D-Link Forums

The Graveyard - Products No Longer Supported => IP Cameras => DCS-931L => Topic started by: acellier on February 05, 2014, 06:00:12 PM

Title: new 931L, old problem.
Post by: acellier on February 05, 2014, 06:00:12 PM
We bought a new DCS-931L to try again - having tried and returned one last summer.
These cameras have an unfortunate tendency to garble the image, as you can see at:

<url>http://smg.photobucket.com/user/acellier/media/DCS-931L_Errors/CamN_2014020517293501.jpg.html</url>
<url>http://smg.photobucket.com/user/acellier/library/DCS-931L_Errors</url>

It's really too bad, as otherwise this model; has a lot going for it.
BTW, we did install the latest v1.02 firmware.
Title: Re: new 931L, old problem.
Post by: JavaLawyer on February 07, 2014, 06:48:18 AM
Do the image artifacts appear on both MJPEG and H.264? My camera is usually set to MJPEG and I haven't encountered this issue.
Title: Re: new 931L, old problem.
Post by: acellier on February 07, 2014, 10:21:30 AM
This is in mjpeg.
I guess I keep getting the bad cameras.
Title: Re: new 931L, old problem.
Post by: JavaLawyer on February 07, 2014, 11:04:37 AM
LOL. . . I don't use motion detection on that particular camera, so it's possible my camera has the same issue, I'm just not catching it.  I'll need to setup motion detection to see if I'm hitting any false positives.
Title: Re: new 931L, old problem.
Post by: acellier on February 07, 2014, 03:45:02 PM
seems to USUALLY corrupt the first image upon motion detection ...
Title: Re: new 931L, old problem.
Post by: JavaLawyer on February 08, 2014, 09:15:52 AM
I setup motion detection with FTP image notification to try to replicate this issue. For testing consistency, what are your video, motion detection, and image capture settings?
Title: Re: new 931L, old problem.
Post by: acellier on February 08, 2014, 09:54:55 AM
Video:   MJPEG 640x480, VeryHigh, Auto:  Default: MJPEG
Moption Detection:   Enable, Time: Always; 65%; all boxes ON
FTP:   Passive: Yes (does not matter)
    Schedule: Enable upload - Motion
    Image frequency: 1 frames/second
    Date/Time Suffix, 0.5 hour
Title: Re: new 931L, old problem.
Post by: JavaLawyer on February 08, 2014, 03:00:04 PM
Quote from: acellier on February 08, 2014, 09:54:55 AM
Video:   MJPEG 640x480, VeryHigh, Auto:  Default: MJPEG
Moption Detection:   Enable, Time: Always; 65%; all boxes ON
FTP:   Passive: Yes (does not matter)
    Schedule: Enable upload - Motion
    Image frequency: 1 frames/second
    Date/Time Suffix, 0.5 hour


I've mirrored your settings. With what frequency are you getting frames with image artifacts?
Title: Re: new 931L, old problem.
Post by: acellier on February 09, 2014, 07:14:54 PM
not every "event", but too many to count, usually at onset.
A shame that this error persists.
My money is on a circuit timing issue.
Title: Re: new 931L, old problem.
Post by: JavaLawyer on February 13, 2014, 04:41:37 AM
So I ran motion detection with FTP using the settings you provided for approximately 5 days.  The area I was monitoring has some moderate activity during certain parts of the day, so there were a good number of valid motion detection events. All said, the FTP directory contained over 11,000 images, most of which were due to valid activity, while a smaller percentage were due to changing light conditions which I could have eliminated had I changed the sensitivity, but I was mirroring your settings.  

Of all 11,000+ images, there were only "5" that had image artifacts that looked identical to what you experienced, three of which occurred on the same day. All things being equal, that would equate to an average of one artifact per day.
Title: Re: new 931L, old problem.
Post by: acellier on February 13, 2014, 08:32:44 AM
Your experience is *much* more benign that mine.
Just to randomly sample a few 1/2 hour intervals -
Wed, Feb 12, 11:00 AM ... 39 images - 3 bad
Tue, Feb 11, 08:30 AM ... 32 images - 1 bad
Wed, Feb 12, 02:30 PM ... 41 images - 1 bad
Sat, Feb 08, 10:30 AM ... 9 images - 1 bad
Fri, Feb 07, 01:00 PM ... 5 images - 1 bad
Thu, Feb 13, 08:00 AM ... 30 images - 0 bad
Wed, Feb 12, 02:30 PM ... 41 images - 1 bad
Title: Re: new 931L, old problem.
Post by: JavaLawyer on February 13, 2014, 10:41:02 AM
I misread the time stamps. It looks like all 5 instances took place over a two hour window during that five day period, which makes this all the more confounding.

Could these artifacts be caused by some form of intermittent environmental interference (e.g. electrical)? This is the only root cause that I can come up with that would explain the indeterminacy and extreme difference in frequency on different cameras of the same model. It would be interesting to test a camera known to have these issues on another off-premise network to see if there's any change in behavior.

 ???   :-\   ???
Title: Re: new 931L, old problem.
Post by: acellier on February 14, 2014, 09:49:44 AM
this happens with 2 different routers, to 2 different servers
Title: Re: new 931L, old problem.
Post by: JavaLawyer on February 17, 2014, 09:53:29 AM
I passed my findings, as well as all of the related DCS-931L and DCS-933L links, to D-Link.  I'll post back if I hear anything.   ;)

As an FYI, aside from posting in this forum, D-Link often escalates issues based on the quantity of reports they receive through their support desk. A large volume of tickets with the same issue raises a flag. You may want to log this issue with technical support for that reason alone: 1-888-851-6464
Title: Re: new 931L, old problem.
Post by: gajego on March 12, 2014, 08:57:14 AM
Quote from: JavaLawyer on February 17, 2014, 09:53:29 AM
I passed my findings, as well as all of the related DCS-931L and DCS-933L links, to D-Link.  I'll post back if I hear anything.   ;)

As an FYI, aside from posting in this forum, D-Link often escalates issues based on the quantity of reports they receive through their support desk. A large volume of tickets with the same issue raises a flag. You may want to log this issue with technical support for that reason alone: 1-888-851-6464

I've been waiting in a queue for 40 min. gave up ...
Title: Re: new 931L, old problem.
Post by: JavaLawyer on March 12, 2014, 08:59:57 AM
Quote from: gajego on March 12, 2014, 08:57:14 AM
I've been waiting in a queue for 40 min. gave up ...

Wow, I've never waited for more than 5-10 minutes (usually less than that even). I suppose it depends on the time of day you are calling. The times I've called were typically first thing in the morning and early evening.
Title: Re: new 931L, old problem.
Post by: JavaLawyer on March 24, 2014, 08:34:54 AM
D-Link expanded the release notes for the latest DCS-931L firmware version 1.03b08 (security update). There are additional changes that were not documented in the first set of release notes: DCS-931L - Cumulative Firmware Release Notes (http://forums.dlink.com/index.php?topic=53797.0)

There are two particular bug fixes that leave me wondering if the issue described in this thread was addressed. The release notes entries are a bit ambiguous:


I'm not sure how to interpret this. . .  :-\

Have any DCS-931L owners who updated with the security update noticed any improvement on the image artifacts discussed in this thread?
Title: Re: new 931L, old problem.
Post by: JavaLawyer on March 25, 2014, 05:27:15 PM
Any updates on this issue from DCS-931L owners using the latest firmware release?
Title: Re: new 931L, old problem.
Post by: JavaLawyer on March 31, 2014, 12:35:59 PM
Good news. At least one user of the new v1.03b08 firmware has reported that this latest firmware has corrected the image artifacts that caused false motion detection triggers in earlier firmware versions.

If you installed this latest firmware, please post your findings!  ;)
Title: Re: new 931L, old problem.
Post by: JavaLawyer on April 08, 2014, 06:46:58 AM
Call for user feedback...

At least two users of the new v1.03b08 firmware has reported that this latest firmware has corrected the image artifacts that caused false motion detection triggers in earlier firmware versions.

If you installed this latest firmware, please post your findings!