Scientists develop a system to synthesize realistic sounds for computer animation

An Interesting tool to synthesize realistic sounds for computer animation

Scientists develop a system to synthesize realistic sounds for computer animation

An Interesting tool to synthesize realistic sounds 

Recently, several parameters have been removed from the model of mechanical contact of the sound wave to provide dynamic sound synthesis control from virtual environment simulation. Sound can be synthesized in real-time using rapidly busy waveform changes. Many advanced models and perceptual experiments have been described to validate the synthesis of sound veracity. Developed models and results show proof of concept and show the likelihood of this approach.

The computer-generated imagery usually depends on the sound recorded to complete the illusion. When they collide, the objects vibrate, which can tilt further into the grip and the sound. It is computationally expensive to solve all the frictional contact incidents between fast vibrational objects. In order to speed things up, their algorithm imitates only the number of contacts and vibrations to synthesize the sound. However, the recording can limit the sounds produced by you, especially in the virtual reality environment in the future where you can not always know beforehand what action will be. To synthesize the collision sounds and noise of confrontation calculates computer forces that will apply the objects generated from the computer, if they were genuine, then how do these powers vibration objects and how the companies move in the air to make the sound. Progress in computer-generated imagery has brought vivid, realistic animations to life, but sounds related animations simulated on screen, such as two object collisions, are often recorded. Now researchers from Stanford University have developed a system that automatically provides accurate sounds for a wide variety of animated events.

Realistic sounds for computer animation

Stanford University researchers said in a major conference on computer graphics and interactive techniques, "In previous efforts, it is often considered that the contact objects were rigid, but in reality, there is no such thing as a hard object. Depending on the part of this sound synthesis system will present its work. Films and virtual reality are alive to the world. Apart from creating, this system can also help in the prototype of engineering companies that how the product will sound before being physically produced and hope to encourage design, calm and less disturbing.

Doug James, a professor of computer science at Stanford University, developed computer algorithms to synthesize sound on "fly" based on simulated physics models. Their research group has already invented sound synthesis methods for liquids, fracturing glass and other solid substances, and noise-thin shells such as sheet metal. Now he has composed ways to synthesize more realistic voices of hard work and roar of the fire.

Mr. Doug James said in a statement: "Being able to simulate reality for humans is a holy grail in calculating. We can animate scenes and present them in terms of physics and computer graphics with vision, But, generally, they usually seem to be, at present, realistic synchronized for complex animated content, such as splashing water or collision objects. To generate sounds, there is no way. It fills the void. He also pointed out that "you set a box for so many people who contact it on a flat surface. We identify a small set of physically loyal contacts, which is the easiest set to accomplish the task. Similarly, algorithms select objects in many different ways to vibrate, which ignores people who produce an inaudible sound ".

Jui-Hsien Wang, a graduate student in the Jame's Lab at the Institute for Computational and Mathematical Engineering (ICME), explained "I have tried to solve partial differential equations - that governs how well the hands propagate It's actually a place where you do not solve the equation but after doing it you can actually hear it. This is really exciting and funny for me. 

Predicting sound

Predicting realistic sounds for computer animation

What is Predicting sound

The system tells the vibrations of each object as statistics, such as geometry and notice at physical speed, such as loudspeakers, they vibrate sound waves. It calculates pressure waves by fast-moving and vibration surfaces, but the room does not repeat the acoustics. Therefore, although it does not rebuild the echo in a grand cathedral, it can solve elaborate sounds from scenarios such as a crashing symbol, a bouncy bowl spinning on one stop, glass filled with water or a virtual character talking into a megaphone device.

"We essentially have a render button with a pre-processing, which treats all objects together in an acoustic wave simulation," said Ante Qu, co-author of the paper and a graduate student in James' Laboratory. The method is highly detailed, it takes into consideration the sound waves generated by each object in the animation, but also estimates that in the sequence Those waves based on their interactions with objects and sound waves, how are the bends, which throws or die.

Most sounds related to the animations rely on pre-recorded clips, which require a huge manual effort to synchronize with the action on the screen. These clips are limited to existing noise - they cannot predict anything new. Other systems that produce and predict, look exactly like James and his team, work only in special cases, or assume that geometry is not very bad. They also require a long pre-calculation phase for each separate object.

 Are there any Challenges ahead?

Mr. James said, "The sound of the first water produced with the system was the best invention that we emulated - and water is a major challenge in the sound generated by the computer. We thought we had a small improvement But it is dramatically better than previous approaches, also it was really striking in a great way".
In its present form, it takes some time to create a finished product in the group process. But, now that they have proven the capability of this technique, they can concentrate on performance optimization, such as implementing their method on parallel GPU hardware, which should make it faster.

The members of the group say, "Although the work of the group has presented the voice of various objects spinning, falling and colliding in each other, more complex objects and conversations - inverted sound-realistic models of a Stradivarius violin". So they have to wait for future solutions.

By: Mahtab Alam Quddusi