[Elphel-support] limiting the frame rate

Sebastian Pichelhofer sebastian.pichelhofer at gmail.com
Wed Feb 9 01:36:39 PST 2011


Thanks for questions and answers!

I collected everything into a new wiki page:
http://wiki.elphel.com/index.php?title=FPSControl

Regards Sebastian

On Wed, Feb 9, 2011 at 03:37, Andrey Filippov <andrey at elphel.com> wrote:

> Adam,
>
> this is the camera problem - seems like some old features (supported in
> GUI) are broken in the later firmware, some new features are not controlled
> from GUI.
>
> Sensor has 2 operation modes - free running and triggered. In the camera
> you can set trigger to FPGA trigger generator (or you can trigger from the
> external source, i.e. another camera).
> The first (free running) mode can achieve higher frame rate and exposure
> can overlap frames (while you are reading the current frame exposure of the
> next one already began). In triggered mode the minimal frame period is the
> frame readout time (same as the frame period in the free running mode) plus
> exposure time plus 8 extra lines (some sensor internal overhead).
>
> On the other hand it is much easier to set any required frame period  in
> triggered mode than in free running - you can just specify the period in
> single pixel clock increments (as long as it is above the minimal for this
> image size and exposure, of course). In the free running mode the software
> can only adjust frame period by changing vertical and horizontal blanking>
> It is also possible to change the sensor clock frequency, but sensor "does
> not like" when the clock is adjusted while it is running and may hang).
>
> This is why when you do not need the highest fps it is usually more
> convenient to run sensor in triggered mode. It is not controlled by the
> camera GUI, but you can change it from the custom script or with
> parsedit.php. If you just open the
> http://192.168.0.9/parsedit.php page it will show you some sample links to
> control different parameters, just select the "External Trigger Controls".
> You may even reduce the number of parameters using the following link:
>
>
> http://192.168.0.9/parsedit.php?embed=0.1&title=External+trigger+controls&TRIG&TRIG_PERIOD
>
> Each parameter has a short description that would appear as a tooltip when
> you put the mouse pointer over it.
>
> The triggered mode is turned on by setting TRIG=4, free running mode
> TRIG=0.
>
> TRIG_PERIOD - frame period in pixel clock periods, by default the clock is
> set to 96MHz.
>
> If you set something too long, camera may wait and wait... Refreshing the
> page will not help, at some point all 3 currently specified instances of PHP
> will be used (something did not work with dynamically increasing number of
> instances) and you'll not be able to open any PHP-based pages (i.e.
> parsedit.php). To recover from that condition (without power-cycling the
> camera) you may telnet to the camera and run
> killall -9 lighttpd # - that will kill web server and all PHP instances,
> then restart them all.
>
> That PHP restart still will not "unfreeze" the camera from waiting the
> trigger, the driver responsible for changing the run time parameters is
> driven by the sensor frame sync interrupts, you may check the status of the
> camera/sensor by refreshing the http://192.168.0.236:8081/pointers page -
> it will work even when the main web server and php are busy.  If frame
> number is not changing - the sensor is not being triggered. In that case you
> can open external trigger control page, set TRIG=0 and before applying
> change the value in "Program ahead" column from 3 to 0 - in that case the
> driver will not wait the frame sync interrupts before applying the
> parameter(s).
>
> Andrey
>
>
>
>
> _______________________________________________
> 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/20110209/135d93ba/attachment-0002.html>


More information about the Support-list mailing list