Experiment: Build Your Own Stream

Status
Not open for further replies.

johnmoe1

Member
Joined
Aug 5, 2012
Messages
369
Location
Dakota County, MN
John,

Any thought of deploying clients to other RR members who are in different areas with it talking back to your server/setup?

Yes, but it is not an easy download an app and click go kind of thing.


My plan right now is to soon have 3 locations.
Dakota County(now): Anoka, Carver-Scott, Dakota, Goodhue, Hennepin East, Hennepin West*, Minneapolis N/S, Minneapolis City Center*, Ramsey, Rice* and Washington.

SE Rochester (this week?): Elba?, Elkton?, Goodhue, Olmsted, Troy-Utica and Wabasha [I don't understand why I am not seeing any good Dodge County reception and will look into this next time I am in Rochester.]

Winona (next week?): Alma, Garvin, La Crescent, Rolling Stone and Wilson

* Less than perfect reception.



Current setup:
Gateway Desktop 3.10GHz Quad Core 1TB |DX4885-UR2F
Gateway Desktop 3 10GHz Quad Core 1TB DX4885 UR2F | eBay

bladeRF x40
https://www.nuand.com/blog/shop/
 

sfd119

Member
Joined
Jul 18, 2009
Messages
1,763
Yes, but it is not an easy download an app and click go kind of thing.

Never guessed that it would be. But if people desperately want it, perhaps you could configure it and provide the means for it to talk back to your setup, and they pay for the equipment required to host it.
 

anthonymodrow

Member
Joined
Jul 4, 2013
Messages
80
Location
Hopkins,MN
Yes, but it is not an easy download an app and click go kind of thing.


My plan right now is to soon have 3 locations.
Dakota County(now): Anoka, Carver-Scott, Dakota, Goodhue, Hennepin East, Hennepin West*, Minneapolis N/S, Minneapolis City Center*, Ramsey, Rice* and Washington.

SE Rochester (this week?): Elba?, Elkton?, Goodhue, Olmsted, Troy-Utica and Wabasha [I don't understand why I am not seeing any good Dodge County reception and will look into this next time I am in Rochester.]

Winona (next week?): Alma, Garvin, La Crescent, Rolling Stone and Wilson

* Less than perfect reception.



Current setup:
Gateway Desktop 3.10GHz Quad Core 1TB |DX4885-UR2F
Gateway Desktop 3 10GHz Quad Core 1TB DX4885 UR2F | eBay

bladeRF x40
https://www.nuand.com/blog/shop/
What are you running on software side?

Sent from T-Mobile Note 4 via Tapatalk
 

ryolsen8

Active Member
Feed Provider
Joined
May 1, 2011
Messages
510
Location
BEE07
iPad

Also is there a way so I can lock my iPad but not have the Audio stop?

My audio doesn't stop on my iPad, when I lock it... However, if you're using Safari, it will. If you are using Google Chrome, it doesn't. ;-)
 

scanmn

Member
Joined
Oct 18, 2015
Messages
7
Any chance you could add some of the TAC channels?

L TAC 1 - 504
L TAC 2 - 506
L TAC 3 - 24104
L TAC 4 - 24106
ME TAC 1 - 32
ME TAC 2 - 34
ME TAC 3 - 36
ME TAC 4 - 38

Also, if you've got the time, could you add some of the Hennepin County cities' car-to-car channels? Maybe under a separate section?

Thanks
 

sobe4u8k

Newbie
Joined
Dec 12, 2015
Messages
2
Location
Champlin, MN
Hey John, I'd also like to extend my thanks for this wonderful site. I would love to be able to listen to some metro plow trucks during storms this winter. Would it be possible to add that to the site?
 

johnmoe1

Member
Joined
Aug 5, 2012
Messages
369
Location
Dakota County, MN
That's awesome. When we want channels added does it put a stress on your setup as far as overwhelming on performance

Short answer: probably.

Long answer:
The limits are more related to the amount of audio than the number of talkgroups. A mostly silent talkgroup doesn't cause any problems except for a cluttered list.

There are a few of parts to this system.
1. The part that listens to 10 MHz of radio spectrum, pulls out the individual channels, decodes control channels and writes the raw audio to disk. This part doesn't care which talkgroup anyone is interested in, it saves everything.

2. The part that watches for audio from a talkgroup on the list, decodes audio, makes an mp3 and uploads the audio to a server at Amazon. This part, right now, is able to handle about 8(?) seconds of audio per second. So if more than 8 talkgroups are active for long, it will start falling behind. I don't have good monitoring in place here, so I am not exactly sure how close to the limit I am.

3. The part that streams the metadata to the browser. It is not aware of what talkgroups you are listening to and sends everything (you only download audio for things you are listening to). At some point, this would probably become a problem.


There are tons of improvements that could be made to all of this. Examples:
* I tried to measure how much of the audio is ever listened to. It looked like about 20%. Not decoding/mp3 encoding until the first person requests it.
* Not making mp3s at all. Decode it all in the browser using javascript - probably more javascript than I feel like writing, but would result in better audio quality and less bandwidth used.
 

anthonymodrow

Member
Joined
Jul 4, 2013
Messages
80
Location
Hopkins,MN
Short answer: probably.

Long answer:
The limits are more related to the amount of audio than the number of talkgroups. A mostly silent talkgroup doesn't cause any problems except for a cluttered list.

There are a few of parts to this system.
1. The part that listens to 10 MHz of radio spectrum, pulls out the individual channels, decodes control channels and writes the raw audio to disk. This part doesn't care which talkgroup anyone is interested in, it saves everything.

2. The part that watches for audio from a talkgroup on the list, decodes audio, makes an mp3 and uploads the audio to a server at Amazon. This part, right now, is able to handle about 8(?) seconds of audio per second. So if more than 8 talkgroups are active for long, it will start falling behind. I don't have good monitoring in place here, so I am not exactly sure how close to the limit I am.

3. The part that streams the metadata to the browser. It is not aware of what talkgroups you are listening to and sends everything (you only download audio for things you are listening to). At some point, this would probably become a problem.


There are tons of improvements that could be made to all of this. Examples:
* I tried to measure how much of the audio is ever listened to. It looked like about 20%. Not decoding/mp3 encoding until the first person requests it.
* Not making mp3s at all. Decode it all in the browser using javascript - probably more javascript than I feel like writing, but would result in better audio quality and less bandwidth used.
Adding bus traffic mtc first transit and transit team shouldn't be to bad. The light rail talk groups seam to become over wellming

Sent from T-Mobile Note 4 via Tapatalk
 
Status
Not open for further replies.
Top