Location Submission Not Handled Properly

Status
Not open for further replies.

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
After reading the thread about inaccurate GPS data (https://forums.radioreference.com/arkansas-radio-discussion-forum/367839-gps-inaccuracies-found.html) I submitted a correction for the WV Jail Authority - Eastern Regional Jail in Berkeley County West Virginia. It was listed as being in Jefferson County, and the GPS location was about 15 miles wrong.

The submission has been processed, but only partly correctly. There are now two entries for the jail, one under Jefferson County, which has the correct GPS coordinates, but the actual location is in Berkeley County. There is a second entry for the jail under Berkeley County, but it has wrong coordinates--a location NW of Martinsburg.

How would one go about getting this corrected?
 

mciupa

Database Admin
Moderator
Joined
Nov 5, 2002
Messages
8,301
After reading the thread about inaccurate GPS data (https://forums.radioreference.com/arkansas-radio-discussion-forum/367839-gps-inaccuracies-found.html) I submitted a correction for the WV Jail Authority - Eastern Regional Jail in Berkeley County West Virginia. It was listed as being in Jefferson County, and the GPS location was about 15 miles wrong.

The submission has been processed, but only partly correctly. There are now two entries for the jail, one under Jefferson County, which has the correct GPS coordinates, but the actual location is in Berkeley County. There is a second entry for the jail under Berkeley County, but it has wrong coordinates--a location NW of Martinsburg.

How would one go about getting this corrected?

Best way is to make a database submission. Copy this thread link for reference.
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
I submitted two entries, one to correct the Berkely County location data, and a second one to delete the Jefferson County entry. Hopefully that works.
 

mtindor

OH/WV DB Admin
Database Admin
Joined
Dec 5, 2006
Messages
10,370
Location
Carroll Co OH / EN90LN
Looks like the Berkeley Co DB has state/federal frequencies listed in it. Not sure why, but those should probably be moved under the state/fed views instead of being listed inside the county. (and that would include Eastern Regional Jail). The same freq / location shouldn't be listed in two ccounties anyway, even if it covers two counties. Transmitter is in one county, and if there needs to be a range to cover vehicle travel then that can be done. But the Jail vehicular movements I think occur on SIRN talkgroups. So I have to wonder if that is just for activity at the institution itself. If so, 16 mi is way to broad a range on it as well.

I saw you're submission. I'm a WV admin. But I'll let a WV admin who lives in WV handle it.

Mike
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
The frequency is used primarily for in-house use--guards wanting doors opened, count status, etc. The only vehicle-related use I've heard is arriving vehicles requesting a gate to be opened. 16 mile Range is a reasonable approximation of how far the signal can be heard, but 1 mile Range would probably cover the area of actual user activity.
 

mtindor

OH/WV DB Admin
Database Admin
Joined
Dec 5, 2006
Messages
10,370
Location
Carroll Co OH / EN90LN
The frequency is used primarily for in-house use--guards wanting doors opened, count status, etc. The only vehicle-related use I've heard is arriving vehicles requesting a gate to be opened. 16 mile Range is a reasonable approximation of how far the signal can be heard, but 1 mile Range would probably cover the area of actual user activity.

OK good deal. We aren't supposed to set the range to "how far it can be heard", although I'm sure admins do -- and obviously there is no definitive there. For localized institutions (example like a GM assembly plant) I usually specify 2 mi radius. If somebody wants to be hearing stuff farther away than their curent location, they should simply disable location based scanning during that time.

Thanks for the clarification of use of the frequency. Whomever worked the submissions can take that into consideration.

Mike
 

INDY72

Monitoring since 1982, using radios since 1991.
Joined
Dec 18, 2002
Messages
14,636
Location
Indianapolis, IN
Range should be set if possible to the FCC range radius on the LIC. But for better explanation look in the DB Admin HB. :)
 
D

DaveNF2G

Guest
OK good deal. We aren't supposed to set the range to "how far it can be heard", although I'm sure admins do -- and obviously there is no definitive there.

Why not? Isn't that the point?
 

UPMan

In Memoriam
Premium Subscriber
Joined
Apr 19, 2004
Messages
13,296
Location
Arlington, TX
Why not? Isn't that the point?

No. The range sets how far the signal is relevant, not receivable (which would vary by receiver and receiver antenna, anyway) and should only cover the entity the agency serves.
 
D

DaveNF2G

Guest
That seems like an irrelevant presumption for scanner use. I don't really care that the agency using a frequency can operate within one square block. I want to know what frequencies can be monitored from a given city or county. (I know that the FCC ranges are often sufficient for that purpose, but a very tiny range should be evaluated for reality.)
 

UPMan

In Memoriam
Premium Subscriber
Joined
Apr 19, 2004
Messages
13,296
Location
Arlington, TX
In that case, you would set your selection range to a higher value. If you set the entity's range in RRDB very large (because you think it can be received from farther out) there is no way to prevent selection of that entity based on range.

If you are not using a location-based scanner (LBS), it is irrelevant, as the range and location isn't used anyway. If you are using a Whistler, it is irrelevant, as they just select all systems in the county you are in (even if you are much closer to systems in another county or are very far away from systems in the same county).

So, this really matters to Uniden models. See How it Works: Location, Location, Location

Selection depends on the overlapping of two circles. The circle defined in RRDB and the circle defined by the location/range you set in the scanner. If the circles overlap at all, the departments (subcategories in RRDB speak) and sites are selected. If you set the scanner's range to 0, then only departments/sites whose RRDB circles overlap your precise location are used. So, if you only want to hear the city you are in, you set to 0. If a 5-mile diameter's city's range was set to 200 mi in RRDB because someone lives on top of a mountain and has an antenna on a mast pointed toward that city and therefore can receive it, then anyone using LBS within 200 or more miles will get that city in their scan list, even though it is a) irrelevant to where they are and b) likely not receivable by them.
 

wa8pyr

Technischer Guru
Staff member
Lead Database Admin
Joined
Sep 22, 2002
Messages
6,983
Location
Ohio
In that case, you would set your selection range to a higher value. If you set the entity's range in RRDB very large (because you think it can be received from farther out) there is no way to prevent selection of that entity based on range.

If you are not using a location-based scanner (LBS), it is irrelevant, as the range and location isn't used anyway. If you are using a Whistler, it is irrelevant, as they just select all systems in the county you are in (even if you are much closer to systems in another county or are very far away from systems in the same county).

So, this really matters to Uniden models. See How it Works: Location, Location, Location

Selection depends on the overlapping of two circles. The circle defined in RRDB and the circle defined by the location/range you set in the scanner. If the circles overlap at all, the departments (subcategories in RRDB speak) and sites are selected. If you set the scanner's range to 0, then only departments/sites whose RRDB circles overlap your precise location are used. So, if you only want to hear the city you are in, you set to 0. If a 5-mile diameter's city's range was set to 200 mi in RRDB because someone lives on top of a mountain and has an antenna on a mast pointed toward that city and therefore can receive it, then anyone using LBS within 200 or more miles will get that city in their scan list, even though it is a) irrelevant to where they are and b) likely not receivable by them.

I can't improve on that explanation, except to add that the Handbook is pretty clear on this (Section 6.6, specifically); range is to be set to the jurisdictional area of the entity in question, not how far away it can be heard. Nor should the FCC data be used unless it happens to match the proper coordinates and range exactly (highly unlikely).

And to forestall the inevitable "manufacturer-specific" argument someone is bound to raise, location data is not manufacturer specific. We provide the location data and how manufacturers use it in their radios is up to them; Uniden has their way, and Whistler has a different way.

For any other ways you'll have to buy Cincinnati Chili (yum) and dump it on top of spaghetti. :D
 
Last edited:

UPMan

In Memoriam
Premium Subscriber
Joined
Apr 19, 2004
Messages
13,296
Location
Arlington, TX
The exceptions to this are site location/range on a multisite system and on a simulcast site. For a multisite system, the site's intended coverage range should be used (but should not fall outside the jurisdiction's boundary). And, for a simulcast "site" that consists of multiple repeater sites acting as one, the intended coverage of all repeaters that make up the site should be used (in that case, the lat/lon used is unlikely to correspond to the physical location of any of the repeaters).
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
I think the strongest argument in favor of what Upman & Tom are saying is that "receivable range" is impossible to define unless you also define:

• A "standard antenna" of some kind (what type? what elevation above ground level? what gain factor?)

• A dBm threshold that constitutes "receivable" (which obviously varies from receiver to receiver)

• A standardized methodology for factoring in the effects of atmospheric conditions, weather, terrain, vegetation, and buildings on RF propagation.

One quickly gets into "nailing jelly to a waterfall" territory trying to deal with all this.
 

UPMan

In Memoriam
Premium Subscriber
Joined
Apr 19, 2004
Messages
13,296
Location
Arlington, TX
Also, you can always increase your range setting to select more distant systems. But, if you set the range to 0 and it is still selecting distant systems, there isn't anything you can do about it (other than a lot of avoids). Overselection is already the biggest issue related to LBS (due to the use of circles, which imperfectly model most geotagged elements).
 

jonwienke

More Info Coming Soon!
Joined
Jul 18, 2014
Messages
13,416
Location
VA
You could get a lot closer if n-sided irregular polygons were allowed, rather than just circles, even if you had a limit on the number of vertices (say 10).
 

UPMan

In Memoriam
Premium Subscriber
Joined
Apr 19, 2004
Messages
13,296
Location
Arlington, TX
Calculations become very intense for irregular shapes. Scanner CPU is not enough (with everything else it is doing). However, we do allow for any number of rectangles, which is trivial to calculate and much better than circles.
 
Status
Not open for further replies.
Top