>> Data and Metadata
Thread views: 1724 View all threadsNext thread*Threaded Mode

(Stranger )
05/11/14 12:35 PM
Conversion for unknown data Report this article as Inappropriate to us !!!Login to Reply

I have recently taken a developer's position at a utility location facility.  Though I am a programmer, I have 4 years+ as a GIS specialist back in the mid-90s.  I came from ArcInfo and MapInfo background.  With my client seeing that in my resume, I been asked to help with a long running problem.
They locate utilities in the Seattle, Wa area using OSP Viewer, an outdated program circa 1999.  It uses PIC files are its images.  PIC was made for PC Paint and never went much further.

I believe the data simply stores the pixel coordinates of the node and the OPS viewer takes you there.

Having said that, my client would simply like to have the locations in Lat. Long., which would be suitable for their purposes, since that is what they provide the locators with.

The Cross Wire Boxes each have a data entry of:


1757 249
-4768562.33943067 14987023.69146397

1758 245
-4782490.88858055 14990527.45872711

2601 11TH W
1759 248
-4772809.78108297 14992442.11638163

3209-02 17TH W
1759 247
-4774955.87643112 14994841.67435944

The first figure(6866543) is a sequential index that I can see being referenced through the program.  But it is resolved to the information below.  

The second is the street address for Seattle area(1204 QUEEN ANNE N). Only included in the Wire Center data. Not in the utility data so there is no hope in geocoding addresses. 

The 1757 249 may be Row and Col of the drawing but I am not sure. (just took on this challenge)

The last value pair looked suspicious.  I called the data providers, who were tight lipped, and he assured me they were meaningly database values inserted from Oracle!?  So what I did was charted the first 100 wire centers on an Excel scatter chart.  I then geocoded the same 100 addresses using Google Maps, by hand, saving the points on teh map.  I set the alpha of the chart to 50% (transparency) and was able to overlay an image of the scatter chart directly over the google map, matching point by point to the addresses I had saved.  So, without a doubt, these are projection (or coordinate) values.  Using Google Map's "What's Here" function, I can see that the first address is: 47.629606, -122.356529 (LatLon).

If I knew the projection, (the north/south looks nearly like a latitude), I should be able to convert these values to Lat. Long.
At this point, the only GIS tool I have is an old version of MapInfo 6 but I think I have access to ESRI suite or I can write a conversion myself.  I am dealing with several hundred thousand drawing, so Ican't do this by hand.

Does anyone recognize the projection values?






08/13/14 10:23 AM
Re: Conversion for unknown data new [re: SpringerRider]Report this article as Inappropriate to us !!!Login to Reply

I would guess that it's just a coincidence - the second number is the latitude and the first number is the longitude.

The numbers are too big for meters - feet seems in the ball park but still off.




(Stranger )
08/14/14 08:43 AM
Re: Conversion for unknown data new [re: scrabblehack]Report this article as Inappropriate to us !!!Login to Reply

Try other values for "feet". There are International Feet, and Survey Feet, just for starters; do a search for exact values and other unit possibilities.  Since this is older data, that may be what units the data are in; the slight differences in values could account for the variation in these numbers.  Good luck!

05/11/15 07:34 AM
Re: Conversion for unknown data new [re: SpringerRider]Report this article as Inappropriate to us !!!Login to Reply

I realize this is late but it's an interesting problem. 

There is only a 0.0002% difference between the International and the U.S. Survey foot. 

But it is 9868 units from 2261 Montavista to 2601 11th.  Various mapping sites show the actual distance to be in the neighborhood of 8858 ft.  The unit is about 11% smaller than a foot.  As I said this is in the ball park. 

If we expand the values by 11% and divide by 364806 (the number of feet in a degree of latitude) we get an origin somewhere in the neighborhood of 2 degrees north.  14994841 * 11% / 364806 ~= 45.6.  I suspect the true reference latitude is the equator. 

The reference longitude is 4774956 * 11% /  246578 (feet in a degree of longitude at 47.6 N) = 21.5 degrees.  That could be a reference longitude of 99W, or 102W, or maybe 100W. 

Whatever it is it is not UTM13 (ref longitude = 105 W) or UTM14 (ref longitude = 99).  It would seem you need a projection with a reference latitude at the equator and 11% linear distortion at 47.6 degrees North latitude.  There are many possibilities.

Is it possible to just geocode the addresses?

Edited by scrabblehack on 05/11/15 07:36 AM.

View all threadsNext thread*Threaded Mode
Jump to


Teledyne: - Countless CAD add-ons, plug-ins and more.

Internet Business Systems © 2017 Internet Business Systems, Inc.
595 Millich Dr., Suite 216, Campbell, CA 95008
+1 (408)-337-6870 — Contact Us, or visit our other sites:
AECCafe - Architectural Design and Engineering EDACafe - Electronic Design Automation TechJobsCafe - Technical Jobs and Resumes  MCADCafe - Mechanical Design and Engineering ShareCG - Share Computer Graphic (CG) Animation, 3D Art and 3D Models
  Privacy Policy