Sound Ecology and Acoustic Health, Part 4: Time Domain Analysis
Bibliographic record
Abstract
Introduction to periodical article: We have spent the last while working towards a mobile phone application to help identify a local noise nuisance problem. We joked with Mike’s neighbour’s kids that the record and play-back .3GPP file WAT_AN_APP application, Fig. 1A, was to impress them that “Without Any Teenage Assistance Necessary, we could write an Android APP” CC Issues ###. We then added just enough additional code (JEAC) to store an audio record for later analysis CC Issues ###. To continue the friendly tease in CC Issues ###, we pretended that the project code was actually designed to detect – “Things that Go BOOm at Night+ - how many “TGBN ghosts” there are in the neighbourhood (Fig. 1B). As they say “Be careful what you wish for!” Our neighbours got interested in the community noise issues we were really trying to measure. They had their teenagers explore the acoustic health of their home using our work-in-progress. Late yesterday, a knock on the door revealed our neighbours asking for help. Their eldest teenager had gone to the University of Pennsylvania. According to the Penn Arts and Sciences website sites.sas.upenn.edu/ghosts-healing, a group of scholars from literature, art history, nursing, archaeology, religious studies, science and medicine wants to take research on ghosts seriously. So our neighbour’s kid decided to volunteer with this group. This turned into a term project -- working on analysing room acoustics as a possible source of “that friendly spectral feeling”. Hence the frantic email message they wanted to pass on -- Term’s nearly over! Could you please get Mike to hurry up and fulfill his promise in that first CC article of providing enough information to do some “real” digital signal processing (DSP) analysis? While he was at it – could he get Adrien to add some graphics’ capability to display the frequency characteristics of the sounds in a room to make my term report more interesting! In Canada, it always good to keep on the right side of the neighbour’s kids as they are a good (inexpensive) labour source for shovelling snow off sidewalks. So we decide to write a RoomAcoustics Analysis Capability addition. Actually we wanted to be able to say that we had Penn-ed some code (sorry for the pun ( :-). ). First we will explain how to reliably excite a room resonance that can be captured by our existing TGBN detector code. We will graphically display the room audio signal to give us a first chance to compare resonance characteristics in different rooms. We found that looking for small differences in the captured signals displayed as a function of time meant working (slowly) with a lot of data. So we added a way to generate frequency information signal of captured signals using a discrete Fourier transform (DFT) algorithm code we grabbed from the web. Fig 2A shows the background noise recorded in our university lab. Having noticed a possible small 727 Hz ghost sleeping next to our desk, we tried to move around the room to better record its characteristic, Fig. 2B. The frequency characteristics of our two records look too similar for us to be sure that we have a non-snoring ghost close by. We decided to wake it up by outputting a three second Chirp, a sound burst from 50 to 1000 Hz. Fig. 2C show the frequency response of the Chirp signal, but there is not much there other than measuring the poor low frequency of our phone’s speaker. However, we accidently got close enough that we woke up the sleeping ghost which significantly changed the frequency response of the room, Fig 2D.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.004 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.003 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.017 | 0.005 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".