iKALiZER is an audio engine for softwares, applications, games, R&D, advanced VoIP...
It provides highest audio quality, ultra low latency and others exclusive features.


Under development iKALiZER "generation 2"
The audio engine is divided in 4 parts while using the output mode :
- Modeler
- Mixer
- Assembler
- Writer

The modeler is a "real world audio modeling unit" for objects, materials, geometry, etc.
- Audio is simulated with an advanced ray-tracing engine based on waves and particles.
- All programming needs are provided by a very simple and easier dedicated reality modeling API.
The modeler creates an audio transform like a long term FIR filter.

The mixer could be compared to the most advanced digital mixer with wide routing and FX possibility.
It also provides sample rate conversion (using Lanczos and Riesz interpolation) and format conversion.
Routing possibilities are based on multiple submix, inserts...
The effect processing is based on an user programmable audio DSP and the most used effects are already embedded in its internal memory like occlusion, obstruction, absorption, reflection, distance attenuation, reverb/echo, doppler/shifting, EQ/FIR filter...
Under worst conditions of fast parameter changes, some distorsions could be heard. A smooth transition processing always maintains the highest possible audio quality by avoiding these issues.

The assembler transforms all high level computing commands to be used by the processor.
This processing is done using the iKALiZER embedded JIT x86/x64 compiler.

The writer incorporates 2 features : Output audio transport and device driver.
Output audio transport performs the final audio processing (compressor/limiter) and format conversion (with dithering).
The device driver is an abstraction layer between Microsoft® DirectX®, Microsoft® WaveOut, Steinberg ASIO™ and Linux OSS/ALSA (1).


Other features :
- Ultra low-latency mixing with auto-adaptative buffer underrun management.
- 32bit high quality audio.
- 2D(2) and 3D spatialization(3) using HRTF, Holographic, Bauer delay, ITD, IIT-ILD and panning processing.
- Easy message based API.
- Capture/record capability with DSP processing.
- Multithread architecture.
- Microsoft® Windows® and Linux support.
- System based application with configuration panel.
- Under study "Off-loading server side processing.

(1) x86 Apple® Mac OS™ support could be possible in future.
(2) Best results with headphones and 4+ channels (4.0, 5.1...).
(3) Only works with headphones.

Note : Possible iKALiZER "gen 1.5" could be used the Mixer, Assembler and Writer


Currently available iKALiZER "generation 1" features :
- Full 2D/360 spatialization
- x86/x64 Windows 2000/XP/Vista support
- High speed / high quality 32-bit mixer
- Ultra low-latency mixing with auto-adaptative buffer underrun management
- Audio limiter (up to +20dB gain)
- Allow dynamic distance (air absorption and distance attenuation) and underwater effect
- Smooth transition effects
- Sample rate conversion (SRC) with 16-point Lanczos window interpolation
- Dedicated thread (allow good use on dual/quad core)
- Highly optimized voice mixer for SSE
- Preemptive dynamic voice allocation
- Hybrid HRTF for Headphones and 2-speakers
- Holographic and Virtual holographic mode for Headphones
- DPL I & II realtime compatible encoder
- Quadriphonic on 4-speakers set
- 5.0 and 5.1 home-theatre sound
- Hybrid 32bit API via Windows DLL
- Closed source library
- Could be used in Urban Terror or ioQuake3
Note : iKALiZER "generation 1" is only a proof of concept.

 
 
Javascript is currently disabled. You need to enable it in order to use this website.