LCT Dereverb™: Dereverberation has been a long-standing challenge within the communications and audio industry. The presence of reverberation, caused by audio reflecting off surfaces and objects, significantly degrades voice quality, speech intelligibility, user experience, and speech recognition accuracy. Despite the advancements in technology, dereverberation remains an unsolved problem, and the majority of smartphone brands and video conference apps available in the market still struggle with noticeable reverberations as you can hear below.
To address this issue, LCT has developed an innovative solution known as LCT Dereverb™, which encompasses denoising capabilities as well. LCT Dereverb™ stands out as an advanced software solution that does not need prior knowledge of room size, speaker location, or microphone distance.
LCT Dereverb™ is now available for licensing or purchase by smartphone, chips, IoT, speech recognition, voice control, and automobile companies. It can convert a smartphone to a conference phone, significantly improve voice quality in video conference, support far-field voice capturing, and improve speech recognition accuracy. For more information, please don’t hesitate to contact us at admin@licreativetech.com. We look forward to discussing how LCT Dereverb software can improve your product audio and increase your market shares.
|
Demo 1 – Large Conference Room
|
Raw recording (iPhone 12 Pro): After LCT dereverberation and denoise: |
|
Raw recording (Regular mic, no noise reduction) After LCT dereverberation and denoise: |
|
|
Demo 2 – Small Conference Room
|
Raw recording (Blue Yeti, Condenser Mic): After LCT dereverberation and denoise: |
|
Raw recording (iPhone 12 Pro): After LCT dereverberation and denoise: |
|
|
Raw recording (Samsung Galaxy S20): After LCT dereverberation and denoise: |
LCT Voice Separation: The objective of this task is to separate the voices of two or more speakers into distinct audio output channels, ensuring that each channel predominantly contains one person’s voice. To accomplish this, multiple microphones and various audio DSP techniques are employed to achieve optimal performance. Over the years, LCT has conducted extensive research in this field, resulting in outstanding and unparalleled results as evaluated by our customers. We offer two options for custom implementation: providing software for integration with customer’s existing microphone setup or delivering the CrispMic® PCB or device with pre-installed software. The following demonstrations exemplify the application of voice separation in vehicle. You can expect the same or better results for in-office voice separation because offices usually have less background noise than cars.
For detailed information of CrispMic®, please visit www.crispmic.com.
|
Demo – In-Vehicle Voice Separation:
|
Recoding of a regular microphone, purchased in the market, at the same location as the CrispMic. The regular microphone output is mixed two-speaker’s voice with background noise: Recording of LCT CrispMic microphone (Model No: CrispMic-Interview). CrispMic has stereo output, where the Left channel is Speaker 1’s voice and the Right channel is Speaker 2’s voice. Background noise is cancelled in real time: |
LCT Audio Solutions and Customer Services
Upon request, LCT can bundle LCT Dereverb with other LCT audio DSP functions as a all-in-one solution to meet customers’ needs in audio signal processing:
- Dereverberation
- Denoise / noise reduction
- Microphone array
- Acoustic dynamic beamforming
- Voice separation
- Adaptative noise cancellation
- Echo cancellation
- Equalization
- Automatic gain control
- Spatial audio; immersive audio
- 3D Soundstage audio