially with the required resolution and high AA settings.

It is so appropriate values of Adaptive subdivision sampler are the ones that are set in it by default. That is why in the universal V-Ray starting configuration they should not be changed by overstating or understating. They must be left as they are initially.

Antialiasing filters

Another important parameter is the Image Sampler antialiasing filter, it can be chosen from the Antialiasing filter section, in the V-Ray:: Image sampler (Antialiasing) rollout.

V-Ray renderer supports almost all standard filters of 3ds Max.

When the Image Sampler calculated the required amount of sub-pixels, the next process is the conversion of information to pixels. Color of the pixels can be formed immediately by averaging the color of all samples, belonging to a particular pixel, or filtered by a special algorithm that makes a certain adjustment in the formation of pixel colors from the subpixels. These algorithms are the antialiasing filters.

To select one of the filters is sufficient to mark the checkbox On and select from the drop down list the desired AA filter.

3ds Max has many filters with different algorithms. The most popular, and even causing "holy wars" among beginners and even experienced visualizers, are the Catmull-Rom and the Mitchell-Netravali, named after the developers of current algorithms.

These wars do not cease to this day, stirring up the violent disputes about which filter is better and which of them should be used for their renderings.

With the application of Catmull-Rom or Mitchell-Netravali filter, the received image becomes little sharper, as if view focuses a little on it.

For example, Catmull-Rom filter makes very sharp edges of objects, sometimes thus improving the general form of rendering, but it is not significantly affecting their internal parts.

So, if you want to raise up the sharpness, for example in Photoshop, to the entire image including the inside of objects and not just at its edges, the unwanted defects in the form of ultra-sharp boundaries of objects may appear, what will look unnatural and ugly.

Another very significant problem of AA filters is that they work in RGB space only, unable to perform properly rasterization with HDR-rendering. Performing the calculation of pixel color from the sub-pixel floating-point information, the computation errors arise and they seen as awful artifacts. They are distinctly visible on the borders of bright objects, for example on objects with VRayLightMtl materials or visible light sources, such as VRayLight. Moreover, another defects, which even more depressing - the reflections of those bright lights on the reflecting materials in the form of evident black borders around reflection spots.

Look closely at the image of an eye of this frog (yeah, it's a frog :) ). To the left is the rendering, performed without the use of AA filter, in the middle the rendering with the AA filter Catmull-Rom and on the right the rendering with the AA filter Mitchell-Netravali. The green part of the object is an ordinary VRayMtl, yellow is a bright VRayLightMtl. The left image looks slightly blurred, but without any unwanted artifacts. Easy blurring of the left image is elementary solved by sharpening in Photoshop. At the same time, the middle and right images clearly show the sharp boundary in the form of a terrible multi-colored line between green and yellow materials. This disadvantage cannot be eliminated, unless by long and tedious retouching in a 2d editor, but ... What for?

Applying the AA filters greatly reduces the freedom of post-processing of the image, making it impossible to have high quality sharpness of desired extent manually.

In addition, the filters that add sharpness on the edges of objects, very amplify the spurious moiré effect, even at high settings of an Image sampler.

It is for these reasons we should not use AA filters at all. Only in some cases, such as animation, where there is no possibility of adding sharpness to each of hundreds of frames by your hand, the use of AA filters is justified.

In the universal V-ray settings, the anti-aliasing filter should be completely disabled.

V-Ray Environment Rollout

Next in line is the V-Ray:: Environment rollout.

This rollout has important for rendering options that allow us to substitute GI glow of the environment, the reflection/refraction, and the refraction separately. These are the GI Environment (skylight) override, Reflection/refraction environment override and Refraction environment override parameters respectively.

The essence of these parameters is that, if necessary, we can put the right map to the appropriate slot, or specify the desired color, thereby replacing the map or color that is in the standard 3ds Max Environment and Effects (8 key) dialog.

Suppose we want to render a car in the exterior scene, and we want its body to mirror the clouds and surrounding buildings. All we need is to put into the Reflection/refraction environment override slot the corresponding spherical or cube map with the desired image and it will be reflected on the body, or rather on all reflecting materials, of the car. For more information on the panoramic and cubic maps, see the Rendering 360° Panorama Complete Guide.

The same can be done with global illumination of the scene.

Global light may be a simple fill lighting of one solid color,* bitmap RGB* or HDR map.

Anyway, it is the feature of the some particular scene and the goals that its creator, so all the parameters of this rollout in universal V-Ray settings should be left intact, and changes needed only in special cases.

V-Ray Color mapping Rollout

The next V-Ray:: Color mapping rollout is a very important tool, which work got to know and understand each professional 3d rendering artist.

There are the parameters responsible for the rendering exposure and gamma correction in this rollout.

Exposure is a term that came to computer graphics from photography. The exposure means the amount of light, which fell on the film during the shooting of photo. Photo can be well exposed, i.e. look good, underexposed - to be too dark and dull, and overexposed - to be overly bright. In computer graphics exposure means roughly the same thing - namely, the brightness and color saturation of generated images.

Gamma correction is a necessary correction and rendering images with a certain degree of nonlinearity of the gradient tones from dark to light. This feature is inherent to modern color information transmitting and displaying systems. This aspect can be controlled in the Color mapping of the render-engine V-Ray.

V-Ray has many exposure control algorithms. They are Linear multiply, Exponential, HSV exponential, Intensity exponential, Gamma correction, Intensity gamma, and Reinhard.

The most popular of these are the Exponential and Linear multiply.

Exponential Color mapping algorithm outputs the brightness values of all image pixels exponentially. Its essence is that the illumination of the scene appears nonlinear, with a little brighten of dark areas and darken very bright ones, so averaging the overall result, trying to display all levels of brightness in the limit of RGB colors. This algorithm allows to get rid of darkness and, at the same time, to eliminate overexposure of very bright areas, such as visible light sources or flashes on the scene geometry from them.

Another feature of an Exponential algorithm is that the color saturation of pixels is calculated based on their brightness.

In practice, this algorithm results very faded and unsaturated images, satisfying the necessary level of photorealism and presentable only in rare cases.

As a rule, Exponential Color mapping can be suitable only for some interior renderings, where undertones are more important, rather than specific colors of individual objects.

It is because of dullness and desaturation of the image, Exponential Color mapping completely unsuitable for presentable product rendering, where the representing for each color of the object, bright and saturated as much as possible, is important. For such rendering, the Linear multiply Color mapping algorithm is much better.

Linear multiply Color mapping algorithm is a linear algorithm that multiplies the each pixel color on its brightness. Thus, we obtain a linear, bright and vivid picture.

However, because of its linearity, the use of this algorithm leads to strong and sometimes horrible overexposure in bright areas of the generated images, making it very unsuitable for middle-tone interior visualizations.

Theoretically, to correct the defect of Exponential Color mapping algorithm, namely the colorlessness of an image, and to approach saturation of the Linear multiply Color mapping algorithm, the HSV exponential algorithm was designed. However, in practice, it remains just as uncomfortable as the Exponential or Linear multiply Color mapping ones.

There is another, not very popular, especially among beginner 3d artists, yet the most powerful and flexible algorithm for exposure control in the V-Ray, called Reinhard Color mapping.

Reinhard Color mapping is a hybrid algorithm, ideally combining in itself the other two basic algorithms: the Exponential and Linear multiply. It is this algorithm is able to give not overexposed, without the "nuclear" overbrights, and at the same time rich and vibrant presentable images, making renderings truly photo-realistic and tasty :)

By the way, you can download frog 3d model from these examples for free.

You can see the three images, created with the use of different color mapping algorithms.

On the left is the Exponential, in the center is the Linear multiply, and to the right is Reinhard color mapping. Green rectangle on top of each example shows a sample of desired colors, which are in the Diffuse slot of a frog’s green VRayMtl material.

It is easy to see that the left image is dim and pale, and even specially backlit scene background looks gray instead of white, despite the fact, that in the Diffuse slot of its material is pure white RGB 255.255.255 color.

The middle image has a more lively appearance, the frog is bright and juicy, and the white background looks clean and white. However, compare the bright and oversaturated frog color with a rectangle sample. The disadvantage of this color mapping method is also obvious. Despite its richness, the picture turned out very overexposed.

Quite another thing is the right visualization. It looks nor dull, nor overexposed. The white background is really white. Look at the color sample; it exactly coincides with the color of the frog. This is precisely the result, which must be obtained. It is presentable and lively. Absolutely the same situation will occur at interior or exterior 3d rendering. This is thanks to the use of a good and adaptive Reinhard algorithm.

To work with this algorithm, it is necessary to find Type section in V-Ray:: Color mapping rollout, and select Reinhard from the drop-down list. This will make the control parameters of algorithm to appear.

Multiplier is a parameter that controls the overall brightness of the image. More about applying it, you can learn from the Replacement VRayPhysicalCamera by a Standard tutorial.

Burn value is the most interesting parameter of this algorithm. As mentioned earlier, Reinhard is a hybrid of two other algorithms. Exactly the Burn value determines how Reinhard will behave. If this parameter is zero, then Reinhard will give a result quite similar to Exponential color mapping, if unity – similar to Linear multiply color mapping. Intermediate values move the direction of the work of Reinhard closer to Exponential or to Linear.

Thus, by adjusting Burn value, you can achieve excellent bright and intense visualizations, which Exponential cannot boast, at the same time to get rid of Linear’s overbrights.

In practice, the most suitable is the value of Burn value equal to 0.35. This value will be an excellent choice for most interior and exterior scenes, as well as for product renderings. That value should be used in universal settings of V-Ray.

Gamma is the parameter that allows us to perform gamma correction of the image.

Nominally, the correct value of gamma correction is 2.2, but there is one trick. Despite the theoretical correctness of the concept of rendering with gamma 2.2, even more interesting and beautiful results can be achieved by slightly shifting the gradient of gray on rendering to the black. This will make the shadows more juicy and dense. It is enough to lower the value of the gamma correction from 2.2 to 1.8 - 1.6. Then the rendering becomes even more contrast and realistic.

The nominally universal value, suitable for any scene is the value of color mapping gamma equal to 1.8. That is it should be used in universal V-Ray settings.

Parameter Sub-pixel mapping determines where exactly color mapping will be applied, directly to the final pixels of the image or at the level of sub-pixels that are only after color mapping will be converted into pixels. This feature is recommended to be disabled even by developers of V-Ray, as in this case a more correct result can be achieved. However, this parameter is very useful when we want to render not to HDRI, but to RGB space. Its activation lets to get the satisfactory smoothed edges on the bright image areas, which come beyond the RGB space. Still, this parameter may be disabled for correct HDR rendering.

Clamp output allows clipping of the high values of brightness outside the RGB space. Sometimes it can be useful, for example, to get rid of artifacts of AA filter (see the example of the frog’s eye above) that may appear because of an incorrect work of the AA filter with HDRI. Also is to obviate the problems of AA smoothing when rendering images with dynamic RGB range.

Nominally correct is a rendering and saving the image to the HDR (High Dynamic Range) format, so in universal settings Clamp output option must be turned off, or else when image stored in HDRI, the complete information about the actual brightness of the pixels irreversible loses.

Clamp level parameter determines the level of the cut-off of the information by Clamp output function.

Affect background determines whether to apply color mapping to the Environment map or color. As a rule, if the override background image is set, it is rarely a need to apply to color mapping on it. Therefore, in the universal V-Ray settings, the Affect background function must be switched off.

There is a misconception that there is no any need to configure color mapping at all; that it is sufficient to render HDRI in a linear space and save it in the appropriate format of HDR; after that, using compositing in 2d editor, brush up the image to the desired gamma with the desired exposure. Unfortunately, it is not so.

Firstly, to render one exposure, and then tune it up to another is a blind work. To make a beautiful photo-realistic visualization, not seeing it, its details, colors, illumination, but focusing only on a rough preview with the unfair color mapping it is necessary to be at least a seer or prodigy, who can produce the desired exposure correction in own mind :)

Secondly, the adaptive V-Ray engine, based on the principle of the sampling the most important to the overall result values and cutting off the less important, will simply incorrectly calculate what image area is important and what is not. V-Ray, to save time and computing resources, produces less accurate calculations in dark places with low light energy, in consequence of what they will have a lot of noise. However, since they are still visually dark on the supposed image, then the noise is little or even not noticeable on that image, even when we closely look at it. At that time, the bright spots with high light energy are always in sight and V-Ray produces the most accurate calculations on them, so in bright places there are less of noise and artifacts.

Therefore, if the rendering calculated in gamma correction lower that it would later be observed in, e.g., with a linear gamma 1.0, the picture, after further correction to the bright 2.2 side, will be very noisy and contain many artifacts, as V-Ray considered that image initially as dark.

Nevertheless, the Don't affect colors (adaptation only) function exists for such unusual requirements. The essence of this function is that it allows one to get an image in linear gamma of 1.0, but all the effects and GI are calculated as if they were in gamma 2.2. Thus, we obtain a linear image ready for further correction, but without the presence of the aforementioned artifacts inherent to the dark areas of rendering.

However, this method is more exotic than practically necessary, so in the universal V-Ray configuration the Don't affect colors (adaptation only) feature should be turned off.

V-Ray Camera Rollout

The last rollout of the V-Ray tab of a Render Scene dialog, is a rollout named V-Ray:: Camera.

This rollout holds the camera settings, namely, the additional features that enhance the standard 3ds Max camera. Such as depth of field or motion blur, switching types of cameras to standard, spherical, cylindrical, etc.

Of course, the universal configuration of V-Ray has no relation to this rollout and there is nothing should be changed in it without very specific need.

Conclusion

Now you are a little more familiar with some important aspects of how to configuring V-Ray renderer.

We hope that reading this tutorial makes you see that the previously seemed unfriendly or even challenging interface of V-Ray tab not so hard in reality. Understanding these basic features will make your work at V-Ray much easier and more professional.

See you in the next chapters of this tutorials series!

Advertisements
Comments
i_one_as
Nov 13, 2010
I like how you chewed everything for us! There is nothing left to ask! Thanks guys!
Ellesar
Nov 13, 2010
Very good tutorial, already read twice!! I learned a lot of new things, especially about anti-aliasing filters. I could not find a clear explanation on this topic anywhere. Thanks! Keep going!
hands-creative.com
Nov 13, 2010
The tutorial is really clear and accessible. Thank you.
Elena
Nov 14, 2010
Thank you very much. Very interesting tutorial. I have a question about gamma 1.8 in Color mapping, do I need to change something in the Gamma and LUT of 3ds Max? do I need to change the value from 2.2 to 1.8? Thanks. I look forward to read a next chapter.
RenderStuff
Nov 15, 2010
Elena, hello, it is a good question. No, leave Gamma and LUT settings as it described in our gamma 2.2 tutorial, that is 2.2.
tutindex
Nov 15, 2010
Wow! I 've learned more here than when I studied Image Processing in the University!
RyJek
Nov 15, 2010
Thanks a lot, useful and helpful.
Elena
Nov 16, 2010
It is clear, thank you very much.
zero
Nov 17, 2010
great article, thanks
Evreny
Nov 18, 2010
Thank you!
Bob
Nov 18, 2010
That is really fundamental, thank you!
cadu3d
Nov 19, 2010
Thanks man!
Arq. Raul Cervantes
Nov 20, 2010
Hello, my 3d rendering world was completely turned on 180° by this tutorial, i learned here a lot and i am very grateful to you... The only thing is that in all my interior renders, where i've tried your tips, was one big BIG problem... If I unchecked an 'Sup-pixel mapping' and 'Affect background', my VRaySun produces a jagged edges where the light goes and some damn bothering artifacts (in burned areas)... If i decrease a light intencity a little bit down, my scene gets very very dark, but with no jagged edges on lights... Yeap my renders got better with reinhard and the tip of 1.8 on gamma... i was the 2.2 boring guy.... What can i do, i am really lost... i need help!!!! Thank you very much, because in everything else you made my knowledge grew in just a few minutes.
Adanedhel
Nov 20, 2010
Wow, I was looking for something like that. I always knew which presets to choose, but I didn't know why. Now i know. thanks a lot mate.
RenderStuff
Nov 20, 2010
Thanks all for your kind responses!
RenderStuff
Nov 20, 2010
Hi, Arq. Raul Cervantes,

Thank you for your question, we have made the changes to the tutorial, so there is a little more information regarding Sub-pixel mapping already.
The rendering engine performs the calculation of the pixels brightness in the High Dynamic Range (floating-point values, not integers) and thus the brightness values may significantly exceed the RGB colors range (in which you observe the image on the screen). From the monitor, the human vision percept the colors that are in the so-called Low Dynamic Range (LDR), which the RGB color space is in. Let suppose, that some obtained pixel value is 1000 brightness units, but the RGB can carry only 256 gradations, where the last 255 value (starting from the 0) is the maximum brightness level that we can see as white color on the monitor. The result is so that the 1000 value is represented as the same white color. But… there are the 745 units of actual difference between the 255 and 1000 values.
Though, if we render and save the 3d randered image in a HDRI format, that holds much more gradations of the brightness data, we would be able to see that even the brightest white-appearing in RGB areas have the smooth gradient and no ugly sharp aliasing jags on the edges.

As the tutorial says, the enabling of the Clamp output function cuts off the pixel brightness values, which exceed the 256 gradations limit. That way we can get the right average value of the pixel, if it should represent. If one pixel part’s brightness is 1000 that is white in final and other’s is 0 that is black, then their average is 500, which is still white in result. But, after clamping, the 1000 becomes 255, thus the average of 255 and 0 is 128 that is the correct middle gray color.

So you get a situation where a pixel that needs to be 'gray' on a high-range gradient is clamped to 'white' on an image with a low gradient range. Then instead of a smooth transition between colors, such as between adjacent dark gray and light gray pixels, the visualization will have a sharp transition from dark gray to literally white. And this is exactly the jaggedness you are talking about - a sharp and contrasting color change.

Again, when the renderer does supersampling, it computes a scene with a higher resolution than the final image resolution. This means that, for example, four pixels (subpixels) in the oversampling phase will become one pixel in the final render. To calculate the final pixel color, the render calculator takes the average of the four subpixel colors [(subpixel_1 + subpixel_2 + subpixel_3 + subpixel_4) / 4 = pixel], and obtains the final pixel color. If one of these sub-pixels has a very bright value, then the average value will be much higher than expected, as a result, the final pixel on the final render will be too bright.

VRaySun and VRaySky are HDR color sources. It generates a very bright colors. Not surprisingly, you see artifacts on visualizations by looking directly at them or at their reflections.

What to do? Sometimes there is nothing you can do to avoid overbrights on a render. The ultimate remedy is to render an HDR image, without clamping. If you still want to render to LDR, disabling 'Sup-pixel mapping' may make things even worse, because with 'Sup-pixel mapping', colors is being clamped on suppixel level and the final pixel color will be less bright, but with 'Sup-pixel mapping' off, you will get the just described situation, then one 'unclampel' suppixel with a very bright value can make the final pixel color an excessively bright even if other 3 subpixels is dark.
Arq. Raul Cervantes
Nov 20, 2010
Hello, well I'm really amazed that my "great" knowledge about 3D and rendering, has been thrown back to earth... And i am REALLY grateful to you, that stimulates me to study more and more and more... I truly thank you for your work and your time to answer my humble question. I really need to learn a lot, I studied all these processes on my own. It was a painful way, but at the end of each render I feel great... I hope that these days, after reading and re-reading all this, my works will be better, by the way, I have not used HDR in my work process, I will definitely try... THANKS A LOT MY FRIEND!!! Cheers from Mexico and keep up the good work for us, the mortal humans....
Arq. Raul Cervantes
Nov 20, 2010
I'm really sorry, but i have one new doubt, i am the type of guy who saves it's image as JPEG... that's wrong??? i don't get how to save it as an HDR??? may be i'm a rookie of 3D. Damn all the time i learn new stuff...
RenderStuff
Nov 21, 2010
Yes, your doubts are well-founded. Normal JPG, GIF, BMP images are raster 8-bit image formats. Therefore, color data, along with brightness gradations of one pixels, are stored with 8-bit accuracy. Of course, an 8-bit image is an LDR. There are 16-bit formats, such as PNG, for example. They can store pixel color data with 16 bits. However, 32 bits are required to store true HDR color data. Only 32-bit formats allow you to store HDR color data generated by V-Ray (unclamped color mapping). For example, Radiance Image file (HDRI; *.hdr, *.pic) or OpenEXR Image File (*.exr). However, they are not completely usable immediately after rendering, and they need post-processing in an HDRI-compatible raster editor in order to edit and convert them to the usual 16-bit or 8-bit formats for later use.
Arq. Raul Cervantes
Nov 22, 2010
Thank you a loooot!!! I'm a big new fan of Renderstuff!!! thanks for your kind tips, however i'll start looking of how to save in HDR, i will start saving it like that and editing in photoshop. Thank you very much!!! Really, was a lot of help
RenderStuff
Nov 23, 2010
You're welcome!
V3N0M
Nov 23, 2010
Hi, thanks for this tutorial! I have a problem when I'm trying to render an image without AA, the edges come out very sharp and they look ugly to me. Is it posibile to make them look better without using AA and without increasing the rendertime? I used the exact settings from this tutorial, the colors come out nice but the edges look bad :( can you help me?
RenderStuff
Nov 23, 2010
Hi, V3N0M V-Ray nominally cannot render image without AA. AA is always applied, but you could choose Fixed rate sampler and set Subdivs value to 1. That way V-Ray “antialiases” the each pixel with single sample, which is, of course is equal to not using antialiasing at all. But, this is surely not the result you want to achieve. AA is necessary if we want to have smoothed edges on rendering, because no other post-process can enhance them without knowing the origin 3d objects data with the same accuracy. The AA settings screenshot (that is in the AA algorithms section) shows the optimal settings, which you may use to get good result. Please notice that tutorial advices to turn off the AA filter, not the actual AA.
trizet
Nov 25, 2010
Judging by the fact that it is written that it is part of the lesson is to be continued?
RenderStuff
Nov 25, 2010
Yes that's right. There are still two, stay tuned;)
Сергей
Nov 26, 2010
Reinhard algorithm is good, but to highlight areas where the contrast gives strong teeth and how to get rid of them is not clear, and Exponential does not chipping, but it faded (
RenderStuff
Nov 26, 2010
However, Sergey, when rendering banal concepts and conservation in the LDR, almost linear Reinhard on bright vyhodyaschiyh for RBG range of sites can give the wrong anti-aliasing. Actually, that can be corrected by activation Clamp Output and Sub-pixel Mapping.
WehrLich
Nov 28, 2010
Thank you for the excellent article, I was looking like :) But the question I have made in the settings range of 2.2 and rendering, and render the texture color a bit lighter (as opposed to specimen) Duck ... so I do not understand ... So I should be or not properly configure something?
RenderStuff
Nov 28, 2010
Hello, WehrLich, really hard to say for sure in the blind, it is necessary to see your scene. It is possible that your texture is just presvechena lighting in the scene, that's all. But, in any case, if you want to make a contrasting texture, a little podtemniv them to render, go to the "Gamma and LUT" and darken incoming texture, see Bitmap Files => Input Gamma. If you have everything set up right, now you must stand Input Gamma 2.2 (see Using gamma 2.2 ), and you put 2.3-2.5, the taste in general. Then, all the textures are a bit dim at the entrance to the rendering engine.
WehrLich
Nov 28, 2010
I Tried ... still remains at 2.2 scale :) By the way, when will come the following lessons render settings? Just can not wait to have read about other settings ...
RenderStuff
Dec 12, 2010
Subscribe to RSS (yellow icon top right) and you'll be aware of all the updates on RenderStuff;)
Hudruk
Dec 12, 2010
Thank you for the lesson, break it down, it is available))) When continuation ???
indian
Dec 21, 2010
Hello. When you plan to publish Settings and Indirect illumination Settings? Before New Year or after?
RenderStuff
Dec 21, 2010
Hello, indian. Section Indirect illumination is very voluminous and requires careful design that will take some time. Therefore, the continuation of this lesson on the V-Ray - wait after the New Year;)
Hudruk
Dec 22, 2010
uuuuuuuuuuuuuuu :(
RenderStuff
Dec 22, 2010
Production costs, sorry.
sam
Dec 25, 2010
Thank you for the lesson, it is gorgeous!) All clearly and chewed everything in one place) We are waiting for the continuation!)
foot
Dec 26, 2010
About AA settings: I tried to do everything as written here, made a picture resolution of 1280 by 960, in principle permission is not small, but in many corners there were ladders (((Before doing this rendering of the same resolution, but Adaptive AA was DMC.S it was a miscalculation of course much longer, but for something all the faces were rovnye.Poetomu question, if I want to do imaging with Adaptive subdivision AA, that would smooth the corners, I need another resolution uvilechivat have pictures or tweak settings is higher? thanks in advance )!
RenderStuff
Dec 27, 2010
Hello, foot. resolution should be increased. 1280 by 960 pixels - very low resolution for final rendering, especially saturated large number of CMV small parts, such as interior. Make sure to increase it to provide high-quality and well-smoothed image. When it comes to photo-realistic visualization, the sufficient resolution is at least 2500 pixels on the longest side, and more. Alternatively, of course, you can not change the resolution to increase the AA settings. In particular, if there is insufficient anti-aliasing on the edges of objects (the corners), it is possible to activate the Object outline function for AA selective amplification in these areas, or simply to raise the value Min. and Max. rate The, however, a variant with a resolution to be more rational and give you more freedom in post-processing of the final rendering, for example, in Photoshop.
Tanya823
Jan 12, 2011
An enormous thank you for the lesson !!!
cypress
Jan 19, 2011
All clear, thank you :) Question: When will the continuation?
RenderStuff
Jan 20, 2011
Continuation of the already written, is in the wording, it will take some time. Get ready to read 12,000 words;)
Lexus0103
Feb 8, 2011
When about to be continued? I liked very much !!!
RenderStuff
Feb 11, 2011
The next lesson on the optimal settings V-Ray read here .
Hudruk
Feb 22, 2011
Hello! I do not understand what HDRI and RGB space? It can be a little bit more.
RenderStuff
Feb 22, 2011
Hello! Although it is difficult technical terms in this context must be understood is that the standard RGB format can bear a very limited number of brightness values ​​on 8 bits for each primary color. Consequently, a very large portion of bright pixels obtained in photorealistic rendering, it simply can not accommodate itself. In turn, HDR formats spared from this drawback and can carry a very large range of brightness values.
SLONIK_ZX
Feb 28, 2011
"It is for these reasons, it is not necessary to apply the AA filters at all. Only in some cases, such as animation, when there is no possibility of inducing sharpness for each of hundreds of frames by hand, AA filters use is justified" Maybe I did not understand, but, for example, in contrast AfterEffect sequence is very simple twist ..
RenderStuff
Mar 2, 2011
SLONIK_ZX, you knew all right. Simply, as the long-term practice, not all, even experienced 3d artists are able to use professional kompouzingovymi programs. Especially if we are talking about is not an animated film, but about simple fly-through presentation of architecture;)
kossbloody
Mar 8, 2011
Hello! I wanted to ask a question about the size of the final image, but have found the answer to the comments. What to do on the long side at least 2500 pixels. And I have a different question. Usually I show people pictures of the monitor and its size is less than the final image. I have always rendered the 1200h900. (He came to it with time and higher resolution, I do not need). I understand that the size of 2500 pixels., Needs to accurately render a detail and then I can reduce it, if necessary. And increasing the size of images that are more efficient in time than the increase of IM parameters and LC, to obtain the same result by 1200h900 image pixels. Did I understand correctly? If in the end I anyway need the picture in 1200 by 900 pixels, with the same result, that will be faster over time: 1. to increase the parameters of what IM and LC. (If so, what that I received on the quality of the image is the same as 2500 pixels. For your advice). 2. render 2500 and decreased to 1,200 (it sounds crazy, but still :) PS I apologize for the complexity of the formulation. If you do not understand, I try to reformulate the question. Thank you.
RenderStuff
Mar 8, 2011
Hello, kossbloody! Increasing the resolution to render even the lowest setting at least give more details on the image, which greatly adds visual quality saturation composition. Large resolution, in addition to its visual benefits, gives a lot of freedom for further processing, since we are talking about the subsequent operations, such as changing the size of the picture. As part of the most visualization 3d, increasing the resolution render, without changing the quality settings, automatically makes the best IM, AA, reduces the noise, but as with any improvement, of course, increases the rendering time. Specific figures for "hit" in your resolution with the same quality simply can not in principle be, if you have the desire, you can experiment and learn it better for your particular situation is to increase the quality settings. If you berezhesh time, the nominally Use these settings to the maximum resolution available :)
kossbloody
Mar 8, 2011
Thanks for the answer! :) As I understand it, all the same, do you recommend to render at least 2,500 pixels, as a lower resolution will be difficult to achieve photorealistic quality. So? )
RenderStuff
Mar 8, 2011
We recommend to render a resolution of 6000 pixels, but in 2500 - is a tentative compromise between quality and render time. You can safely rely on it. If time and resources allow, the principle more = better to a certain extent universal
kossbloody
Mar 8, 2011
Thank you.
Slon
Mar 9, 2011
Very useful article!
Darya
Mar 16, 2011
Thank you all very clearly explained and without unnecessary remarks. Thank you!))
Nantaruck
Apr 2, 2011
Hi , Just wanna say thank... this article is the best V-Ray tutorials ever. I just finally rendered a scene that make me feel great about using V-Ray. You the MAN!
RenderStuff
Apr 2, 2011
We are glad to know this. Thank you for your feedback!
Anastasiya
Apr 13, 2011
Thank you for the lesson! For a long time looking for a good lesson, where the settings described in detail, because basically they write in the lessons: "place and not otherwise" and do not explain why this value
Podr
May 15, 2011
Hey. A very useful series of articles. I read in one breath. For the first time about the many times read the human language. Came across another article:
Demystifying V-Ray DMC Sampler
According to it, 'Adaptive DMC' works better than 'Adaptive subdivision'. But your advice is the opposite. Where is the truth, man?!
RenderStuff
May 15, 2011
Hello, Podr, thank you!
This article is great and contain a lot of theory, but we are practitioners, and our advice is focused on practice only. You need to make an desidion not by 'what is best in an ideal world', but on the basis of daily results. As we wrote, somtimes 'Adaptive DMC' can give a better results. However, as practice shows, these cases are exceptions rather than regularities, so we still advise you to nominally use the 'Adaptive subdivision' sampler, aand only if you have some problems with it, use different image sampler.
pixelerator7
May 20, 2011
This was hell of an article! Very well explained. I wanted to get a better concept of vray color mapping methods and it's what I got. We're making an animation short series rendering with Vray, and I'm sure it'll help me alot. THANKS!
denny
Jun 26, 2011
greeting. in VRay 1,5 SP4A in the Color Mapping is a function Linear Workflow. You could explain that it gives. From HELP little understood, in practice, incl. colors are more luscious and the rendering time is reduced by 35-40%.
RenderStuff
Jun 27, 2011
Hello! If, in general, this feature is designed to convert color materials settings created by using a different workspace. That is, if, for example, the scene was created by someone in the Gamma 1.0, and then you discover it on your computer with a workspace in the Gamma 2.2, you do not have to reconfigure the materials for the new gamma value. Activating the checkbox Linear workflow in Color Mapping automatically reconfigures all. To be more precise, it will apply to old materials inverse value of gamma correction, that is, in fact, cancel it, after the application of the current value range in which you're working on now, is equivalent to the original setting materials in the workspace with the current value of gamma;)
denny
Jun 28, 2011
in fact a lot of things on the VRay written and all in different ways, which proves that there is no universal settings and magic buttons. For each scene something of their own, and each has its own approach to the settings. From the last sensible explanation I've read, written and interesting details.
RenderStuff
Jun 28, 2011
With regard to the universality of magic buttons and it is true. Read chapter about V-Ray settings universality myths. Just pay attention to the fact that this chapter about Antialiasing and Color mapping is only the first part of a series of articles about the V-Ray renderer. In order to understand the whole picture, carefully read all the articles in this series. Do not stop at the first.
Andrey
Jul 29, 2011
Hello ! please explain this phrase
However, it can be very useful when rendering not HDRI, and in the RGB color space
namely, that means HDRI space and how it differs from the RGB and where space is selected in the settings? thanks =)
RenderStuff
Jul 29, 2011
Hello, Andrey! Your question has already been asked here before, look in comments above. There discussed how to handle an HDRI rendering.
Andrey
Jul 31, 2011
Thanks for the instruction, already read!
Andrey
Jul 31, 2011
Prompt please: the parameters min rate, and max rate only when Adaptive subdivision, Adaptive DMC how they are configured?
RenderStuff
Aug 1, 2011
Hello! Parameters Min. rate and Max. rate parameters are not global, it is only the specific parameters of the algorithm, in this case, anti-aliasing algorithm - Adaptive subdivision. They simply works as multiplier or coefficient. In our case, this means how much larger or smaller resolution of calculations should be compare to the final rendering. Adaptive DMC, for example, uses the concept of resolution. Adaptive DMC operates with the concept Subdivs.
Andrey
Aug 8, 2011
Thanks for the answer ! =)
Alek85
Aug 26, 2011
((((Been reading your nastroek..chto something has helped improve the quality and degrade some items, of course it's something I do not where tweaking, but one of them, based on your settings included in the image sampler adabdiv sabvizhn instead adaptive LCA and, for example, the far wall (plaster), all kind of blurred and she rudely oshtukatena gone, the problem decided included items INTACHMENT, but you have written that the strikers include only when it is very necessary .... shorter straight and do not know what to do, ie adaptive sabvizhn smoothes everything, somewhere well as where on the contrary do not need to (((... Shoot billion settings ...: D when it will make the "make nice"?
Anton (RenderStuff)
Aug 26, 2011
If you think that some of your specific situation in a particular scene better than any other setting, feel free to put it. The tutorial describes only universal start setting, which means that they should start and, if necessary, change. If there is a need, - changes, there is nothing to reflect. But do not confuse the versatility and uncompromising. The Board regards the fact that the vast majority of scenes is recommended to put the Adaptive subdivision sampler, does not mean that it is absolutely better to Adaptive DMC sampler. By contrast, Adaptive DMC sampler more accurate and high-quality algorithm, but, unfortunately, is much slower. When you need to render two shelves on a flat wall, you can set and Adaptive DMC sampler, but when the scene is really hard, and it can even rendering with Adaptive subdivision sampler to go 4-6 hours to view, idealism and uncompromising quickly go by the wayside. In a complex scene, replacing subdivision on the DMC can easily cost two-fold increase rendering time that the commercial rendering, with clearly defined deadlines, it is not acceptable. If we talk about uncompromising quality, the best choice would be not to Adaptive subdivision sampler, and not even the Adaptive DMC sampler. The best quality will give Fixed rate sampler with up stranded Subdivs maximum. However, the rendering time is just terrifying, even in the most elementary stage. Need Whether this difference in quality at this price - it's up to you;)
Alek85
Aug 26, 2011
Understood, thanks for the answer
supun
Nov 14, 2011
so nice, thanks so much for this
George M.
Nov 16, 2011
Hello, Again a problem with "don't affect colors(adaptation only)" I always save the result as OpenExr/hdr 32bit full float but not ticking this option the VFB "burns" the color mapping's gamma into the exr resulting in an overexposed image when opened in Nuke/Photoshop. My renders stays linear only if I tick this option. In VFB the image looks ok (sRGB off). and also when saved as 8 bit images looks ok. Another problem is again the subpixel+clamp output option. Not ticked again creates jagged edges and bright artefacts even if saved as exr full float. LUT is set as in tutorial. Help :|
RenderStuff
Nov 17, 2011
Hello, What does the Don't affect colors (adaptation only) function, you can read from this tutorial. All above that, is that you do not need to enable this function without any special need. How to know if there is a special need? Very simple. If you do not fully conscious about how this function work, then you surely keep it off :D Just do not pay attention to it, this is more exotic and theoretical capability, than really useful thing. As for the Sub-pixel mapping + Clamp output, you should turn off these option if you render to HDR format only. In its turn, you want to render to HDR format only if you plan to post-process your render in 2d editor lately. In other case, there is no need to output rendering to inconvenient HDRI. After all, you will have to convert it to usual LDR and that’s it. Talking about post-processing, in this context, the only point here is the correction of gamma and exposure. The other post-effects as well as color balance, can be added and/or altered on the 8bit image with the same success. What do we mean, when saying exposure? Suppose, you render the interior scene and you have a vase with flowers on the windowsill. In this situation, the brightness of window’s pixels is much higher that the brightness of the vase’s ones, and they come out beyond the standard RGB brightness range. However, the monitor, which displays your rendering, work in the RGB space and the over-brighted pixels are shown as a usual white color. Even if those pixels not really the whitest of all, they already will be shown as white. This is the effect we can observe on the vase contour, when it is very sharply jagged. But, if you open this HDR image in Photoshop, for example, and, using the Exposure tool, move the same named slider left, then you’ll notice that the vase contour is pretty smooth. You just don’t see it on the limited RGB range. This time, you will have the smooth contour, but the quite dark vase body. To return the natural vase brightness, we need to selectively (using masks) change the exposure, only where it is really needed. Speaking about practice, this technique is used to get rid of the overbrights on the ceiling near the window, or on the transparent tulle in front of it. It even can be useful to get some light in the too dark areas, far from the scene lights. If you activate the Clamp output option, renderer will automatically darken the overbright zones to make them fit in the RGB range. Of course, it will be done as aptly, as the automatics allow it. Only if you are not satisfied with the automatic result, or you wish to change the brightness in the specific render’s places, compensating the excess of scantly illumination, then you must render with the Clamp output checkbox off. If not, the HDR-rendering will not only be inconvenient because of manual conversion of the TIFF and PNG formats, but also it is much resource-consuming, because the float format rendering is more hard to perform, than the RGB. As a resume, you have to decide if you really need the HDR rendering. If yes, then turn the Clamp output off, render the HDRI and control the exposure of you renders by manually altering the desired areas and get rid of the all flaws, including the jagged edges by oneself. If not, the turn the Clamp output on and render directly to LDR format. This is less professional and flexible approach, but it is also reasonable ;)
George M.
Nov 17, 2011
I'm glad I found this site :D Of course it is not difficult in adjusting in post-processing the exposure to match the VFB result. The main reason that I asked is that every book and tutorials I read so far explains how images in linear space looks dark when viewed due to nonlinear displays and only full floating point images will display correctly on post editing softwares. I was expecting that a correct linear workflow will result in a HDR image that will look exactly like VFB result without further post work and 8 bit images should look dark without applying the gamma correction on output. Didn't expected to have a bright HDR result similar to an image that had 2x times gamma correction applied (that is my case) and VFB-like 8 bit images. I didn't found complete workflow informations for gamma 2.2 and HDR post editing so far that's why I have some gaps on this. Perhaps you will make a complete tutorial on this subject, from rendering to post editing hdr images. :D Thank you.
RenderStuff
Nov 17, 2011
It is great that you enjoy this site. We'll be happy to answer your questions in comments. Yes, the tutorial on HDRI rendering is the logical continuation of the best vray settings series and it is the knowledge, which we will share after finishing those ;)
Nocnik
Dec 29, 2011
Thank you for this tutorial, your work, knowledge and experience you share! Happy new year and take care!
RenderStuff
Dec 30, 2011
Thank you too, Nocnik! Happy New Year
Manoj
Jan 3, 2012
Wow man this is really wonderful. No one has ever before be able to explain this with that depth... You solve 100% problems i had... Looking forward for more good stuff from you!!!
3dlib
Jan 4, 2012
Great article, thank you very much .
SERDGIUS
Jan 6, 2012
Thank you very much for the most reasonable explanation. A low bow to you for your work.
lamb
Feb 24, 2012
This is absolutely amazing!
nayk
Mar 29, 2012
Hello! excellent lessons, respect you! * THUMBS UP * I have a question about filters ,,, where frog eyes I think the picture for Mitchell and ketmul room you swap ,,, maybe I oshybayus but tried sharpness The appearance by the large room with ketmul and with Mitchell smaller :)
Anton (RenderStuff)
Mar 29, 2012
Hello! There is nothing I do not need. It seems to you: D All screenshots to all our lessons illustrate exactly what is written on them. Of course, depending on the materials, the model and configuration, etc., result obtained you can vary slightly;)
nayk
Mar 29, 2012
Oh responded quickly thanks) well, then will continue eksperemintirovat and learning;)
Woodrocker
Jun 11, 2012
Very helpful! I like the way how you describe the processes and the problems in your own words, makes it easy to understand and confidential! - Thanks a lot!
JoeX
Jul 15, 2012
Awesome guide, I learned tons of stuff in such a short time, make more stuff like this please! :) Thank you!
ami
Mar 22, 2013
sir its a fabulous information not information it is a really wanted "STUFF"... ;) i am an interior visualizer and my suggestio is to u will include a good a rare studio exterior rendering set up using vay hope u willl upload it soon... thank you keep it up
shahin
Sep 2, 2014
very nice artcile!, i have an doubt, when i rendered it looks more yellowish with using only ny vray camera...without any lights.. give me some solution pls
modar
Oct 6, 2014
Hi, I have a problem with v-ray v2.0 all (Linear multiply , Exponential , HSV Exponential ... )in color mapping , have the same effects . so what is the solution ? and where is the problem ?
RenderStuff
Oct 7, 2014
Dear Usco Thank you very much for the tip. I watched this video, so if I have small patterns like those threads, and my arch I can get both on the road, I do it with the help of PathDeform modifier. But I have a problem that my pattern, in this case one and large, I need whatever it has acquired this form itself, she bent under the arch I needed. I do not know if I could explain my situation. : | Thank you for your time :)
modar
Oct 7, 2014
hi , thank you for replying on my question yes I didn't enable «Don't affect colors (adaptation only)» I will put you in the image : sub- pixel is off clamp out is off affect background is on Don't affect colors (adaptation only)» is off and then if i tried to change between (Linear multiply , Exponential , HSV Exponential.. ) or even in the sub parameters for each of them (dark multiplier , bright multiplier , Reinhard burn value ) nothing works , the results are the same. so maybe the problem in my v-ray version ? or there is an option i have to enable or disable ?
RenderStuff
Oct 8, 2014
This probably indeed is a bug of your V-Ray version if you do not use the automating plug-ins like SolidRocks that take over all setups on themselves. Anyway, we haven't met this kind of a problem on our practice.
DomoVoi
May 11, 2017
Burn value is the most interesting parameter of this algorithm. As mentioned earlier, Reinhard is a hybrid of two other algorithms. It's Burn value that determines how Reinhard will behave. If this parameter is zero, then Reinhard will produce the result of The lesson ends .. A PIT!
Yury E
May 14, 2017
Good afternoon! The article is super! But it's been about 7 years and in the actual vray (v3 +) renderer a lot has changed. For example, the first time I could not figure out how to turn on the Adaptive subdivision image sampler mode and set the number of minsubspies equal to less than zero ... :) Now the concept, as I understood it, "changed slightly" - less than 1 set It does not work out, but the Min Shading Rate parameter appeared (important, as I understand it) .. In short, a huge request, if possible, to update a single post with universal settings for a new engine - what is there for that now responds (without repetitions, only changes). Thank you in advance! ;)
Alena
Jun 10, 2017
Hello. Tell me please, the optimal settings - what values ​​in which case do you use Progressive or Bucket?
Dmitry
Oct 21, 2017
It would be very cool if the article was updated in accordance with the current version of V-Ray (3+)
This discussion has been closed.