Locations where additional Ka segments (ie. segment 6) are needed

NorEaster18

"No more government subsidies for active basses."
Advanced User
Joined
Sep 5, 2018
Messages
1,970
Reaction score
5,160
Location
The North Shore
Definitely, a late model marked patrol sedan has this unit. Have seen it again on my commute. DFR7 reads 36.11. I have to try to catch it on my R3. Could it be that far out of tune? Likely some artifact confusing the DFR7 or dual antennas causing a harmonic. This DFR7 does read correct on other Ka signals, so not sure about this crazy one. It shouldn't even be scanning up there.
Not to take the report lightly, but see if you can confirm it with your R3. I never found the frequency display on the DFRs to be accurate. My R3 consistently reports my K band gun around 24.107 and my DFR6 reports the same gun as anything between 24.10 and 24.16. I will be checking the same with my 35.5 Decatur in the next week.
 

sdrawkcaB

Defender of Light 📶
ModSec
Premium Plus
Lifetime Premium
Advanced User
Joined
Mar 6, 2018
Messages
6,571
Reaction score
14,206
Location
Here

Too3zln

PSL +5
Premium Plus
Intermediate User
Joined
Jun 27, 2019
Messages
292
Reaction score
434
Location
Nebraska
Makes a guy not want to segment anymore. Last week changed to 2/5/8 because I haven't seen too many stray too far and low and behold I ran into a 34.798. Too close for comfort. Back to 24568.
 

DrHow

Going “Plaid” (Tesla S). RDT refugee
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
May 18, 2018
Messages
3,896
Reaction score
6,580
Last edited:

Bloovy One

"Today didn't have to end in eels"
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Feb 8, 2016
Messages
7,152
Reaction score
16,073
Location
Western NC
@Bloovy One Was this changed with FW v4.1019?

It has. You can now manually adjust 33.8 like the others as of v4.1019. Some might want 33.650 to 33.950 for example. :)

When typing in "segments", you no longer need to go with the Gen1 segment boundaries either.
You can now set 34.7 to 34.600 to 34.800 instead of the Gen1 oscillator boundaries of 34.593 to 34.803, etc.
The defaults in JBV1 are for the Gen1 units.
 
Last edited:

Transporter

ModWight Transporter
Premium Plus
Lifetime Premium
Advanced User
Joined
Jul 3, 2018
Messages
4,266
Reaction score
7,252
Location
In front of you but behind a Rabbit - Hot'Lanta,GA
I might have see that this week. G2 can’t segment the “33.8” area. Thought I had alert down lower in 33.x area south of Dalton, north of I285. I could be wrong on exact location. I thought it was odd.


Yes. The two counties on I-75 South out of Chattanooga south of Dalton have County SDs using some way off-center 33.8 equipment. If one is running the short Segment List, one can get a ticket because ones RD will NOT Alert as these guns are below the Segment 2 lower boundary.

.
 

DrHow

Going “Plaid” (Tesla S). RDT refugee
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
May 18, 2018
Messages
3,896
Reaction score
6,580
Yes. The two counties on I-75 South out of Chattanooga south of Dalton have County SDs using some way off-center 33.8 equipment. If one is running the short Segment List, one can get a ticket because ones RD will NOT Alert as these guns are below the Segment 2 lower boundary.

.
excellent report! Thanks for both of you verifying this mess. Exactly where it was. will go back on JBV1 logs. Might have it there. I did not ignore. Had I been running the normal 2,4-5-6,8ish on other devices (Indiana has off freq Stalkers), would have never seen the alert. Since these GA Radar are so far off 2, looks like on purpose? They must have seen RDF member posts on how to retune radar. Ha ha.


Side note (supposition).. Unless Theia could auto spot the “signature“ of that radar, it would not alert. Yes, Theia does not alert based on frequency. Wonder when messing around with radar like this would have unique or different look. Since Theia device has NO AI built in (does not self learn and look at what is different from what it has been trained by the mother ship), modifications that could cause a different “hardware signature“ could fool Rai until mothership sees in national reporting a new pattern?
 

OBeerWANKenobi

This is not the car you're looking for......
ModSec
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Mar 20, 2018
Messages
7,100
Reaction score
20,650
Location
Outer Rim - Hiding from 35.5 I/O
Side note (supposition).. Unless Theia could auto spot the “signature“ of that radar, it would not alert. Yes, Theia does not alert based on frequency. Wonder when messing around with radar like this would have unique or different look. Since Theia device has NO AI built in (does not self learn and look at what is different from what it has been trained by the mother ship), modifications that could cause a different “hardware signature“ could fool Rai until mothership sees in national reporting a new pattern?


I think the answer is based on how RAI is trained. If, for instance, RAI was only supplied "images" of guns within a narrow frequency range and if frequency is weighed highly in it's inferencing, it could fail to identify guns falling outside that range. However, if RAI was trained on other, more repeatable identifying factors of the signal, then an out of tune gun will still be identified correctly with high probability.

There's a video or interview somewhere out there where Radenso mentions that they had problems after training where Theia would only identify a specific serial number gun, ignoring others of the same brand and type. This was due to it learning characteristics of that particular gun and weighing them highly.

Could components changing affect Theia's ability to identify a gun? I think the answer is "it depends". The component has to change something in the signal that Theia can "see" and Theia has to be weighing that aspect of the signal highly as an identifying factor.

After training and reverse propagation an AI,
can sometimes get the right answer but for the wrong reason. How you train it to mitigate that and to get the right answer for the right reason is a major factor in the final accuracy.
 

DrHow

Going “Plaid” (Tesla S). RDT refugee
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
May 18, 2018
Messages
3,896
Reaction score
6,580
I think the answer is based on how RAI is trained. If, for instance, RAI was only supplied "images" of guns within a narrow frequency range and if frequency is weighed highly in it's inferencing, it could fail to identify guns falling outside that range. However, if RAI was trained on other, more repeatable identifying factors of the signal, then an out of tune gun will still be identified correctly with high probability.

There's a video or interview somewhere out there where Radenso mentions that they had problems after training where Theia would only identify a specific serial number gun, ignoring others of the same brand and type. This was due to it learning characteristics of that particular gun and weighing them highly.

Could components changing affect Theia's ability to identify a gun? I think the answer is "it depends". The component has to change something in the signal that Theia can "see" and Theia has to be weighing that aspect of the signal highly as an identifying factor.

After training and reverse propagation an AI,
can sometimes get the right answer but for the wrong reason. How you train it to mitigate that and to get the right answer for the right reason is a major factor in the final accuracy.
Yep... Rai tech as part of Theia at the announcement/SEMA demo said it is not frequency dependent. Generally it is signature dependent (not AI, or self learning). I saw that conversation about temporary issues with Rai tech to specific gun, but not others in same brand/class. Do not know if this is totally solved.

If i were Radenso, I would ask all RDF tinkerers who have modified radar units to help R test them. Plus, work in anticipation of the cat/mouse game. Be ready for radar makers to foil the signature IDs. Where there would be constant fighting back and forth (like VPR Dragon Eye does) with how the radar deals with speed sensing. Because the device itself is not doing the true AI, I believe this is one way the long term game will be played. Once Theia gets a back log of orders, gets popular, certain radar makers will respond. Or new ones (like jack boot in GA started with the laser jamming game) might pop up.

I do not care if the battle rages on. So long at falsing/BSM is solved with Rai tech once and for all. With no issues with this popping up due to new radar tech appearing.
 

Transporter

ModWight Transporter
Premium Plus
Lifetime Premium
Advanced User
Joined
Jul 3, 2018
Messages
4,266
Reaction score
7,252
Location
In front of you but behind a Rabbit - Hot'Lanta,GA
Side note (supposition).. Unless Theia could auto spot the “signature“ of that radar, it would not alert. Yes, Theia does not alert based on frequency. Wonder when messing around with radar like this would have unique or different look. Since Theia device has NO AI built in (does not self learn and look at what is different from what it has been trained by the mother ship), modifications that could cause a different “hardware signature“ could fool Rai until mothership sees in national reporting a new pattern?


IIRC from postings or one of the videos (not going looking for it now), Theia actually has a wider scan range than current RDs. The reason, to catch way Off Center Guns and because it will not false like frequency based RDs because it is specifically looking for actual Radar Signals so opening up the band ranges slightly does not hurt performance or cause additional false alerts. Remember Theia is actual looking at Signals, not frequency. So if it is a Police Radar signal and it falls within the bandpass of the receiver, Theia will Alert no matter the Frequency especially if said signal has a Police Radar signature (and no matter how far off center frequency a Police Radar is, its signal still looks like Police Radar). In my personal opinion, only a fool would Segment the Theia if it will be possible to do that. There is absolutely no reason whatsoever to Segment Theia! There are no Ka Band False Alerts to filter out via Segmenting because Theia only alerts to Police Radar! Its superior sensitivity and what I believe will be a new standard in so called extreme range will also rule out any need to Segment.

Has anybody else noticed that Theia anticipation is leaking into just about every Thread?

.
 

DrHow

Going “Plaid” (Tesla S). RDT refugee
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
May 18, 2018
Messages
3,896
Reaction score
6,580
IIRC from postings or one of the videos (not going looking for it now), Theia actually has a wider scan range than current RDs. The reason, to catch way Off Center Guns and because it will not false like frequency based RDs because it is specifically looking for actual Radar Signals so opening up the band ranges slightly does not hurt performance or cause additional false alerts. Remember Theia is actual looking at Signals, not frequency. So if it is a Police Radar signal and it falls within the bandpass of the receiver, Theia will Alert no matter the Frequency especially if said signal has a Police Radar signature (and no matter how far off center frequency a Police Radar is, its signal still looks like Police Radar). In my personal opinion, only a fool would Segment the Theia if it will be possible to do that. There is absolutely no reason whatsoever to Segment Theia! There are no Ka Band False Alerts to filter out via Segmenting because Theia only alerts to Police Radar! Its superior sensitivity and what I believe will be a new standard in so called extreme range will also rule out any need to Segment.

Has anybody else noticed that Theia anticipation is leaking into just about every Thread?

.
Yea, I stipulated that Theia is not a frequency based detector. It is a signature based detector in a general freq range (programmable as needed for future), uber fast processing of wide swath of RF. So, if the same signature for say a 33.8 FCC approved like device/maker shows up anyplace Theia is watching, it could pick it up. That was not my point.

well, unicorns always work themselves into all kinds of threads. So does the term “two weeks”. :)
 

Monus Brewer

Learning to Drive
General User
Joined
Dec 10, 2018
Messages
27
Reaction score
13
They're not intended to be accessible to all. They're simply a few example threads that have come up recently on the subject. Any nuggets of information about frequencies and locations that would be good to post in this thread will be copied over here as well so you won't miss anything.

@hiddencam @KASHER1979 , I know you guys have seen some stuff lately as well. Could you let me know the frequencies and locations?
I kinda hate that, I am more of a page stalker than poster. 😭 Oh well maybe someday I can reach the "Golden Circle". 😂 Stay safe guys!
Post automatically merged:

Segmenting older detectors actually helped with picking up signals faster because of a reduction scan time. With modern detectors, there isn't a noticeable performance benefit to segmenting. There is a false alert reduction benefit though so if you will never see real police radar in a particular segment, turning that segment off will keep anything else in that range from alerting.

If you want to be extra safe, 2,4,5,6,8 removes segment 1 falses and is pretty safe for the entire U.S. except for NY, NJ and a Unicorn in KY where some police guns have drifted down into segment 1.
Thank you! I might try this set up and see what I get. (If I can remember 😂)

Have a great day!
Post automatically merged:

Yes. The two counties on I-75 South out of Chattanooga south of Dalton have County SDs using some way off-center 33.8 equipment. If one is running the short Segment List, one can get a ticket because ones RD will NOT Alert as these guns are below the Segment 2 lower boundary.

.
So turn on Seg 1 also?
 
Last edited:

Bloovy One

"Today didn't have to end in eels"
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Feb 8, 2016
Messages
7,152
Reaction score
16,073
Location
Western NC
Time to recommend 34.400 to 35.000....
 

Choose Life

35.5 hater R3 Day One User
Premium Plus
Lifetime Premium
Advanced User
Joined
Jan 14, 2017
Messages
8,407
Reaction score
10,234
Location
Charlotte North Carolina
Time to recommend 34.400 to 35.000....

Theres no requirement for calibration in New York state

@Vortex
NC has went to Stalkers and Segment 6 could be a good bet

All of Tennessee stalkers have been 34.7 or better
Picking up 34.778 already in a small area over just a few weeks
Segment 6 a necessity
 

Bloovy One

"Today didn't have to end in eels"
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Feb 8, 2016
Messages
7,152
Reaction score
16,073
Location
Western NC
You misunderstand me. :)

Some people here "segment" to avoid falses in between. 35.0 to 35.4 and 33.9 to 34.4 for example.
 

Brainstorm69

TXCTG - 2016 MOTY
Premium Plus
Lifetime Premium
Advanced User
Joined
May 23, 2015
Messages
11,480
Reaction score
28,500
Location
Lone Star State

Vortex

Making Videos
Observer
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Jul 19, 2012
Messages
21,141
Reaction score
54,622
Location
Washington State
Theres no requirement for calibration in New York state

@Vortex
NC has went to Stalkers and Segment 6 could be a good bet

All of Tennessee stalkers have been 34.7 or better
Picking up 34.778 already in a small area over just a few weeks
Segment 6 a necessity
Great, added. Thanks. :)

I really miss the RDFGS where people could go in and update the database directly. That's way better than having to have someone be a middleman, keeping the resource updated periodically.
 
Last edited:

OBeerWANKenobi

This is not the car you're looking for......
ModSec
Premium Plus
Lifetime Premium
Corgi Lovers
Advanced User
Joined
Mar 20, 2018
Messages
7,100
Reaction score
20,650
Location
Outer Rim - Hiding from 35.5 I/O
Great, added. Thanks. :)

I really miss the RDFGS where people could go in and update the database directly. That's way better than having to have someone like a middleman, keeping the resource updated periodically.
Yeah, I've been thinking abut RDFGS lately as well. I'd love that to be back online and editable again. It's really useful and departments are changing all the time. A local department to me always ran C/O 35.5 but now they've started using Falcon HRs. It would be nice to give RDF peeps a heads up.
 

Discord Server

Latest threads

Latest posts

Forum statistics

Threads
90,802
Messages
1,381,878
Members
22,838
Latest member
akoe
Top