Raspberry Pi Forums

Posted: **Wed Nov 29, 2017 3:25 am**

arnauld wrote: ↑
Tue Nov 28, 2017 4:14 pm
and edited "scripts/startup":
MOUNT_DISK=untitled

Didn't catch this on your first post, but MOUNT_DISK needs be a device name such as sda1, sdb1, or wherever
the disk appears. But you say you get videos on the drive, so I'm not sure if something else is going on.

What "mount" lines do you get in the log file?
What does ls -l on /media/pi and ls -l /media/pi/untitiled show?
Are there any files in /medai/pi/untitled/thumbs

In pikrellcam.conf try setting: media_dir /media/pi/untitled/media

Posted: **Wed Nov 29, 2017 6:24 pm**

thank you @billw, I found one of your older post:

I have tried to find a post about this but couldnt find, so heres a question:

How do i make the pikrellcam so it saves the recorded media files to USB stick or NAS ?
The media_dir is configured in ~/.pikrellcam/pikrellcam.conf and you can have an automatic mount
enabled in ~/pikrellcam/scripts/startup.

For my setups, I leave the .pikrellcam.conf media_dir at its default of "media" and I edit the startup
script so that MOUNT_DISK=sda1
I have a USB stick plugged in which appears as /dev/sda so the startup script mounts /dev/sda1
on ~/pikrellcam/media and so all my video files are stored on the USB stick. This should work
the same for a NAS. If you want to mount the NAS or USB stick somewhere else, edit media_dir
as needed.

...did the same and it is working fine now, videos and pics on my usb key and I can see the thumbnails on the webpage. Thanks again.

Posted: **Wed Nov 29, 2017 10:04 pm**

Billw

I have been using these on pi zero >

https://www.aliexpress.com/item/Overfly ... eLevelAB=]

Without any problems on V4.0.5 until i upgraded to V.4.1.1

Now it is recording, using V.4.1.1 but only for the first 30 to 50 seconds of a 3 min video

To replicate this i have rewritten an image with V4.0.5 already loaded then made a video 3 min long which had sound all the way through, then upgraded to V4.1.1 using the button on webpage,

Result was same, the mic recording stops after 30-40 seconds

If you can take a look please

Posted: **Wed Nov 29, 2017 11:20 pm**

DaveDJ wrote: ↑
Wed Nov 29, 2017 10:04 pm
Without any problems on V4.0.5 until i upgraded to V.4.1.1

Now it is recording, using V.4.1.1 but only for the first 30 to 50 seconds of a 3 min video

What do you get for space in /run? Run

Code: Select all

df -h /run

Run it before starting a record and then again after a record has run for the 30 to 50 seconds.
Also after a record has run for the 30 to 50, do a

Code: Select all

ls -l /run/pikrellcam

to check the size of the mp3 being recorded. Continue doing a ls -l every 10 seconds or so
to see if the mp3 continues to grow in size.

Posted: **Thu Nov 30, 2017 6:11 pm**

[email protected]:~$ df -h /run
Filesystem Size Used Avail Use% Mounted on
tmpfs 186M 4.5M 181M 3% /run
[email protected]:~$ df -h /run
Filesystem Size Used Avail Use% Mounted on
tmpfs 186M 4.5M 181M 3% /run
[email protected]:~$ ls -l /run/pikrellcam
total 292
-rw-r--r-- 1 pi pi 19071 Nov 30 17:53 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:53 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 500 Nov 30 17:52 motion-events
-rw-r--r-- 1 pi pi 15915 Nov 30 17:51 motion_2017-11-30_17.51.41_2.jpg
-rw-r--r-- 1 pi pi 253952 Nov 30 17:53 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 648 Nov 30 17:52 state
[email protected]:~$ ls -l /run/pikrellcam
total 368
-rw-r--r-- 1 pi pi 18918 Nov 30 17:53 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:53 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 646 Nov 30 17:53 motion-events
-rw-r--r-- 1 pi pi 16291 Nov 30 17:53 motion_2017-11-30_17.51.41_2.jpg
-rw-r--r-- 1 pi pi 331776 Nov 30 17:53 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 648 Nov 30 17:52 state
[email protected]:~$ ls -l /run/pikrellcam
total 440
-rw-r--r-- 1 pi pi 18932 Nov 30 17:53 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:53 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 646 Nov 30 17:53 motion-events
-rw-r--r-- 1 pi pi 16291 Nov 30 17:53 motion_2017-11-30_17.51.41_2.jpg
-rw-r--r-- 1 pi pi 405504 Nov 30 17:53 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 648 Nov 30 17:52 state
[email protected]:~$ ls -l /run/pikrellcam
total 524
-rw-r--r-- 1 pi pi 18993 Nov 30 17:54 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:54 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 787 Nov 30 17:54 motion-events
-rw-r--r-- 1 pi pi 16291 Nov 30 17:53 motion_2017-11-30_17.51.41_2.jpg
-rw-r--r-- 1 pi pi 491520 Nov 30 17:54 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 648 Nov 30 17:53 state
[email protected]:~$ ls -l /run/pikrellcam
total 616
-rw-r--r-- 1 pi pi 18918 Nov 30 17:54 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:54 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 787 Nov 30 17:54 motion-events
-rw-r--r-- 1 pi pi 16291 Nov 30 17:53 motion_2017-11-30_17.51.41_2.jpg
-rw-r--r-- 1 pi pi 585728 Nov 30 17:54 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 648 Nov 30 17:53 state
[email protected]:~$ ls -l /run/pikrellcam
total 616
-rw-r--r-- 1 pi pi 18931 Nov 30 17:54 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:54 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 787 Nov 30 17:54 motion-events
-rw-r--r-- 1 pi pi 16291 Nov 30 17:53 motion_2017-11-30_17.51.41_2.jpg
-rw-r--r-- 1 pi pi 585728 Nov 30 17:54 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 648 Nov 30 17:53 state
[email protected]:~$ ls -l /run/pikrellcam
total 616
-rw-r--r-- 1 pi pi 18890 Nov 30 17:55 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:55 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 787 Nov 30 17:54 motion-events
-rw-r--r-- 1 pi pi 16291 Nov 30 17:53 motion_2017-11-30_17.51.41_2.jpg
-rw-r--r-- 1 pi pi 585728 Nov 30 17:54 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 648 Nov 30 17:54 state
[email protected]:~$ ls -l /run/pikrellcam
total 604
-rw-r--r-- 1 pi pi 17948 Nov 30 17:55 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:55 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 793 Nov 30 17:55 motion-events
-rw-r--r-- 1 pi pi 588192 Nov 30 17:55 motion_2017-11-30_17.51.41_2.mp3
-rw-r--r-- 1 pi pi 646 Nov 30 17:55 state
-rw-r--r-- 1 pi pi 0 Nov 30 17:55 video_converting
[email protected]:~$ ls -l /run/pikrellcam
total 28
-rw-r--r-- 1 pi pi 18001 Nov 30 17:56 mjpeg.jpg
-rw-r--r-- 1 pi pi 0 Nov 30 17:56 mjpeg.jpg.part
-rw-r--r-- 1 pi pi 793 Nov 30 17:55 motion-events
-rw-r--r-- 1 pi pi 646 Nov 30 17:55 state

Hi Billw

I have seen the problem, when watching the playback the video freezes for a few seconds at the exact point the sound cuts out and does not restart the mic from the point it froze

Still using Jessie

Just to add i am running 8x pi zero and 4x pi zero w, every time i upgrade one to V4.1.1 the video freezes momentarily and the mic stops
It isn't limited to one pi it's all 12 of them

Regards
Dave

more info - if the usb mic is removed V4.1.1 works perfectly on Jessie/pi zero

Posted: **Fri Dec 08, 2017 2:22 am**

Billw

I have found the problem now, it isn't a bug in V4.1.1 but if you can advise how adjust to record the full mic recording up to the end in settings that would be great

the mic records until the last motion event while the camera keeps recording to the end of the 'Hold' period in the screenshot the mic records until 1.02 then the video keeps going for another 44 seconds with no sound

i can live with that

Regards
Dave

Posted: **Fri Dec 08, 2017 11:10 pm**

I have just done a PikrellCam software upgrade in one of my birdboxes which is running pikrellcam with RasPi noir. There's a second camera (usb webcam) which also has an integral microphone. With the audio upgrade I now get sound synced with video with no extra effort, which is frankly awesome...

I was wondering if it would be possible to..
1) Capture audio event to an audio file format even if there is no movement detected?
2) Run an external script 'on audio event' (I couldnt see something like that in pikrellcam.conf)

The scenario I was considering is when I get a bird / squirrel having a go at the outside of the box, I dont have any way of recording that since there's no external camera. Being able to trigger recording of an audio file, e.g. audio over a certain threshold / or 'x' times per 'y' period would be interesting. I could also see an application for logging my Tawny owls night calling to each other where they regularly shout at each other near the nesting boxes without necessarily entering the box and triggering motion via the camera.

Thanks,

Chris

Posted: **Mon Dec 11, 2017 3:43 pm**

DaveDJ wrote: ↑
Fri Dec 08, 2017 2:22 am
I have found the problem now, it isn't a bug in V4.1.1 but if you can advise how adjust to record the full mic recording up to the end in settings that would be great

What are all of your Times settings: confirm_gap, pre_capture, event_gap and post_capture?

Posted: **Mon Dec 11, 2017 3:58 pm**

Nestboxtech wrote: ↑
Fri Dec 08, 2017 11:10 pm
I was wondering if it would be possible to..
1) Capture audio event to an audio file format even if there is no movement detected?
2) Run an external script 'on audio event' (I couldnt see something like that in pikrellcam.conf)

The scenario I was considering is when I get a bird / squirrel having a go at the outside of the box, I dont have any way of recording that since there's no external camera. Being able to trigger recording of an audio file, e.g. audio over a certain threshold / or 'x' times per 'y' period would be interesting. I could also see an application for logging my Tawny owls night calling to each other where they regularly shout at each other near the nesting boxes without necessarily entering the box and triggering motion via the camera.

It's been on my todo to add an audio trigger to start a motion video and it should be one of the next things I do. In your case
for night or audio out of video view you would get uninteresting video but have the audio you want. I'm inclined to do this rather than
have separate audio file only handling. Aside from the more complicated user interface to handle separate audio files, an audio
trigger needs to start a video in case there will be subsequent motion.

The existing on_motion_begin/on_motion_end would run, but I'll have a way for the begin script to know if it was an audio
trigger and the end script to know if there were only audio trigger events.

Would that be good enough for you?

Posted: **Mon Dec 11, 2017 5:08 pm**

Would it be possible, in the event only an audio trigger starts the video, to temporarily select a lower frame rate and/or bitrate for the video recording? This would allow saving storage space in the case where audio events are less likely to result in interesting video. I'm not sure if changing frame rate on the fly like that might complicate things, if you have already recorded a video pre-buffer with different settings.

Posted: **Mon Dec 11, 2017 6:19 pm**

jbeale wrote: ↑
Mon Dec 11, 2017 5:08 pm
Would it be possible, in the event only an audio trigger starts the video, to temporarily select a lower frame rate and/or bitrate for the video recording? This would allow saving storage space in the case where audio events are less likely to result in interesting video. I'm not sure if changing frame rate on the fly like that might complicate things, if you have already recorded a video pre-buffer with different settings.

Maybe a video could play if the bitrate is changed, but I'm not sure. At least there would be a lot of places in the code that would need
to be changed. The camera is restarted when the bitrate is changed and right now the circular buffer is reinitialized and there are calculations in various places based on bitrate. This is something I would have to look at later as an experiment.

Posted: **Mon Dec 11, 2017 6:35 pm**

Ah, I'd say it's probably not worth it then, if the camera must be restarted to change the parameters. Sounds like it would be better to have an entirely separate application to deal with audio-only recording. I have been entertaining some thoughts along those lines.

Simply recording audio 24/7 is pretty easy, such as viewtopic.php?t=153520&p=1006200#p1006200

It would be fun to have the Pi look for "interesting" sounds and record those separately. I suspect that audio processing neural networks may now be capable of (for example) detecting human footsteps inside a house, as distinct from dogs, cats, HVAC, dishwasher, random creaks, and outdoor weather sounds. I'm not aware of this being done yet, but it's an intriguing possibility.

Posted: **Wed Dec 13, 2017 11:10 pm**

billw wrote: ↑
Mon Dec 11, 2017 3:58 pm

Nestboxtech wrote: ↑
Fri Dec 08, 2017 11:10 pm
I was wondering if it would be possible to..
1) Capture audio event to an audio file format even if there is no movement detected?
2) Run an external script 'on audio event' (I couldnt see something like that in pikrellcam.conf)

The scenario I was considering is when I get a bird / squirrel having a go at the outside of the box, I dont have any way of recording that since there's no external camera. Being able to trigger recording of an audio file, e.g. audio over a certain threshold / or 'x' times per 'y' period would be interesting. I could also see an application for logging my Tawny owls night calling to each other where they regularly shout at each other near the nesting boxes without necessarily entering the box and triggering motion via the camera.
It's been on my todo to add an audio trigger to start a motion video and it should be one of the next things I do. In your case
for night or audio out of video view you would get uninteresting video but have the audio you want. I'm inclined to do this rather than
have separate audio file only handling. Aside from the more complicated user interface to handle separate audio files, an audio
trigger needs to start a video in case there will be subsequent motion.

The existing on_motion_begin/on_motion_end would run, but I'll have a way for the begin script to know if it was an audio
trigger and the end script to know if there were only audio trigger events.

Would that be good enough for you?

If it would be possible to distinguish an 'audio only' vs 'audio + motion' event that would be useful. I had envisaged running an external command that logged such events to a MySql database together with some sort of ID that linked to the video file + with timestamp. Is there some means that an external script can also capture the name of the output video file at the same time?

Posted: **Wed Dec 13, 2017 11:13 pm**

hopefully a simple one this time..
Is there a way to move the text overlay form the top to the bottom of the video?
I couldn't see anything in 'Annotate Text Options '

thanks

Posted: **Wed Dec 13, 2017 11:48 pm**

Nestboxtech wrote: ↑
Wed Dec 13, 2017 11:13 pm
Is there a way to move the text overlay form the top to the bottom of the video?

The text is limited to the top by the mmal library layer because it is reusing an original camera debug feature that wasn't intended to be general purpose.

I intend that a video triggered by audio events only will be tagged in some way to show it is just audio.

Posted: **Thu Dec 14, 2017 4:01 am**

Bill

Those times are as follows

confirm-gap=4
pre-capture=5
event-gap=60
post-capture=5
motion-time-limit=0

I have upgraded all pi zero to V4.1.1 now, but when i was on V4.0.5 the sound did always record up to the end of the hold period with the same settings as above

I did try Stretch, it seems a lot slower than Jessie which has a rapid response with your software, i have rewriiten Jessie bare minimum images and no plans to upgrade

Regards
Dave

Posted: **Sat Dec 16, 2017 3:19 pm**

Hello,

I try to archive loop-video with the at-commands.conf but it does not work...

In the web interface on the loop panel I can do it by selecting the loop-video and then press archive button : it works.

The log :

16:10:00 : command_process: archive_video day yesterday
16:10:00 : execl[wait:0]: /home/pi/pikrellcam/scripts-dist/_archive-video day yesterday /home/pi/pikrellcam/media/archive /home/pi/pikrellcam/media /home/pi/pikrellcam/www/FIFO /tmp/pikrellcam.log
16:10:00 : command_process: archive_still day yesterday
16:10:00 : execl[wait:0]: /home/pi/pikrellcam/scripts-dist/_archive-still day yesterday /home/pi/pikrellcam/media/archive /home/pi/pikrellcam/media /home/pi/pikrellcam/www/FIFO /tmp/pikrellcam.log
16:10:00 : Bad command: [archive_loop] [day yesterday]
archive-still day 2017-12-15 to /home/pi/pikrellcam/media/archive/2017/12/15
archive day 2017-12-15 to /home/pi/pikrellcam/media/archive/2017/12/15

If I understand the archive_loop does not exist as command.

Is it correct ?

Thanks

Posted: **Sun Dec 17, 2017 5:07 pm**

Ouba wrote: ↑
Sat Dec 16, 2017 3:19 pm
If I understand the archive_loop does not exist as command.

Is it correct ?

I had thought only selected loop videos with motion would be worth archiving and not a full day of loop videos. But I can add it. I'll probably modify the existing archive_video script so it can archive videos from the media_dir or loop_dir and not have a separate archive_loop script.

Posted: **Sun Dec 17, 2017 6:21 pm**

Hello Bill,

For my purpose it's ok but for the others users maybe not ? ...

If you can keep the choice for what to archive : still / video / loop (but in this case you have to make an archive_loop script) ... everything is possible ....

I also want to participate to your great software but I 'm only a user. So just a feedback : found that in this topic (H264 streaming) :

viewtopic.php?f=43&t=115583&start=425#p862399

to make it work I had to modify the first line of the script start-gst-variable-rtsp-server.sh :

/root/gst-gateworks-apps-master/bin/gst-variable-rtsp-server -d 99 -p 8555 -m /stream.high \

become

/root/gst-gateworks-apps/bin/gst-variable-rtsp-server -d 99 -p 8555 -m /stream.high \

Maybe useful to modify or mention it somewhere ... ?

Sorry for my english.

Thanks again you for this great software

Posted: **Sun Dec 17, 2017 9:28 pm**

Hello @Billw & @DaveDJ

I have found the problem now, it isn't a bug in V4.1.1 but if you can advise how adjust to record the full mic recording up to the end in settings that would be great

Image

the mic records until the last motion event while the camera keeps recording to the end of the 'Hold' period in the screenshot the mic records until 1.02 then the video keeps going for another 44 seconds with no sound

The same thing is happening to me, no sound on the video when there isn't any movement.

see my video, at 23secs the sound is off...https://youtu.be/O2yCjtKNX44

Posted: **Mon Dec 18, 2017 7:08 pm**

arnauld wrote: ↑
Sun Dec 17, 2017 9:28 pm
Hello @Billw & @DaveDJ
the mic records until the last motion event while the camera keeps recording to the end of the 'Hold' period in the screenshot the mic records until 1.02 then the video keeps going for another 44 seconds with no sound
The same thing is happening to me, no sound on the video when there isn't any movement.

see my video, at 23secs the sound is off...https://youtu.be/O2yCjtKNX44

It's not supposed to record to the end of the hold period (event_gap) but only through the post_capture time. So the audio length was actually OK and the bug was the video data was too long - should be fixed now.

I got a dual mic a while back but just haven't got around to looking into the 2 channel bug. Will try to get to it soon.

Posted: **Tue Dec 19, 2017 7:21 am**

billw wrote: ↑
Mon Dec 18, 2017 7:08 pm
I got a dual mic a while back but just haven't got around to looking into the 2 channel bug. Will try to get to it soon.

No rush on that one, I'm probably the only guy who ever tried a stereo mic with it. Hope you have a great holiday season and thanks again for all the amazing work you've done on PiKrellCam!

-john

Posted: **Tue Dec 19, 2017 4:21 pm**

It's not supposed to record to the end of the hold period (event_gap) but only through the post_capture time. So the audio length was actually OK and the bug was the video data was too long - should be fixed now.

thank you.

Posted: **Wed Dec 20, 2017 10:53 pm**

billw wrote: ↑
Mon Dec 18, 2017 7:08 pm

It's not supposed to record to the end of the hold period (event_gap) but only through the post_capture time. So the audio length was actually OK and the bug was the video data was too long - should be fixed now.

Thanks for that Bill
Runs like clockwork, much appreciated

Posted: **Thu Dec 21, 2017 12:33 am**

jbeale wrote: ↑
Tue Dec 19, 2017 7:21 am

billw wrote: ↑
Mon Dec 18, 2017 7:08 pm
I got a dual mic a while back but just haven't got around to looking into the 2 channel bug. Will try to get to it soon.
No rush on that one, I'm probably the only guy who ever tried a stereo mic with it. Hope you have a great holiday season and thanks again for all the amazing work you've done on PiKrellCam!

-john

John, i have just ordered one to experiment with

https://www.aliexpress.com/item/New-Min ... 0.0.Br6Afi

and one of these>

https://www.aliexpress.com/item/3D-Hot- ... 0.0.Br6Afi

It's going in the car on a zero w

Raspberry Pi Forums

PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface

Re: PiKrellCam: motion vector detect + OSD web interface