20250731

It conjures up different ideas, that. Burial in the ground, keeping things from floating away or spinning off, overwhelming size, serious business, orbit-shaping, force of attraction, pull, bending light, intensity of character, aura, fundamental.

These notions came up in a conversation about gravity that I was a party to. A question that was posed at the end and never answered was how to use or create gravity in music? That question has been on my mind a bit lately. I came up with some basic ideas about how one might incorporate gravitational concepts in music. Technically, the circling around of a sound in the auditory space would function as though you are being orbited by a sonic element. If recording, you would want to use the correct microphone and could walk around or spin the microphone. It could be accomplished with panning and frequency adustments over time. Deep and bulbous bass sounds might invoke a sense of overwhelming size. A series of chords that do not resolve create tension perhaps that of being suspended between two entities.

Skill sharing can take several forms. Here is a story of a problem someone else had, their attmpted solution, and my addressing using one better tool for the job.

A tenant union was trying to use AI and google street view to identify smaller multi-family dwellings to produce better results for their organization. This would take time and money without guarantees of success. Money was short, but they were invested in pursuing this. Was there training based on images of known smaller multi-family dwellings in the area that could somewhat create a match? This approach would also require verification, sendng someone out to determine if yes indeed there was a match. I did not get the particulars of the AI process. Didn’t really want or need them. I instantly knew that I knew there was a better way to go about this effort. I replied to them that I could look at their issue after work and that I thought it would be a simple, and much more straightforward process to get them highly relevant and accurate information. Land records are public information. The county you live in keeps the history of parcels of land (real estate) in large volumes or deed books that are kept by the tax assessor’s office. It’s very likely that the current data has now been digitized, spatially referenced, and can be loaded into Geographic Information System (GIS) software. This data is typically provided as a downloadable file or a web service that can be added to a project via a GIS server. I should probably provide a brief overview of spatial data here. There are two main data types: rasters and vectors. Rasters are data containers that are pixels. Each pixel contains a value. A pixel can be very large for datasets that cover a large area (earth), or very small for datasets that cover a smaller area (city). Vector data are represented by geometry: points, lines, and polygons. The cool thing about spatial data (to me) are that they have a spatially referenced geometry (it will appear in the correct place on a model of the earth if it has the correct reference information) and they have a table that contains information for each piece of geometry. The dataset from the tax office contains polygons that represent land parcels in the real world, and the table contains all of the relevant information that the tax office deems necessary. My day to day work involves GIS in an enterprise environment. The license cost for this is astronomical, and even for a single user license it’s not terribly cost friendly. Plus the last time I had one, it took days to activate the license. I decided to forgo the paid route. Thankfully, there is a free, open source, and very capable version of GIS software called QGIS. I downloaded and installed the latest version so I could familiarize myself enough with the layout to help out. The basic gist of what I needed to do was get the data via connecting to a GIS server and importing into a project file; make a local copy (export); and run a SQL query to pull out the correct information. I found a couple of guides online for referencing where the buttons or menus are that I needed or what processes to follow. This took about 20 minutes. I connected to the county tax office GIS service and brought the parcel layer into the project. I opened up the table and tried to locate a field that contained a description of what the land-use is. Often land-use determines tax rate, so identifying this in the data can make the process much smoother. The landuse field was encoded by numbers - 022, 023, 034, 041, etc. This made it impossible to understand without a guide. After looking online through the county website, I stumbled across an older document containing a partial list of values. This was enough for a start in building out the query, but I knew I’d need to obtain the full list. I contacted the tax office and the GIS people had left for the day. I tried back thenext week and after some mild annoyances from the staff, I was able to get ahold of the GIS person there and make a request for a document that listed all the codes and their english language equivalents. This was the key to ‘unlocking’ the data. I wrote up a guide detailing the process, including screenshots, some other useful tidbits, and handed it off to the tenant union contact.