To fais may be useful wants to rend

To fais may be useful





wants to render a vectorial scene through raster images. Unfortunately an unavoidable information loss, due to the different way of representing objects occurs, as a consequence of the vectorial-raster passage. In order to minimize the problem of pixel use, that is, the typical “jaggedness” of image edges, very high rendering resolutions could be used: thanks to higher resolution a table edge could then be less jagged. Of course this is a good way to handle the problem, but it involves a greater demand of hardware and time resources. In any case, aliasing occurs using very high resolution as well.
There are two main aliasing types:
Spatial aliasing: when artefacts are considered in the single static image;
Time aliasing: when artefacts are considered in a sequence of images (animations). This problem leads to shaking or flashing objects. Sometimes it might occur in particular situations, such as the animation of a wheel turning in the opposite way compared to the direction it is moving in. This problem is due to the fact that small objects appear in some frames but not in some others: this is typical phenomenon coming from an incorrect antialiasing.
As everybody knows 3ds max works with primitives, tiangles and lines: these are adjoining objects, because of their vectorial nature. They turn to a pixels ensemble during the raster only. This turning is named sampling and it yields a visual artifact called aliasing. During the sampling process the color of a pixel is fixed referring to the centre of the pixel itself. The central point of the pixel is taken from the geometry lying behind the pixel grid fixed by the output resolution.
Figure 3.29 Aliasing is a phenomenon that takes place trying to built a continuous image through the use of discrete values: information loss occurs, and this leads to poor image quality and to jaggedness.
Figure 3.30 A low resolution image. After the rasterization the image appearing in the monitor is unintelligible. Every square is a pixel that takes the color from the vectorial geometry lying behind. It takes the color corresponding to the pixel’s geometric center, represented by the while dot.

176/p161
The aim of the antialiasing technique is to improve image quality by reducing or removing artefacts due to aliasing
There are a several techniques for the computation of antialiasing. One called oversampling, also called supersampling.
Through supersampling each pixel is divided into subpixels. The color of each subpixel is separately computed. To obtain the final color of a pixel it is necessary to compute the average of the colors of each subpixel. The ensemble of subpixels is called subpixel mask (e.g. 2x2=4 subpixels, 4x4=16 subpixels). Antiliasing quality is higher, the higher the number of subpixels for each pixel. A 4x4 mask is supposed to lead to a very good visual quality, but requires a large raster time.
Figure 3.31 the subdivision concept is represented in this diagram. It is a frequent parameter among Vray’s tools. The subdivision in smaller parts, both for AA and for Area Shadows etc… the real number of subdivisions is given by the square of input value. Subpixel scheme.
Figur 3.32 vectorial line rendered before and after antialiasing.
In the upper example, all the pixels of the render are divided in subpicels using a 3x3 subpixel mask, even those that do not need it, such as totally black ones or the ones without any vectorial data. Incase it is possible to avoid the subpixel division and therefore aloe the AA calculation to be faster. This is why the idea of adaptability has been introduced. Based on a threshold fixed bye the user, Vray seeks out the pixels of the image which require antialiasing. After, these pixels are divided in subpixels.
Vray tries to find the subpixels that need another subdivision, skipping the sufficiently detailed ones, according to the threshold fixed by the user. This process continues until the maximum subdivision number set has been reached.

177/p162
Figure 3.33 System of Adaptive Sampling
Working only on trictly required area allows less computations and therefore faster renderings. This way, in the red segment example, AA computation time in the black undetailed area would be much shorter, while Vray would exclusively focus on red pixels or on those directly adjoined.
VRay provides another AA calculation technique, also adaptive, but which uses a method called undersampling. Until noe the pixel was divided into parts, 2x2 subdivision means a pixel division in 4 parts; while a 5x5 subdivision means 25 subpixels, without adaptability, and so on. The undersampling works in the opposite direction: one sample is used for more than one pixel. For instance, applying a-1 undersampling value, 1 sample is used for 4 pixels, applying a-2 undersampling value 1 sample is used for 16 pixels etc…plainly this method leads to excellent quality/time results in the case of an undetailed scene or one presenting very smooth surfaces.
Figure 3.34 Undersampling is not just used for AA baut for Irradiance Map and for glossy effects of materials as well.
Vray provides 3 types of algorithms for antialiasing computation: Fixed rate sampler, Adaptive QMC sampler and Adaptive subdivision sampler. Moreover almost all 3ds max filters are supported.
Figure 3.35 On this board it is possible to choose the sampling type, in order to avoid aliasing on the rendering.

178/p163
Parameters
Image samplers
Fixed rate sampler
It is thee asiest sampling method. It computes the AA of each individual pixel, according to the number of subpixels set. It globally performs a subdivision of all pixels in the image, without any adaptability. Regardless of geometrical detail, each pixel is always divided into the same number of subpixels.
Subdivs – it fixes the number of samples for each pixel. If this value is 1, then only sample is considered for each pixel. If the parameter subdivs is greater than1, then each pixel is divided bye the square of the set value. It represents an AA direct calculation method. It is advisable to use this AA model when working with images full of effects, such as Area Shadows, SSS, Motion blur, DOF and very complex textures.
Adaptive QMC sampler
This algorithm works on a varying number of samples for each pixel, according to the intensity difference among the neighbouring ones. Like all blurry effects, this method is closely influenced and related to the parameters in rollout Vray: rQMC sampler, especially the adaptive amount and noise threshold parameters. The first of these determines how much the sampler is adaptable: adaptability is greater when this parameter is higher. The second one dealks with choosing which subpixels is useful or not for antialiasing computation: the quality is higher as the value is lower and vice versa. It is recommended not to use values lower than 0.0002. values which are too low result in a very long rendering time. It is preferable to apply the adaptive QMC sampler to very detailed image, like VrayFur, glossy, depth of field (DOF), Motion blur etc… moreover, this sampling uses less RAM than the next method: the adaptive subdivision sampler.
Min subdivs – it fixes the minimum sample number to be used in each subdivision of pixels. The higher its value, the more samples are considered. Therefore quality is higher at the cost of penalizing rendering time. This parameters sets the minimum number of subdivisions per pixel below which Vray is not allows to go. Obviously the minimum value is 1. Increasing its value, e.g. Subdivs = 2, Vray uses at least 4 subpixels for each pixel to perform the antialiasing computat

Hasil (Bahasa Indonesia) 1:

To fais may be useful wants to render a vectorial scene through raster images. Unfortunately an unavoidable information loss, due to the different way of representing objects occurs, as a consequence of the vectorial-raster passage. In order to minimize the problem of pixel use, that is, the typical “jaggedness” of image edges, very high rendering resolutions could be used: thanks to higher resolution a table edge could then be less jagged. Of course this is a good way to handle the problem, but it involves a greater demand of hardware and time resources. In any case, aliasing occurs using very high resolution as well.There are two main aliasing types:Spatial aliasing: when artefacts are considered in the single static image;Time aliasing: when artefacts are considered in a sequence of images (animations). This problem leads to shaking or flashing objects. Sometimes it might occur in particular situations, such as the animation of a wheel turning in the opposite way compared to the direction it is moving in. This problem is due to the fact that small objects appear in some frames but not in some others: this is typical phenomenon coming from an incorrect antialiasing.As everybody knows 3ds max works with primitives, tiangles and lines: these are adjoining objects, because of their vectorial nature. They turn to a pixels ensemble during the raster only. This turning is named sampling and it yields a visual artifact called aliasing. During the sampling process the color of a pixel is fixed referring to the centre of the pixel itself. The central point of the pixel is taken from the geometry lying behind the pixel grid fixed by the output resolution.Figure 3.29 Aliasing is a phenomenon that takes place trying to built a continuous image through the use of discrete values: information loss occurs, and this leads to poor image quality and to jaggedness.Figure 3.30 A low resolution image. After the rasterization the image appearing in the monitor is unintelligible. Every square is a pixel that takes the color from the vectorial geometry lying behind. It takes the color corresponding to the pixel’s geometric center, represented by the while dot.176/p161The aim of the antialiasing technique is to improve image quality by reducing or removing artefacts due to aliasingThere are a several techniques for the computation of antialiasing. One called oversampling, also called supersampling.Through supersampling each pixel is divided into subpixels. The color of each subpixel is separately computed. To obtain the final color of a pixel it is necessary to compute the average of the colors of each subpixel. The ensemble of subpixels is called subpixel mask (e.g. 2x2=4 subpixels, 4x4=16 subpixels). Antiliasing quality is higher, the higher the number of subpixels for each pixel. A 4x4 mask is supposed to lead to a very good visual quality, but requires a large raster time.Figure 3.31 the subdivision concept is represented in this diagram. It is a frequent parameter among Vray’s tools. The subdivision in smaller parts, both for AA and for Area Shadows etc… the real number of subdivisions is given by the square of input value. Subpixel scheme.Figur 3.32 vectorial line rendered before and after antialiasing.In the upper example, all the pixels of the render are divided in subpicels using a 3x3 subpixel mask, even those that do not need it, such as totally black ones or the ones without any vectorial data. Incase it is possible to avoid the subpixel division and therefore aloe the AA calculation to be faster. This is why the idea of adaptability has been introduced. Based on a threshold fixed bye the user, Vray seeks out the pixels of the image which require antialiasing. After, these pixels are divided in subpixels.Vray tries to find the subpixels that need another subdivision, skipping the sufficiently detailed ones, according to the threshold fixed by the user. This process continues until the maximum subdivision number set has been reached. 177/p162Gambar 3,33 sistem adaptif samplingBekerja hanya pada trictly wilayah memungkinkan kurang perhitungan dan karena itu lebih cepat rendering. Dengan cara ini, dalam contoh segmen merah, AA perhitungan waktu di daerah terperinci hitam akan jauh lebih pendek, sementara Vray akan berfokus pada merah piksel atau orang-orang yang secara langsung terhubung.VRay menyediakan AA perhitungan teknik lain, juga adaptif, tetapi yang menggunakan metode yang disebut undersampling. Sampai noe pixel dibagi menjadi bagian, 2 x 2 subdivisi berarti Divisi piksel dalam 4 bagian; Sementara subdivisi 5 x 5 berarti 25 subpixels, tanpa kemampuan beradaptasi, dan sebagainya. Undersampling bekerja di arah yang berlawanan: satu sampel digunakan untuk lebih dari satu piksel. Misalnya, menerapkan nilai a-1 undersampling, contoh 1 digunakan untuk 4 piksel, menerapkan a-2 undersampling nilai 1 sampel digunakan untuk 16 piksel dll... jelas metode ini mengarah ke hasil kualitas waktu yang sangat baik dalam kasus adegan terperinci atau satu menyajikan permukaan yang sangat halus.Gambar 3,34 Undersampling tidak hanya digunakan untuk AA baut untuk Irradiance peta dan mengkilap efek dari bahan-bahan serta.Vray menyediakan 3 jenis algoritma untuk perhitungan antialiasing: tetap tingkat sampler, adaptif QMC sampler dan adaptif subdivisi sampler. Selain itu, hampir semua 3ds max filter yang didukung.Gambar 3.35 di forum ini mungkin untuk memilih jenis sampling, untuk menghindari aliasing pada rendering. 178/p163ParameterGambar SamplerSampler tingkat bunga tetapIni adalah metode sampling asiest engkau. Itu menghitung AA setiap piksel individu, sesuai dengan jumlah subpixels set. Global melakukan subdivisi dari semua piksel dalam gambar, tanpa adaptasi apapun. Terlepas dari geometris detail, setiap pixel selalu dibagi menjadi jumlah yang sama subpixels.Subdivs-ini memperbaiki jumlah sampel untuk setiap pixel. Jika nilai ini 1, maka hanya contoh dianggap untuk setiap pixel. Jika parameter subdivs adalah 1 lebih besar, maka setiap pixel adalah dibagi bye kuadrat dari nilai yang ditetapkan. Ini merupakan metode perhitungan langsung AA. Disarankan untuk menggunakan model AA ini ketika bekerja dengan gambar-gambar yang penuh dengan efek, seperti Area bayangan, SSS, Motion blur, DOF dan tekstur yang sangat kompleks.Adaptif QMC samplerAlgoritma ini bekerja pada berbagai jumlah sampel untuk setiap pixel, menurut intensitas perbedaan antara yang tetangga. Seperti semua efek kabur, metode ini erat dipengaruhi dan berkaitan dengan parameter dalam peluncuran Vray: rQMC sampler, terutama jumlah adaptif dan kebisingan ambang parameter. Yang pertama ini menentukan berapa banyak sampler beradaptasi: adaptasi lebih besar bila parameter ini lebih tinggi. Dealks kedua dengan memilih subpixels yang berguna atau tidak untuk perhitungan antialiasing: kualitas yang lebih tinggi seperti nilai lebih rendah dan sebaliknya. Disarankan untuk tidak menggunakan nilai-nilai yang lebih rendah daripada 0.0002. nilai-nilai yang terlalu rendah mengakibatkan waktu sangat lama render. Ianya lebih baik untuk menerapkan sampler QMC adaptif untuk gambar yang sangat rinci, seperti kedalaman mengkilap, VrayFur, field (DOF), Motion blur dll... apalagi, sampling ini menggunakan RAM kurang daripada metode berikutnya: sampler adaptif subdivisi.Min subdivs – ini memperbaiki nomor sampel minimal untuk digunakan di masing-masing subdivisi piksel. Semakin tinggi nilai, sampel lain yang dianggap. Oleh karena itu kualitas lebih tinggi dengan menghukum render waktu. Ini set parameter jumlah minimum subdivisi per pixel di bawah ini yang Vray tidak memungkinkan untuk pergi. Jelas nilai minimum adalah 1. Meningkatkan nilai, misalnya Subdivs = 2, Vray menggunakan minimal 4 subpixels untuk setiap pixel untuk melakukan antialiasing computat