What This Project Actually Is
San Angelo Odyssey is a multimedia oral history and mapping project that documents the socioeconomic shifts, cultural memory, and infrastructural changes in Tom Green County, Texas. It was initiated around 2021 as a collaborative effort between local historians, the San Angelo State University oral history program, and a small team of independent researchers. The scope covers roughly three decades of material change, from the early 1990s oil price collapse through the recent diversification into wind energy and healthcare sectors. The core deliverable is an interactive web platform paired with a collection of recorded interviews, geolocated photographs, and municipal document scans. The project exists because standard historical records for mid-sized Texas cities tend to gloss over the lived experience of working-class residents, immigrant communities, and people affected by industrial decline. The official city archives focus on mayoral records, zoning approvals, and development deals. Nobody was systematically recording what happened to the families on East Side streets when the major employers pulled out or moved operations elsewhere. I spent about fourteen months contributing raw interview transcripts and field notes to the project, mostly around the North Central district and the area near the old textile mill sites. The process of getting from a raw audio file to something the platform actually uses involves several steps that are not well documented anywhere.
How The Interview Collection Process Works
The project uses a standardized consent and metadata framework before any recording happens. Each participant signs a digital release that specifies exactly how their words can be used, what time limits apply to archival retention, and whether they want their location data attached to their contributions. This is not optional. I once tried to do a quick hallway interview at a community event without the full paperwork because the interviewer on site had left early. The research coordinator rejected the entire file two weeks later during quality control, and we had to rebuild that segment from scratch with a formal appointment. Recording setup is straightforward but specific. The project requires WAV format at 44.1kHz minimum, stereo, with no compression. You cannot upload MP3s or iPhone voice memo files directly. They will fail the automated validation check. I use a Zoom H5 with external windscreens for outdoor recordings and a simple USB condenser mic indoors. Budget about eighty dollars for a basic setup that meets spec if you do not already own equipment.
The Transcription and Tagging Pipeline
Once an interview is recorded, it goes through a three-stage pipeline. First is transcription, which the project does in-house using a combination of Otter.ai for the initial draft and human reviewers for accuracy. The human reviewers catch proper names, local place references, and technical terms that automated tools consistently miss. Second is metadata tagging, where each segment gets coded for themes like employment, housing, religion, family, and civic life. Third is geolocation linking, where mentions of specific addresses, businesses, or landmarks get matched to coordinates on the project map layer. The tagging system is where most people who try to contribute run into friction. The project uses a custom schema based on adapted Library of Congress subject headings, but with additions specific to West Texas regional history. Terms like "RRC pumping curtailment," "Permian Basin production decline," and "Edson steel closure" are valid tags that exist only in their controlled vocabulary. If you tag something as "oil industry" instead of the more precise RRC-related term, the content still gets accepted, but it will not surface in focused searches on regulatory history. That distinction matters if you are building a research path through the archive.
Get the Full Details
Contributing Your Own Materials
The platform accepts submissions from the public in several categories: photographs, personal documents, maps, and audio recordings. The upload portal is at sanangeloodyssey.org/contribute. You need to create an account and verify your email before anything sticks. The file size limit is five hundred megabytes per upload, and the system supports JPG, PNG, TIFF, PDF, WAV, and MP4 formats for different media types. When I submitted a collection of my father's receipts and work schedules from the late nineties, the review process took eleven days. The team contacted me twice to clarify whether the dates on some documents were clear enough to date-stamp accurately. They asked for higher resolution scans on three items where the text was borderline illegible. This is normal. The quality bar exists because the project plans to make these materials permanently available, and blurry scans do not survive that standard. Factor in roughly two weeks of turnaround whether your submission gets fully accepted, partially accepted with revision requests, or rejected entirely.
Common Problems and What to Do About Them
The biggest issue I encountered personally involved geolocation mismatches. A participant described a neighborhood bakery that had operated on South Guthry Street, but the street numbering system changed significantly after the 1997 flood remodeling project. The automatic geocoder placed the location about two blocks west of the actual site. I verified the correct position by cross-referencing the Tom Green County Appraisal District's historical GIS layers from 1995 against the current map. The workaround was to submit a correction ticket through the platform's feedback form with the appraisal district PDF attached. The mapping team updated the coordinate within four business days. They do not announce this process anywhere on the help pages, so if you run into the same problem, you know what to do. Another recurring issue is audio quality from certain recording environments. The old community centers in San Angelo have problematic HVAC systems that create low-frequency hum in recordings. I learned to record during the evening hours when the systems were off, which cut the post-production cleanup time by roughly seventy percent. The project's audio engineering guidelines mention this but do not give the specific time window. If you are doing interviews on site, plan for 6 PM to 8 PM availability to avoid the noise.
What the Platform Can and Cannot Do
The interactive map layer is functional but limited. It handles point-based locations well, but linear features like old railroad corridors and discontinued road alignments are spotty. The project never fully digitized the 1982 TxDOT right-of-way maps, so any story that depends on road geography pre-1990 will have gaps. The search function is also narrow. It indexes titles, tags, and full transcript text, but it does not support Boolean operators or proximity searching. You cannot run a query like "air conditioning AND shutdown" to find discussions of industrial cooling system failures across hundreds of interviews. This is a known limitation the team has discussed internally but has not prioritized for development given their funding constraints. The project is funded through a combination of Texas Historical Commission grants, private foundation support, and university overhead. This means it is not going away, but feature development moves slowly. New submission categories or advanced search capabilities should not be expected on any timeline. The core archive is stable and searchable, which is the main requirement for most users. If you are looking for comprehensive municipal policy analysis, you should use the Tom Green County Clerk records alongside the Odyssey materials rather than relying on the project alone. The oral histories fill gaps the official record ignores, but they do not replace it. I have found that the strongest research comes from treating the platform as one source among several, not as the definitive word on any given topic.
