[Elphel-support] Problems to detect timestamps

Sun Jan 10 11:12:57 PST 2016

Hello,

Thank you very much for the information and quick response. We study in
detail what you proposed.

I'm going to explain a little what it will be our project.
Our goal is a multi-view synchronized capture for  3D analysis with 16
Elphel cameras (An example that has already been made in this university
<https://imatge.upc.edu/web/resources/free-viewpoint-video-fvv-multiview-data>
 )

   1. These analyzes can be of recordings stored on disk or in real time.
   Therefore the timestamps are very important in our project.
   2. Millisecond resolution would almost enough for what we want to do,
   but if we have a microseconds resolution would be better.
   3. The real-time processing does not want to do in the camera, we want
   to do on remote machines (inside a LAN) GPUs, etc.

What would be your recommendation for this particular project?

Kind regards,

Jennifer Valle

2016-01-09 20:07 GMT+01:00 support-list <support-list at support.elphel.com>:

> Hello Jennifer,
>
> Unfortunately we did not had enough resources to get into the
> videosteramer code - we do not use it ourselves because of the multiple
> limitation of the streaming. It strips frames of the Exif headers that we
> use not only for the time stamping (with 1 microsecond resolution), but for
> other metadata too (like IMU or GPS data). Additionally videostream is
> designed for just synchronous streaming, while camera has a large
> videobuffer and it is possible to control it precisely, for example
> acquiring higher resolution/quality using longer time and then "catching
> up" either skipping frames completely or requcing resolution/quality.
>
> Can you please describe your overall application, it may be possible to
> achieve the same goals differently, using approach similar to what we use
> ourselves?
>
> There are several tools available for capturing and synchronizing images:
> 1. camogm (http://wiki.elphel.com/index.php?title=Camogm ) allows to
> record sequences of images, each precisely timestamped and having optional
> GPS data (if GPS is connected). Multiple cameras can be synchronized
> together with microsecond precision, so each camera sequence of images will
> have frames marked with the absolutely the same timestamps.
>
> 2. imgsrv ( http://wiki.elphel.com/index.php?title=Imgsrv ) - optimized
> image server available at http://<camera_ip>:8081 - it has a mini-help
> page when you open this URL without additional parameters. This is the
> fastest way to grab images over the network that fully utilizes 19MB buffer
> in the camera, can be used with the client-based scripts or the in-camera
> PHP code. Among the operation modes it has multi-part JPEG (/mimg suffix in
> imgsrv request) that is supported by multiple browsers - it is very easy to
> decode as it is just a sequence of individual JPEG images with complete
> Exif headers.
>
> 3. Event logger (http://wiki.elphel.com/index.php?title=Event_logger) is
> the FPGA-based system that combines image data timestamping (from multiple
> cameras), GPS data (including precise absolute timing) and high-speed IMU
> (we use 2.5K samples/s) that allows absolute timestaping of each image with
> the accuracy not possible with just internal camera oscillator.
> Additionally the logger accepts external signals over the dedicated input
> and/or over the network (with network the timing uncertainty of the logged
> event will be higher - around 2 ms).
>
> Andrey
>
> ---- On Sat, 09 Jan 2016 01:12:24 -0800 *Jennifer Valle
> <jvs1192 at gmail.com <jvs1192 at gmail.com>>* wrote ----
>
> Good Morning,
>
>  A month ago I wrote a message about the problems I had with timestamps
> that obtained with the RTSP protocol live555. It is a project of validation
> of cameras in space-time and the synchronization is very important. I would
> be very grateful if you could solve these problems before February, when I
> present the final degree project.
>
> Kinds regards,
>
> Jennifer Valle
>
> 2015-12-01 22:32 GMT+01:00 Oleg <support-list at support.elphel.com>:
>
> Hello Jennifer,
>
> Were you able to solve your problem?
>
> I think there has been a confusion, when I referred to the "original" code
> live555, I was referring to the latest version without changing code. The
> problem is that when I connect to the camera with the instruction openrtsp
> -m, I get files with the timestamp (PresentationTime) in the filename and
> these values are incorrect.
>
> video-JPEG-1-1448298848.836308
> video-JPEG-1-3534277304.245330
> video-JPEG-1-3534277304.281407
> video-JPEG-1-3534277304.317484
> video-JPEG-1-3534277304.353561
> video-JPEG-1-3534277304.389627
>
>
> The first file has a correct timestamp, but in the following files have
> incorrect "times" (year 2081).
>
>
> I was able to reproduce your problem. Installed live555 version
> (dated 2015-11-09).
> The first correct timestamp is the timestamp of the PC.
> Might be something with the RTSP server implementation in the camera. To
> find out I need to have a closer look at the code:
> http://elphel.cvs.sourceforge.net/viewvc/elphel/elphel353-8.0/apps/astreamer/
> Will take some time - busy with other things.
>
> Best regards,
> Oleg Dzhimiev
> Electronics Engineer
> phone: +1 801 783 5555 x124
> Elphel, Inc.
>
>
> _______________________________________________
> Support-list mailing list
> Support-list at support.elphel.com
> http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://support.elphel.com/pipermail/support-list_support.elphel.com/attachments/20160110/7565508b/attachment-0002.html>