This is the wiki for a new multi-year project in LAB lab on phonetic adaptation. The proper full name of the project is Adaptations in Conversations and it started circa mid-2022.
We have split the project into three main portions:
AdaptiCon-Vowel is mainly concerned about adaptations in English.
AdaptiCon-Tone is mainly concerned about adaptations in Cantonese tones.
AdaptiCon-AI is mainly concerned about adaptations between a human and a non-human voice.
This is a list of all literatures we have read and reviewed.
This is a list of useful resources we have found during our research.
Audio:
- Rode Wireless Go II with Lapel mics, each participant wears one.
- NT2-A, set to Cardioid mode, each participant gets one pointing to them.
- Recorder: TASCAM DR70D
- Note: there's self-noise from the mic
Computer (to be confirmed):
- Mac Mini with 2 virtual machines, each participant get one with separate controllers and displays
- 2 x VESA monitors, VESA to C-Stand mounts, and actual C stands
Video Recording:
- 2 cameras pointing at the participants face.
- 2 cameras outputs to ATEM mini pro, 2 monitors as well (using HDMI split).
- HDMI recorder (Atomos Ninja V) recording outputs of the ATEM mini pro.
Equipment list (TODO)
- General Noise Floor measurement
- NTG2 (Left/Top) vs NT2a (Right/Down) spectrogram comparison
