Technology will continue to develop, and we will continue to improve our (pulp)
November 19, Pulse, Glove Lead Engine Programmer participated in the Gustav and lifted the various technologies that were used for next-generation engines and developed. And explained how new technologies have been applied to new projects such as
High Gwanghyun Lead Engine developers said they have been in a new game project, while there was a need for many parts, such as light sources, flavors, and visual effects, and have focused on meeting these needs efficiently. He also said that it was also focused on the composition of multi-platform environments like the
First, the lightning effect. At the heart of rendering, it is necessary for minor expression and realistic expression. Pulse allowed the bis to reflect the lightning effect in real time without prior treatment to implement a fast lightning effect. Direct lighting directly seen in the light source introduced a cluster shading technique.
The area light introduced AIS to minimize the ALU to minimize the ALU, and provided quad, disk, tube, sphere form for expressions such as fluorescent lamps, electroluminescence, and light gastrointestinal expression. Rayamachid and La tray Shadow techniques have also been applied.
The IES profile is a file form that recorded optical compliance on the light source space. And using the physical unit to add a realistic expression. Direct lighting shadows have been applied to Cash Capacity Shadow Map, Mashing Shadow, Ream arch / Rat Lased Shadow.
Medal Shadow provides texture space in a pool form. A structure that is reconstituted by a distance to the camera movement. Hundreds of light sources can be realized in real time, and the point format is composed of a destination form and more detailed than the tetrahedron.
Next, Ray March Shade. Massive shadows may cause problems with performance if the number of light sources is too large. So I introduced Ream arch Shadow. It is a method of checking the heater to be hit before meeting the light source. The expression that is difficult to represent is treated as a signed distance. Up to 32 per cluster is applicable.
The character uses a capsule shadow representation. A structure that is processed when the sign is tracing. Ray Trace Shadow uses a structure such as ray March Shadow. It is relatively fast and light, using an anime hit query.
Indirect lighting uses a voxel-based scene representation. There was some improvement in dingoes. Consisting of six clip maps and the smallest bosses is 25 cm and the horizontal reference to 512 m. Depending on the purpose, you can configure multiple forms of clip maps.
First, the Defiance voxel is that the diffuser light results are stored in the voxel. Express the multi-bounce effect over multiple frames. Off-clean dispersion processing is a function that represents a region other than the screen. Below is the process to which Radiance voxels are applied to the final result of the
The InidirectidieDions catch voxel only stores a diffuse irregular lightning that does not have an albedo applied. Sky Non-Geographical Boxer is used for sky representation. It is generated with Indirect Indians voxel. The signed distance voxel is utilized in a general signed distance field structure and is also used for crash processing of particle PhD.
Accessible — Alliance Distance Boxer was also introduced for fast tracing. X, Y, Z Store the number of empty voxels to each axis. It is produced in a manner that is hierarchically merged through the tubes, and is advantageous for Ray Malay that proceed parallel to the surface. Diffuse is used for distance representation of Endemic Lightning Ray match.
Diffuse Inidotranting performs spear tracing in the signed distance voxel. If the ray distance is distant, the tracing using the Axis-aligned DESIREE only is applied. Somewhat roughly, but you can handle a wide range of areas. In addition, you can use the screen depth-ricent result to add details to a signed distance.
Lateraling was applied using a voxel cache without a separate shading. Relatively lightweight can be performed at a light cost.
The specular incidirectleptes first use screen depth as in the screen space reflection, as well as diffuse indirect lightning. You can also resemen the cue view map to the configuration. Dino aging is made through a recircular blur and wavelet filter.
The following are waiting. Integrated atmospheric expressions are applied, various atmospheres, heavens, sun, moon, clouds and fogs, and so on. The core is that it is not to treat atmospheric phenomena, not processing at a time. Local, blended two ways.
Near atmospheric expressions use proxies. Direction light shadows use a cascade shadow map. The shadow that occurred in the fog is calculated as the density injected in the proxies. If you want, you can place a variety of fogs. Implementing foggy shapes on clusters and finally optimized.
The photocell is not suitable for remote atmospheric expressions in memory or performance. Ray matches will express the air atmosphere. Air, aerosol, and cloud density included all in one Ray Match. After separate treatment, it is possible to solve the disconnection and awkwardness that can be mixed when mixed.
The following screenshots are a photograph that expresses other weather and time zones in the same place using the art.
I also implemented a new way of upscale. Check the minimum of the 4×4 pixel range, and the maximum depth range is checked to calculate the 16 depth slices in one Ray Match. Therefore, it is a method of representing a slice corresponding to the depth on a full screen. The sky calculates the direction of the pixel position with 16 frame variance.
Through fluid simulation, we could express the naturalness of fog or smoke. The rendering was made through the proxies, and the interaction with the surrounding objects has become natural.
We introduced a MOVED ACTION system for the ability to express a distance. LOD was granulated, and Importer was improved. Millions of objects can be real-time rendered.
Improvement of Imposter was able to extract information from the original, and to reconfigure the information on the scene. For efficient level production, the procedural creation system was automated as much as possible, and the overall level change has been relatively easy. Millions of invitations are automatically generated and deployed, and the correct environmental representation is possible at all when changing settings.
A level of proxy for extreme remote expression also configured and configured the proxy to match the level layer. If the proxy is gained in the area, the sub-layer proxy is loaded after searching. It is important to care about the display of detailed geometric expressions or height maps. I posted the detail through the shadow treatment.
First, the tubular water that is implemented to the present through the physical engine is a four-wheel drive vehicle, carriage, ship, motorbike. Internally, I have been using the subcomponents, and improved the vehicle system based on the base. Simulation stability has also been added and strengthened the naturalness of driving. Some improved systems will be applied to
Pose rigging can be used through setting in DCC Tool. Movement such as gears and pistons of the mechanic character can be expressed in real time. You can maintain the piston movement even if you modify the character movement. It was also used for helper purposes to improve skiing quality.
The brake cover is part of the partial destruction by physical impact. To this end, the appearance of the destruction system in the
Character fabric motion simplifies conflict calculations using an efficient increase in prostration, and protective contacts through long range advice. Minimizes the failure of collision detection at a large relative speed naturally through it.
There is also a subsystem. The maternal system can be used for a shading system that can be controlled by art, and various directing can be linked in conjunction with the sequence editor. For example, I heard the metabus production of the
There was also an improvement in the way to find directions and AI functions. AI obstacle overcoming has been added, and large-scale movements implemented through group-specific flow fields. There was also an improvement in collision avoidance or moving path. Vehicle / pedestrian road moving + signaling system was added.
Interaction with ambient objects for realistic expression was also strengthened. It is possible to express the character of the character, and the fallen leaves are in the foot, You can also express surrounding physical phenomena, such as detection, magic, and explosion. Interacting with wind direction and intensity.
Are you also interested in topics related to Glovi Lead Games?
Gogwanghyeon Reid stated, There are many more features, and noted that technology will continue to evolve and continue to improve our quality also. On many platforms, he said, and prepared to function this function.