Send Motion-user mailing list submissions to
        motion-user@lists.sourceforge.net

To subscribe or unsubscribe via the World Wide Web, visit
        https://lists.sourceforge.net/lists/listinfo/motion-user
or, via email, send a message with subject or body 'help' to
        motion-user-requ...@lists.sourceforge.net

You can reach the person managing the list at
        motion-user-ow...@lists.sourceforge.net

When replying, please edit your Subject line so it is more specific
than "Re: Contents of Motion-user digest..."


Today's Topics:

   1. Re: Missing frames in movies (Jack Christensen)


----------------------------------------------------------------------

Message: 1
Date: Tue, 11 Feb 2020 19:22:43 -0500
From: Jack Christensen <christensen.jac...@gmail.com>
To: motion-user@lists.sourceforge.net
Subject: Re: [Motion-user] Missing frames in movies
Message-ID: <62173f46-ee4a-e1cd-2d52-e19eea6c8...@gmail.com>
Content-Type: text/plain; charset="utf-8"

Thanks for the suggestion. I ran hdparm on four Zero Ws, a Zero, and a
3B. All give very consistent numbers, 22MB/sec, within a couple tenths.
Maybe not surprising, I think they all have similar SD cards, from Micro
Center, so maybe more or less generic types.

A typical result:
Timing buffered disk reads:? 68 MB in? 3.09 seconds =? 22.02 MB/sec

Next I burned a new copy of Raspbian Lite onto a SanDisk Ultra SDHC
UHS-1 card, which is advertised as having a read speed of up to 80MB/s.
The result on a Pi Zero W was practically the same:
Timing buffered disk reads:? 68 MB in? 3.07 seconds =? 22.15 MB/sec

But hdparm is testing read speed; if the concern is more with write
speed, then I'm not sure what to think.

As for overclocking, it doesn't appear to be an option on the Pi Zeros,
at least not via the raspi-config utility.

On 2/11/20 10:57 AM, Joris wrote:
> Guys,
>
> Actually the more I think about this, the more it sounds like a
> storage overrun where the disk is too slow to store all the frames.
> And then motion throws them away before or after compressing.
> I don't know how motion handles situations like this, but the
> behaviour/symptoms you describe make sense. (including low cpu load)
>
> Could you run hdparm -t /dev/mmcblk0 while nothing else is running to
> get a feel for the baseline.
> Either iotop or htop (apt-get install ..) would show you what is going
> on on the system.
> Different sdcards come in different speeds, worth a try...
>
>
> Kind regards from your random penguin caretaker,
>
> Op ma 10 feb. 2020 om 16:40 schreef Jack Christensen
> <christensen.jac...@gmail.com <mailto:christensen.jac...@gmail.com>>:
>
>     Ah yes, I increased log_level to 8 and now I see a message like
>     this for
>     each event:
>
>     event_new_video: Source FPS 4
>
>     I'll let that run for a while and see if it ever changes.
>
>     But it's not that the entire movie is at a different framerate, it's
>     seemingly random frames missing within a movie. For example, here are
>     the timestamps I see on an 8-second movie, going frame by frame:
>
>     10:46:57-02
>     10:46:57-03
>     10:46:58-00
>     10:46:59-00
>     10:46:59-01
>     10:47:00-00
>     10:47:00-01
>     10:47:01-00
>     10:47:02-00
>     10:47:02-01
>     10:47:03-00
>     10:47:04-00
>     10:47:04-01
>     10:47:05-00
>     10:47:05-01
>
>     On 2/10/20 10:13 AM, Jack Christensen wrote:
>
>     > Would I need a log_level greater than 6 for that?
>     >
>     > With log_level 6, I just see framerate in one message at
>     startup, e.g.:
>     >
>     > mmalcam_start: MMAL Camera thread starting... for camera
>     > (vc.ril.camera) of 896 x 504 at 4 fps
>     >
>     > For each event, I only see three lines, e.g.:
>     >
>     > event_newfile: File of type 8 saved to:
>     > /home/jack/motion-output/2-20200209-023725-001.mkv
>     > motion_detected: Motion detected - starting event 1
>     > mlp_actions: End of event 1
>     > event_newfile: File of type 8 saved to:
>     > /home/jack/motion-output/2-20200209-023940-002.mkv
>     > motion_detected: Motion detected - starting event 2
>     > mlp_actions: End of event 2
>     > event_newfile: File of type 8 saved to:
>     > /home/jack/motion-output/2-20200209-024138-003.mkv
>     > motion_detected: Motion detected - starting event 3
>     > mlp_actions: End of event 3
>     >
>     > On 2/10/20 9:50 AM, tosiara wrote:
>     >> Check for FPS in your log when motion records a movie - it will
>     >> reflect the real frame rate that motion is processing
>     >>
>     >> On Mon, Feb 10, 2020 at 4:45 PM Jack Christensen
>     >> <christensen.jac...@gmail.com
>     <mailto:christensen.jac...@gmail.com>
>     <mailto:christensen.jac...@gmail.com
>     <mailto:christensen.jac...@gmail.com>>>
>     >> wrote:
>     >>
>     >>? ? ?I also wondered about resource issues on such modest machines.
>     >>? ? ?Watching
>     >>? ? ?the CPU utilization and load averages, I didn't think I saw
>     too much
>     >>? ? ?cause for concern. The 15-minute load average runs about
>     0.33 and top
>     >>? ? ?shows about 250MB available memory.
>     >>
>     >>? ? ?I wonder if Motion gives any indication of resource issues.
>     I did try
>     >>? ? ?increasing the log_level setting but did not see anything
>     in the
>     >>? ? ?logs.
>     >>
>     >>? ? ?I collected 24-hour load stats from my camera machines. All are
>     >>? ? ?Pi Zero
>     >>? ? ?Ws with the V2 Pi camera. Attached is a chart of the data
>     for cam2,
>     >>? ? ?which had the largest spike. There were several longer than
>     usual
>     >>? ? ?movies
>     >>? ? ?at about that time, so that may account for the spike.
>     >>
>     >>? ? ?They do get busy at times for relatively short intervals,
>     but the
>     >>? ? ?missing frames are consistent, I'm pretty sure they're in every
>     >>? ? ?movie.
>     >>
>     >>? ? ?hostname? avg15? avg05? avg01? max15? max05? max01
>     >>? ? ?cam1????? 0.35?? 0.36?? 0.36?? 0.55?? 0.66?? 1.03
>     >>? ? ?cam2????? 0.32?? 0.33?? 0.33?? 0.59?? 1.00?? 1.54
>     >>? ? ?cam3????? 0.31?? 0.32?? 0.33?? 0.56?? 0.77?? 1.31
>     >>
>     >>
>     >>? ? ?On 2/8/20 5:55 PM, Dave Howorth wrote:
>     >>? ? ?> On Sat, 8 Feb 2020 15:57:55 -0500
>     >>? ? ?> Jack Christensen <christensen.jac...@gmail.com
>     <mailto:christensen.jac...@gmail.com>
>     >>? ? ?<mailto:christensen.jac...@gmail.com
>     <mailto:christensen.jac...@gmail.com>>> wrote:
>     >>? ? ?>
>     >>? ? ?>> Once Motion starts recording a movie, should it continually
>     >>? ? ?capture
>     >>? ? ?>> frames at the configured rate until the end of the
>     event? I am
>     >>? ? ?using a
>     >>? ? ?>> framerate of 4, but missing frames are common, and only
>     seeing one
>     >>? ? ?>> frame per second in the recorded movie is not unusual.
>     If this is
>     >>? ? ?>> expected behavior, what causes it?
>     >>? ? ?> Dunno, but I see similar things on my system and I put it
>     down to
>     >>? ? ?> limited processor power. My system is an older pi and my
>     camera is
>     >>? ? ?> USB-connected, which I believe make my problems worse so I've
>     >>? ? ?never got
>     >>? ? ?> around to investigating further. I see a frame rate of 1 or 2
>     >>? ? ?frames a
>     >>? ? ?> second, very rarely a third.
>     >>? ? ?>
>     >>? ? ?>> I am using Raspberry Pi Zero Ws with V2 Pi cameras. I have
>     >>? ? ?observed
>     >>? ? ?>> this with Motion 4.2.2 both on Raspbian Lite Stretch and
>     >>? ? ?Raspbian Lite
>     >>? ? ?>> Buster, but it seems more pronounced with Buster.
>     >>? ? ?>>
>     >>? ? ?>> Config file attached.
>     >>? ? ?>
>     >>? ? ?> _______________________________________________
>     >>? ? ?> Motion-user mailing list
>     >>? ? ?> Motion-user@lists.sourceforge.net
>     <mailto:Motion-user@lists.sourceforge.net>
>     >>? ? ?<mailto:Motion-user@lists.sourceforge.net
>     <mailto:Motion-user@lists.sourceforge.net>>
>     >>? ? ?> https://lists.sourceforge.net/lists/listinfo/motion-user
>     >>? ? ?> https://motion-project.github.io/
>     >>? ? ?>
>     >>? ? ?> Unsubscribe:
>     >>? ? ?https://lists.sourceforge.net/lists/options/motion-user
>     >>? ? ?_______________________________________________
>     >>? ? ?Motion-user mailing list
>     >>? ? ?Motion-user@lists.sourceforge.net
>     <mailto:Motion-user@lists.sourceforge.net>
>     >>? ? ?<mailto:Motion-user@lists.sourceforge.net
>     <mailto:Motion-user@lists.sourceforge.net>>
>     >>? ? ?https://lists.sourceforge.net/lists/listinfo/motion-user
>     >>? ? ?https://motion-project.github.io/
>     >>
>     >>? ? ?Unsubscribe:
>     https://lists.sourceforge.net/lists/options/motion-user
>     >>
>     >>
>     >>
>     >> _______________________________________________
>     >> Motion-user mailing list
>     >> Motion-user@lists.sourceforge.net
>     <mailto:Motion-user@lists.sourceforge.net>
>     >> https://lists.sourceforge.net/lists/listinfo/motion-user
>     >> https://motion-project.github.io/
>     >>
>     >> Unsubscribe:
>     https://lists.sourceforge.net/lists/options/motion-user
>
>
>
>     _______________________________________________
>     Motion-user mailing list
>     Motion-user@lists.sourceforge.net
>     <mailto:Motion-user@lists.sourceforge.net>
>     https://lists.sourceforge.net/lists/listinfo/motion-user
>     https://motion-project.github.io/
>
>     Unsubscribe: https://lists.sourceforge.net/lists/options/motion-user
>
>
>
> _______________________________________________
> Motion-user mailing list
> Motion-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/motion-user
> https://motion-project.github.io/
>
> Unsubscribe: https://lists.sourceforge.net/lists/options/motion-user
-------------- next part --------------
An HTML attachment was scrubbed...

------------------------------



------------------------------

Subject: Digest Footer

_______________________________________________
Motion-user mailing list
Motion-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/motion-user


------------------------------

End of Motion-user Digest, Vol 164, Issue 24
********************************************

Reply via email to