3D Sound Directshow Filters
These Microsoft Directshow filters are what enables the 3D spatial sound application to function. Directshow allowes for a simple multimedia streaming model that could easily be modified to perform different spatial sounds algorithms. The following is a list of filters that demonstrate the different algorithms that I worked with to perform positional audio.
PLEASE NOTE: You will need the latest Microsoft Windows Media Player for these filters to work
- Interaural Time Delay (ITD) Filter
- Interaural Level Difference (ILD) Filter
- Positional Audio Filter using the ITD and ILD
- Fast HRTF Filter (Full Positional Audio)
Interaural Time Delay (ITD) Filter
The ITD is a major cue used by humans in localizing sound. It is based on the principal that there will be a slight delay in time between the sound waves hitting an ear depending on its location. If the sound is to the left side of the head, the sound wave will reach the left ear before the right, causing a small delay of about 0.6 ms to occur. This filter models this effect. Major features include fast response time (under 10 ms) to user input, as well as the ability to work on mono or stereo sources of any sampling frequency. It also includes a simple property page that can be used to change the sounds position shown below:
Click here to download the ITD filter
Click here to download the ITD filter source
Interaural Level Difference Filter
The ILD is a major cue used by humans in localizing sound. It is based on the principle that a “head-shadow” effect will occur on a sound depending on its location in space. If the sound is to the left side of the head, the sound wave reaching the right ear will be attenuated as well as lose some of its high frequency content. This filter models this effect. Major features include fast response time (under 10 ms) to user input, as well as the ability to work on mono or stereo sources of any sampling frequency. It also includes a simple property page that can be used to change the sounds position shown below:
Click here to download the ILD filter
Click here to download the ILD filter source
Positional Audio Filter
The following is a DirectShow audio filter that demonstrates the effects of both the interaural time and level differences. (ILD and ITD). These two cues compromise a great deal of an individuals head-related transfer function (HRTF). For more information on what each is, you can look at the individual filters for both the ITD and ILD filters. Major features include fast response time (under 10 ms) to user input, as well as the ability to work on mono or stereo sources of any sampling frequency. It also includes a simple property page that can be used to change the sounds position shown below:
Click here to download the Positional Audio filter
Click here to download the Positional Audio filter source
Fast HRTF Filter
The HRTF is the head-related transfer function. The sound waves that reach the human ear drum are greatly affected by diffraction caused by the torso, shoulders, and outer ear (pinna). This diffraction causes a modification of the spectrum of binaural sound signals, which the human brain can interpret for sound localization. The spectra change can be modeled by the HRTF, which is a complex function that varies not only with direction, but is also different for each individual. In the time domain, the same information can be expressed as the head-related impulse response, or HRIR. At the CIPIC labs, extensive research has been done in collection dense sets of FIR filters of a large number of individuals HRIR’s. These filters are taken at 1250 unique positions in space. Each filter is compromised of 200 points FIR filters for each ear, left and right. These filters must be convolved with a sound source to position it in space around a listener. This takes roughly 70 million mathematical operations per second to do this in real-time at a sampling rate of 44.1 kHz. This is usually done using specialized DSP hardware. This filter demonstrates how it is possible to perform positional audio on a personal computer. It is designed with both standard and MMX enhanced algorithms for increased speed. Major features include fast response time (under 10 ms) to user input. This means the filter works completely in real-time with no delay when changing sound positions. It also includes a simple property page that can be used to change the sounds position shown below. The two sliders or pictures can be used to change the location of the sound in 3D space in the azimuth and elevation planes.
Click here to download the Fast HRTF filter
Click here to download the Fast HRTF filter source