Recording your QSO's as an MP3 file

KG5EBI · July 8, 2019, 7:56am

I have put together a <15 minute tutorial on everything needed to record the QSO’s on your Allstar node, move them to an internet accessible web folder and convert them from large .WAV files (the compression used is not all that bad really) into manageable .MP3 files that will play back simply through your browser.

YouTube link: https://youtu.be/SnOBoF85k0k

The process requires the installation of 4 packages.

Apache2 (most likely have this installed already)
rsync (very small, lightweight and reliable)
ffmpeg to convert the audio
lame encoder to convert to mp3 codec
Installation and configuration of everything is demonstrated in the YouTube tutorial video. If you don’t want to type out the bash script it can be downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

The software mentioned above will still need installed as per video instructions. Quick overview is here:

sudo -s

apt update

apt-get install apache2

apt-get install rsync

apt-get install ffmpeg

apt-get install lame

mkdir /var/www/html/library

Enjoy and if you have any questions, please post them on the official Allstar Community forum here: https://community.allstarlink.org/

73’s

Jae KG5EBI

KV4S · July 8, 2019, 10:33pm

I watched this today and I might give it a shot.
I’m assuming you just setup a crontab schedule to kick off the script?

Also, do we need a delete script to remove them if the MP3s are older than a month or week.

73,
Russell Thomas, KV4S
DMR TG: 240218 | YSF: US KV4S | D-STAR: XRF334R
Allstar: 47923 | IRLP: 4535 | EchoLink: KV4S-L
Web: http://KV4S.com

···

On Mon, Jul 8, 2019, 2:57 AM Joshua Nulton kg5ebi@gmail.com wrote:

I have put together a <15 minute tutorial on everything needed to record the QSO’s on your Allstar node, move them to an internet accessible web folder and convert them from large .WAV files (the compression used is not all that bad really) into manageable .MP3 files that will play back simply through your browser.

YouTube link: https://youtu.be/SnOBoF85k0k

The process requires the installation of 4 packages.

Apache2 (most likely have this installed already)

rsync (very small, lightweight and reliable)

ffmpeg to convert the audio

lame encoder to convert to mp3 codec
Installation and configuration of everything is demonstrated in the YouTube tutorial video. If you don’t want to type out the bash script it can be downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

The software mentioned above will still need installed as per video instructions. Quick overview is here:

sudo -s

apt update

apt-get install apache2

apt-get install rsync

apt-get install ffmpeg

apt-get install lame

mkdir /var/www/html/library

Enjoy and if you have any questions, please post them on the official Allstar Community forum here: https://community.allstarlink.org/

73’s

Jae KG5EBI

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

KG5EBI · July 8, 2019, 10:55pm

I hope you do try it Russell, if you run into hiccups feel free to reach out. But after watching some your videos I am confident you can pull this mod off in under 10 minutes. As far as the few questions, I have it running in two different ways on 2 different servers.

Server #1 the script runs every 30 seconds by a cron job and the files are synced to Google Drive, not even hosted on the node server. Since my Drive account has 500GB of space I don’t really need to think about deleting anything.

Server #2 the script runs on command only. I run an MQTT server on the host and a simple button on my phone to send an MQTT message. When the message trips from 0 (false) to 1 (true) the bash script executes for one run.

If you are on a system that you are concerned with file space you could periodically (daily even) run a script that will purge any .mp3 file in said directory that is beyond XX days old. So in bash form it would look something like this with 21 being the day count, feel free to modify that number:

#!/bin/bash
find /var/www/html/library/ -type f -mtime +21 -name ‘*.mp3’ -execdir rm – ‘{}’ ;
done

···

On Tue, Jul 9, 2019 at 5:33 AM Russell Thomas russell@kv4s.com wrote:

I watched this today and I might give it a shot.
I’m assuming you just setup a crontab schedule to kick off the script?

Also, do we need a delete script to remove them if the MP3s are older than a month or week.

73,
Russell Thomas, KV4S
DMR TG: 240218 | YSF: US KV4S | D-STAR: XRF334R
Allstar: 47923 | IRLP: 4535 | EchoLink: KV4S-L
Web: http://KV4S.com

On Mon, Jul 8, 2019, 2:57 AM Joshua Nulton kg5ebi@gmail.com wrote:

I have put together a <15 minute tutorial on everything needed to record the QSO’s on your Allstar node, move them to an internet accessible web folder and convert them from large .WAV files (the compression used is not all that bad really) into manageable .MP3 files that will play back simply through your browser.

YouTube link: https://youtu.be/SnOBoF85k0k

The process requires the installation of 4 packages.

Apache2 (most likely have this installed already)

rsync (very small, lightweight and reliable)

ffmpeg to convert the audio

lame encoder to convert to mp3 codec
Installation and configuration of everything is demonstrated in the YouTube tutorial video. If you don’t want to type out the bash script it can be downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

The software mentioned above will still need installed as per video instructions. Quick overview is here:

sudo -s

apt update

apt-get install apache2

apt-get install rsync

apt-get install ffmpeg

apt-get install lame

mkdir /var/www/html/library

Enjoy and if you have any questions, please post them on the official Allstar Community forum here: https://community.allstarlink.org/

73’s

Jae KG5EBI

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

W9CR · July 9, 2019, 5:19pm

Installation and configuration of everything is demonstrated in the YouTube
tutorial video. If you don't want to type out the bash script it can be
downloaded here:
moveconvert.sh - Google Drive

Cool, you should add this to the wiki.

Another thing I've been thinking, since we both have google accounts, take the
files and send them to the speech to text google service. We could add in a
small plugin to the node log and reference the filename and the text. You
could grep for callsigns

Enjoy and if you have any questions, please post them on the official
Allstar Community forum here: https://community.allstarlink.org/

Well it's not really used by anyone. Lists are superior they are not subject
to the same sort of censorship as a web forum.

···

On 7/8/19 3:56 AM, Joshua Nulton wrote:
--
Bryan Fields

727-409-1194 - Voice
http://bryanfields.net

KG5EBI · July 9, 2019, 6:36pm

First let me say thank you for the response Bryan, you have given this community so much and I personally am grateful for your work.

So you are thinking of transcriptions of the audio recordings? I t would be a fun test! I like the idea. I would be curious how the tool would handle the nuances of radio. Issues such as the horrid static of RF out of range, the robotic narrow bandwidth of digital modes, the accents of world and so on. Again I think the test would be fun.

As far as the forum vs. mail list thing… I just find it difficult to monitor so many channels and maillists have modlists/banlists as I am 90% sure you probably are aware.

···

On Wed, Jul 10, 2019 at 12:19 AM Bryan Fields Bryan@bryanfields.net wrote:

On 7/8/19 3:56 AM, Joshua Nulton wrote:

Installation and configuration of everything is demonstrated in the YouTube

tutorial video. If you don’t want to type out the bash script it can be

downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

Cool, you should add this to the wiki.

Another thing I’ve been thinking, since we both have google accounts, take the

files and send them to the speech to text google service. We could add in a

small plugin to the node log and reference the filename and the text. You

could grep for callsigns

https://cloud.google.com/speech-to-text/

Enjoy and if you have any questions, please post them on the official

Allstar Community forum here: https://community.allstarlink.org/

Well it’s not really used by anyone. Lists are superior they are not subject

to the same sort of censorship as a web forum.

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

KG5EBI · July 9, 2019, 7:02pm

I uploaded a couple sample audio clips to see how quickly it could transcribe. It is nothing that could be done in real time, they took 40-90 seconds on average. The accuracy results on the first file was pretty good, I was impressed. The second file not so much haha… Take a listen. Still it would be fun, now I am looking forward to the humor honestly.

Link to Audio File 1

Resulting Transcription

Link to Audio File 2

Resulting Transcription

···

On Wed, Jul 10, 2019 at 1:36 AM Joshua Nulton kg5ebi@gmail.com wrote:

First let me say thank you for the response Bryan, you have given this community so much and I personally am grateful for your work.

So you are thinking of transcriptions of the audio recordings? I t would be a fun test! I like the idea. I would be curious how the tool would handle the nuances of radio. Issues such as the horrid static of RF out of range, the robotic narrow bandwidth of digital modes, the accents of world and so on. Again I think the test would be fun.

As far as the forum vs. mail list thing… I just find it difficult to monitor so many channels and maillists have modlists/banlists as I am 90% sure you probably are aware.

On Wed, Jul 10, 2019 at 12:19 AM Bryan Fields Bryan@bryanfields.net wrote:

On 7/8/19 3:56 AM, Joshua Nulton wrote:

Installation and configuration of everything is demonstrated in the YouTube

tutorial video. If you don’t want to type out the bash script it can be

downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

Cool, you should add this to the wiki.

Another thing I’ve been thinking, since we both have google accounts, take the

files and send them to the speech to text google service. We could add in a

small plugin to the node log and reference the filename and the text. You

could grep for callsigns

https://cloud.google.com/speech-to-text/

Enjoy and if you have any questions, please post them on the official

Allstar Community forum here: https://community.allstarlink.org/

Well it’s not really used by anyone. Lists are superior they are not subject

to the same sort of censorship as a web forum.

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

Kb9mwr · July 10, 2019, 1:39am

There was some brief discussion back in January about using artificial intelligence (AI) to control Allstar.

It was in response to a hackaday entry about a google assistant on ham radio.

···

On Tue, Jul 9, 2019 at 2:02 PM Joshua Nulton kg5ebi@gmail.com wrote:

I uploaded a couple sample audio clips to see how quickly it could transcribe. It is nothing that could be done in real time, they took 40-90 seconds on average. The accuracy results on the first file was pretty good, I was impressed. The second file not so much haha… Take a listen. Still it would be fun, now I am looking forward to the humor honestly.

Link to Audio File 1

Resulting Transcription

Link to Audio File 2

Resulting Transcription

On Wed, Jul 10, 2019 at 1:36 AM Joshua Nulton kg5ebi@gmail.com wrote:

First let me say thank you for the response Bryan, you have given this community so much and I personally am grateful for your work.

So you are thinking of transcriptions of the audio recordings? I t would be a fun test! I like the idea. I would be curious how the tool would handle the nuances of radio. Issues such as the horrid static of RF out of range, the robotic narrow bandwidth of digital modes, the accents of world and so on. Again I think the test would be fun.

As far as the forum vs. mail list thing… I just find it difficult to monitor so many channels and maillists have modlists/banlists as I am 90% sure you probably are aware.

On Wed, Jul 10, 2019 at 12:19 AM Bryan Fields Bryan@bryanfields.net wrote:

On 7/8/19 3:56 AM, Joshua Nulton wrote:

Installation and configuration of everything is demonstrated in the YouTube

tutorial video. If you don’t want to type out the bash script it can be

downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

Cool, you should add this to the wiki.

Another thing I’ve been thinking, since we both have google accounts, take the

files and send them to the speech to text google service. We could add in a

small plugin to the node log and reference the filename and the text. You

could grep for callsigns

https://cloud.google.com/speech-to-text/

Enjoy and if you have any questions, please post them on the official

Allstar Community forum here: https://community.allstarlink.org/

Well it’s not really used by anyone. Lists are superior they are not subject

to the same sort of censorship as a web forum.

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

KG5EBI · July 10, 2019, 2:03am

Thanks for sharing Steve. The idea is very cool and it would be fun to play with it more. Personally, I am not a fan of bringing the home assistants into the ham space, at least not on the air. I have used Amazon (echo) Alexa and Google Home both to link and unlink from various nodes, announce time/IP and reboot but the variables were pre-programmed with a few of my favorites, kind of like memory stations/channels. I have heard Siri on the air in the past. I do not like putting the audio on the air at all or letting people know that my node can be controlled by anyone’s voice. Remember it is speech recognition, not voice recognition.

I think Pierre Martel is working on a project similar to what you describe, using a keyword to start listening to commands on the node itself - inline with the audio live. So if configured properly right in the middle of his QSO transmission he could say “We’ve been getting a lot of rain here… Allstar Connect to node 45373… Let’s see what John’s weather is like over in Madrid”. Interesting concept, and a great goal to practice scripting, but far too insecure for my liking to be put into play on one of my public nodes. Private node? Perhaps.

···

On Wed, Jul 10, 2019 at 8:39 AM Steve L kb9mwr@gmail.com wrote:

There was some brief discussion back in January about using artificial intelligence (AI) to control Allstar.

It was in response to a hackaday entry about a google assistant on ham radio.

On Tue, Jul 9, 2019 at 2:02 PM Joshua Nulton kg5ebi@gmail.com wrote:

I uploaded a couple sample audio clips to see how quickly it could transcribe. It is nothing that could be done in real time, they took 40-90 seconds on average. The accuracy results on the first file was pretty good, I was impressed. The second file not so much haha… Take a listen. Still it would be fun, now I am looking forward to the humor honestly.

Link to Audio File 1

Resulting Transcription

Link to Audio File 2

Resulting Transcription

On Wed, Jul 10, 2019 at 1:36 AM Joshua Nulton kg5ebi@gmail.com wrote:

First let me say thank you for the response Bryan, you have given this community so much and I personally am grateful for your work.

So you are thinking of transcriptions of the audio recordings? I t would be a fun test! I like the idea. I would be curious how the tool would handle the nuances of radio. Issues such as the horrid static of RF out of range, the robotic narrow bandwidth of digital modes, the accents of world and so on. Again I think the test would be fun.

As far as the forum vs. mail list thing… I just find it difficult to monitor so many channels and maillists have modlists/banlists as I am 90% sure you probably are aware.

On Wed, Jul 10, 2019 at 12:19 AM Bryan Fields Bryan@bryanfields.net wrote:

On 7/8/19 3:56 AM, Joshua Nulton wrote:

Installation and configuration of everything is demonstrated in the YouTube

tutorial video. If you don’t want to type out the bash script it can be

downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

Cool, you should add this to the wiki.

Another thing I’ve been thinking, since we both have google accounts, take the

files and send them to the speech to text google service. We could add in a

small plugin to the node log and reference the filename and the text. You

could grep for callsigns

https://cloud.google.com/speech-to-text/

Enjoy and if you have any questions, please post them on the official

Allstar Community forum here: https://community.allstarlink.org/

Well it’s not really used by anyone. Lists are superior they are not subject

to the same sort of censorship as a web forum.

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

W9CR · July 10, 2019, 2:27am

I uploaded a couple sample audio clips to see how quickly it could
transcribe. It is nothing that could be done in real time, they took 40-90
seconds on average. The accuracy results on the first file was pretty good,
I was impressed. The second file not so much haha.... Take a listen. Still
it would be fun, now I am looking forward to the humor honestly.

Neat. I actually played around with it several months back.

Link to Audio File 1 <http://108.61.224.90/library/2019-07-05023732.mp3>

*Resulting Transcription*

Do tell! What's it transcribe to?

73's

···

On 7/9/19 3:02 PM, Joshua Nulton wrote:
--
Bryan Fields

727-409-1194 - Voice
http://bryanfields.net

Pierre_Martel · July 10, 2019, 2:44am

Yeah, I,ve been able to work with both porcupine (wake word detection) and chetah speech to text, and rhino speech to intent.

They all work well and they dont need any links to a server online like alexa or google home.

My problem is with splitting the audio from the receiver to both app-rpt and wake word and other software . Then sending the answer only to the node where the command been sent from. They answer could realy be some local play command, and a ifttt server that do the script needed when a certain command is understood… But when a node is connected to anynode every node audio could trigger a wake word trigger and do a command…

And thats what slowing me.

···

Le mar. 9 juil. 2019 à 22:03, Joshua Nulton kg5ebi@gmail.com a écrit :

Thanks for sharing Steve. The idea is very cool and it would be fun to play with it more. Personally, I am not a fan of bringing the home assistants into the ham space, at least not on the air. I have used Amazon (echo) Alexa and Google Home both to link and unlink from various nodes, announce time/IP and reboot but the variables were pre-programmed with a few of my favorites, kind of like memory stations/channels. I have heard Siri on the air in the past. I do not like putting the audio on the air at all or letting people know that my node can be controlled by anyone’s voice. Remember it is speech recognition, not voice recognition.

I think Pierre Martel is working on a project similar to what you describe, using a keyword to start listening to commands on the node itself - inline with the audio live. So if configured properly right in the middle of his QSO transmission he could say “We’ve been getting a lot of rain here… Allstar Connect to node 45373… Let’s see what John’s weather is like over in Madrid”. Interesting concept, and a great goal to practice scripting, but far too insecure for my liking to be put into play on one of my public nodes. Private node? Perhaps.

On Wed, Jul 10, 2019 at 8:39 AM Steve L kb9mwr@gmail.com wrote:

There was some brief discussion back in January about using artificial intelligence (AI) to control Allstar.

It was in response to a hackaday entry about a google assistant on ham radio.

On Tue, Jul 9, 2019 at 2:02 PM Joshua Nulton kg5ebi@gmail.com wrote:

I uploaded a couple sample audio clips to see how quickly it could transcribe. It is nothing that could be done in real time, they took 40-90 seconds on average. The accuracy results on the first file was pretty good, I was impressed. The second file not so much haha… Take a listen. Still it would be fun, now I am looking forward to the humor honestly.

Link to Audio File 1

Resulting Transcription

Link to Audio File 2

Resulting Transcription

On Wed, Jul 10, 2019 at 1:36 AM Joshua Nulton kg5ebi@gmail.com wrote:

First let me say thank you for the response Bryan, you have given this community so much and I personally am grateful for your work.

So you are thinking of transcriptions of the audio recordings? I t would be a fun test! I like the idea. I would be curious how the tool would handle the nuances of radio. Issues such as the horrid static of RF out of range, the robotic narrow bandwidth of digital modes, the accents of world and so on. Again I think the test would be fun.

As far as the forum vs. mail list thing… I just find it difficult to monitor so many channels and maillists have modlists/banlists as I am 90% sure you probably are aware.

On Wed, Jul 10, 2019 at 12:19 AM Bryan Fields Bryan@bryanfields.net wrote:

On 7/8/19 3:56 AM, Joshua Nulton wrote:

Installation and configuration of everything is demonstrated in the YouTube

tutorial video. If you don’t want to type out the bash script it can be

downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

Cool, you should add this to the wiki.

Another thing I’ve been thinking, since we both have google accounts, take the

files and send them to the speech to text google service. We could add in a

small plugin to the node log and reference the filename and the text. You

could grep for callsigns

https://cloud.google.com/speech-to-text/

Enjoy and if you have any questions, please post them on the official

Allstar Community forum here: https://community.allstarlink.org/

Well it’s not really used by anyone. Lists are superior they are not subject

to the same sort of censorship as a web forum.

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

KG5EBI · July 10, 2019, 2:52am

Sorry Bryan I didn’t realize that the images weren’t shared/posted. The first audio clip was transcribed almost perfectly. The second one asked “Is that a pinched nerve surgery where they like moving in herb underneath the balls?”… Good ole Google.

Pierre the Sox package should split the audio for you. I agree the wake word would be an issue, but maybe not if it monitors on a private node from your mic only?

···

On Tue, Jul 9, 2019, 9:27 PM Bryan Fields Bryan@bryanfields.net wrote:

On 7/9/19 3:02 PM, Joshua Nulton wrote:

I uploaded a couple sample audio clips to see how quickly it could

transcribe. It is nothing that could be done in real time, they took 40-90

seconds on average. The accuracy results on the first file was pretty good,

I was impressed. The second file not so much haha… Take a listen. Still

it would be fun, now I am looking forward to the humor honestly.

Neat. I actually played around with it several months back.

Link to Audio File 1 <http://108.61.224.90/library/2019-07-05023732.mp3>

Resulting Transcription

Do tell! What’s it transcribe to?

73’s

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

Pierre_Martel · July 10, 2019, 11:58am

Is app_rpt using sox at all? I need to find more on this.

Garanti sans virus. www.avast.com

Le mar. 9 juil. 2019 à 22:52, Joshua Nulton kg5ebi@gmail.com a écrit :

···

Sorry Bryan I didn’t realize that the images weren’t shared/posted. The first audio clip was transcribed almost perfectly. The second one asked “Is that a pinched nerve surgery where they like moving in herb underneath the balls?”… Good ole Google.

Pierre the Sox package should split the audio for you. I agree the wake word would be an issue, but maybe not if it monitors on a private node from your mic only?

On Tue, Jul 9, 2019, 9:27 PM Bryan Fields Bryan@bryanfields.net wrote:

On 7/9/19 3:02 PM, Joshua Nulton wrote:

I uploaded a couple sample audio clips to see how quickly it could

transcribe. It is nothing that could be done in real time, they took 40-90

seconds on average. The accuracy results on the first file was pretty good,

I was impressed. The second file not so much haha… Take a listen. Still

it would be fun, now I am looking forward to the humor honestly.

Neat. I actually played around with it several months back.

Link to Audio File 1 <http://108.61.224.90/library/2019-07-05023732.mp3>

Resulting Transcription

Do tell! What’s it transcribe to?

73’s

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

Luvencl8 · July 10, 2019, 5:46pm

It is usually part of any Linux package.

···

On Jul 10, 2019, at 7:58 AM, Pierre Martel petem001@gmail.com wrote:

Is app_rpt using sox at all? I need to find more on this.

Garanti sans virus. www.avast.com

Le mar. 9 juil. 2019 à 22:52, Joshua Nulton kg5ebi@gmail.com a écrit :

Sorry Bryan I didn’t realize that the images weren’t shared/posted. The first audio clip was transcribed almost perfectly. The second one asked “Is that a pinched nerve surgery where they like moving in herb underneath the balls?”… Good ole Google.

Pierre the Sox package should split the audio for you. I agree the wake word would be an issue, but maybe not if it monitors on a private node from your mic only?

On Tue, Jul 9, 2019, 9:27 PM Bryan Fields Bryan@bryanfields.net wrote:

On 7/9/19 3:02 PM, Joshua Nulton wrote:

I uploaded a couple sample audio clips to see how quickly it could

transcribe. It is nothing that could be done in real time, they took 40-90

seconds on average. The accuracy results on the first file was pretty good,

I was impressed. The second file not so much haha… Take a listen. Still

it would be fun, now I am looking forward to the humor honestly.

Neat. I actually played around with it several months back.

Link to Audio File 1 <http://108.61.224.90/library/2019-07-05023732.mp3>

Resulting Transcription

Do tell! What’s it transcribe to?

73’s

–

Bryan Fields

727-409-1194 - Voice

http://bryanfields.net

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

App_rpt-users mailing list
App_rpt-users@lists.allstarlink.org
http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”
You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

Tgundo2003 · July 10, 2019, 7:12pm

Joshua-

Would you be willing to share how you sync to google drive? I’m pretty green in linux, but I love the idea of recording this and shipping it off to a google drive.

Thanks!

Tom

W9SRV

I hope you do try it Russell, if you run into hiccups feel free to reach out. But after watching some your videos I am confident you can pull this mod off in under 10 minutes. As far as the few questions, I have it running in two different ways on 2 different servers.

Server #1 the script runs every 30 seconds by a cron job and the files are synced to Google Drive, not even hosted on the node server. Since my Drive account has 500GB of space I don’t really need to think about deleting anything.

Server #2 the script runs on command only. I run an MQTT server on the host and a simple button on my phone to send an MQTT message. When the message trips from 0 (false) to 1 (true) the bash script executes for one run.

If you are on a system that you are concerned with file space you could periodically (daily even) run a script that will purge any .mp3 file in said directory that is beyond XX days old. So in bash form it would look something like this with 21 being the day count, feel free to modify that number:

#!/bin/bash
find /var/www/html/library/ -type f -mtime +21 -name ‘*.mp3’ -execdir rm – ‘{}’ ;
done

···

On Monday, July 8, 2019, 05:55:51 PM CDT, Joshua Nulton kg5ebi@gmail.com wrote:

On Tue, Jul 9, 2019 at 5:33 AM Russell Thomas russell@kv4s.com wrote:

I watched this today and I might give it a shot.
I’m assuming you just setup a crontab schedule to kick off the script?

Also, do we need a delete script to remove them if the MP3s are older than a month or week.

73,
Russell Thomas, KV4S
DMR TG: 240218 | YSF: US KV4S | D-STAR: XRF334R
Allstar: 47923 | IRLP: 4535 | EchoLink: KV4S-L
Web: http://KV4S.com

On Mon, Jul 8, 2019, 2:57 AM Joshua Nulton kg5ebi@gmail.com wrote:

I have put together a <15 minute tutorial on everything needed to record the QSO’s on your Allstar node, move them to an internet accessible web folder and convert them from large .WAV files (the compression used is not all that bad really) into manageable .MP3 files that will play back simply through your browser.

YouTube link: https://youtu.be/SnOBoF85k0k

The process requires the installation of 4 packages.

Apache2 (most likely have this installed already)

rsync (very small, lightweight and reliable)

ffmpeg to convert the audio

lame encoder to convert to mp3 codec
Installation and configuration of everything is demonstrated in the YouTube tutorial video. If you don’t want to type out the bash script it can be downloaded here:

https://drive.google.com/open?id=1-z9UgW_L81MKl9eBKqMmDeT0xO9jl7qE

The software mentioned above will still need installed as per video instructions. Quick overview is here:

sudo -s

apt update

apt-get install apache2

apt-get install rsync

apt-get install ffmpeg

apt-get install lame

mkdir /var/www/html/library

Enjoy and if you have any questions, please post them on the official Allstar Community forum here: https://community.allstarlink.org/

73’s

Jae KG5EBI

App_rpt-users mailing list

App_rpt-users@lists.allstarlink.org

http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”

You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

App_rpt-users mailing list
App_rpt-users@lists.allstarlink.org
http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users

To unsubscribe from this list please visit http://lists.allstarlink.org/cgi-bin/mailman/listinfo/app_rpt-users and scroll down to the bottom of the page. Enter your email address and press the “Unsubscribe or edit options button”
You do not need a password to unsubscribe, you can do it via email confirmation. If you have trouble unsubscribing, please send a message to the list detailing the problem.

Buhalo · May 31, 2020, 1:20am

My script allows you to avoid losing a significant part of traffic being recorded at the moment of uploading files for transcoding and deleting sources. Situation occurs if you run the script every minute.

#!/bin/bash
cd /etc/asterisk/RECORDER/1200/
rsync $(ls -t1 | tail -n +3 | xargs) /var/www/html/recorder/ &&
rm -f $(ls -t1 | tail -n +3 | xargs)
cd /var/www/html/recorder
for i in *.WAV;
do name=echo "$i" | cut -d'.' -f1
echo “$name”
ffmpeg -i “$i” -codec:a libmp3lame -filter:a “volume=3.0” -qscale:a 2 “${name}.mp3”
done
rm -f *.WAV
sleep 1