This is the second in a series of four installments on Bill Mangum's presentation, "The Opening of the Digital Pipeline," from the final day of the BYU Genealogy and Family History Conference. Last week in the first chapter we presented the analogy of a pipeline and introduced the stages in the digital pipeline. This week we'll talk about handling gaps in the pipeline and we'll explain the first several processes in the pipeline. In the third chapter we'll continue down the pipeline to the Infobahn stage. In the fourth and final chapter, we'll finish relating Mangum's presentation, talking about Record Search and FamilySearch Labs.
(Credit: chair clipart)
The current pipeline has gaps where the processes are not yet complete. "We have what we call swivel processes," explained Bill Mangum. These are manual processes where someone must accept the output of the pipe at one of the gaps, manually perform a task, swivel around, and send the data on down the pipeline. "As we go, we'll fill in the gaps with automated tasks."
Current Pipeline Processes
Collection management attempts to find record collections at government agencies, archives and other locations that include most of the people in the area in the desired time frame.
Record Services works with archives and various other record custodians to gain permission to acquire copies of the desired records. They write the agreements and assure that all legal requirements are accounted for. They manage the relationships between FamilySearch and the record custodians.
Since 1938 FamilySearch and its predecessor, the Genealogical Society of Utah (GSU), have been microfilming records. In fact, GSU was involved with microfilm from its inception and drove many of the standards that are used.
Cameras capture records across the nation, and in fact, throughout the world. To see the extensive list of locations where FamilySearch (aka GSU) is capturing records, visit the GSU Website at http://www.gensocietyofutah.org/ and click on Worldwide Activities. GSU currently employs over 240 cameras. If I heard correctly, over 200 of the cameras are now digital.
Digital cameras have several additional advantages. Operators can check their work immediately and re-shoot as necessary without extra time and travel. The lights used on the copy stands are cool to the touch, which is much better for original documents. Operators store the images to 1/2 terabyte drives which they send to Salt Lake. The vision is that these images will one day flow through the pipeline so quickly, they will be available to users in a very short time.
Image Processing receives the images in Salt Lake. They receive about a terabyte of images each week. They create an archival copy of the image, de-skew any tilted images, adjust brightness and contrast if necessary, and create a publication copy for the Internet.
Scanning is also proceeding on the existing microfilm collection at the granite mountain vault. An employee of the scanning team in the audience let Mangum know that the number of scanners was now up to 8. They are working their way up to 15. While the scanners used to scan one page at a time and required an operator on each scanner checking every page, the scanners now image an entire film as one long strip and automatically divides the pages. The audience member told us they only require 2 operators to keep all the scanners active.
Next week we'll finish up scanning and move on down the pipeline.
Any indication on how many individual images on average would be in a terabyte of data? The microfilmers would send back enough data to fill 6,000 rolls a month. How many rolls of microfilm, approximately, are they getting the equivalent of in digital images?ReplyDelete
By the way, half-terabyte drives are not uncommon, we'll see one-terabyte drives commonplace very soon, they are getting close to that for drives sold at consumer computer stores.