I am working on a project in which I need to analyse acoustic data taken from voice recordings of a person in different rooms. I wanted to know if there are any parameters that are independent of the environment and therefore do not change by changing the room setting (e.g. standard deviation of acoustic pitches). Also, I would like some reference if you have.
many thanks!
There probaly are, as we would recognise the voice of a person we know through almost any room setting. An example of this is the ability trained listeners have to discriminate between loudspeakers independantly on the room setting. A good source on this is Floys Toole. You may find a talk on youtube, or get his book. Scientific or popular science articles are available too, often in collaboration with Sean Olive at Harman.
it depends on purpose of your acoustic analysis or method used i guess.
If you are trying analyze voice of the speaker, then ceptral measures may remain constant( CPP) or HNR as well. F0 may remain similar, but other parameter such as jitter and shimmer may be influenced by noise and revrbration of the room.
If it is speech of speaker is the focus of analysis, then intonation pattern , no of pauses, may remain unaffected by room changes.
Anders Buen and Suresh Thontadarya, thanks a lot for your useful answers! I'll check about Floys Toole, and I'll try to explore better the features you suggested. Many thanks
- Badges
- Science topic
- Similar topics
- Counseling Psychology
- Confusion