Hello Daniel,
1. Setting the WOI:
1a. Set the width (doesn't work correctly in camvc - so, Parameter Editor
> multisensor):
I could reproduce your problem - that's a bug.
Notes and temporary solution:
1a1. It's ok to change WOI_WIDTH only.
1a2. If the image is scrambled - setting WOI_WIDTH to 5
Of course, this is not my top priority project in the coming weeks.
Thanks again.
Daniel
Le 10 sept. 2013 à 07:36, support-list a
écrit :
> This is really strange, we'll have to try to reproduce it. Oleg is on
> vacation until September, 17 - he worked most with the stereo setup. He
Andrey,
Le 7 sept. 2013 à 20:37, support-list a écrit
:
> Imgsrv has "/trig" command that issues a single trigger if the camera is set
> to single trigger mode, but for testing of the corrupted images it may be
> simpler to just use some definitely slow fps (like 1 FPS) by setting
> TRIG_PE
Daniel,
The imgsrv - image server on port 8081 has a mini-help page - just open
http://:8081 page. This server is for manipulation data in the
circular buffer, it also does not depend on the main lighttpd that can get
stuck when there are several instances of the PHP busy waiting for
Le 5 sept. 2013 à 23:44, support-list a écrit
:
> Daniel
>
> For FPS control I meant using TRIG=4 as described in
> http://wiki.elphel.com/index.php?title=Trigger
> TRIG_CONDITION=0 means that the FPGA timer will generate trigger pulses, not
> the external connector
OK, this is the basic si
Daniel
For FPS control I meant using TRIG=4 as described in
http://wiki.elphel.com/index.php?title=Trigger
TRIG_CONDITION=0 means that the FPGA timer will generate trigger pulses, not
the external connector
Did you try to change height after reducing width (and getting bad images)?
Do both sen
Thanks Andrey,
Le 5 sept. 2013 à 09:36, support-list a écrit
:
> >In fact for my application, I had to buy (at some Chinese website) longer
> >cables and I just installed only one on one channel, keeping the original
> >medium long on the other: is it a possible cause of the problem ?
>
> D
>In fact for my application, I had to buy (at some Chinese website) longer
cables and I just installed only one on one channel, keeping the original
medium long on the other: is it a possible cause of the problem ?
Daniel, yes - it can cause problem - those that are visible on the first image.
Dear Daniel, I'm sorry but I just found out that we do not have any asembled
stereo cameras around to reproduce the problem, we plan to have on in 2-3
weeks. So le's try to troubleshoot the problems with your camera.
What we can try first is to isolate - is that wrong geometry setting or wrong
Dear Andrey,
Thanks for the advices: I'll do what you suggest.
Le 5 sept. 2013 à 06:25, support-list a écrit
:
> I also noticed the another problem on the first image - wrong sensor board
> phase adjustment (it depends on the cable length that is longer in stereo
> cameras). Oleg did such ad
Dear Andrey,
Here: https://upload.obspm.fr/get?k=OKsc283yHuLvGsNuOFu are the pdf of three
parameters/init pages (parameters editor):
Parameters_init0 : is the standard configuration after a cold start ; full
size, the images are "normal"; note that the parameters MULTI_WIDTHn and
are not s
Dear Daniel,
It is definitely possible to run smaller WOI on the camera, can you please send
some detailed step-by-step instructions so we can reproduce your case and get
the same errors?
The camera GUI interface (camvc) is a rather old application that was updated
to accommmodate new sensor c
Dear Andrey and Oleg,
After a lot of testing I'll try to be more specific. I'm not sure that the
problem comes from
the de-Bayer process. Probably no.
My goal is to acquire (through the disk recorder for instance) simultaneous
images
of the two Black and White 2592x1944 Aptina sensors of the
Hello Oleg,
I purchased a stereo configuration equipped with Black and white CCDs.
I want to do as fast as possible image acquisition in reduced format.
Unfortunately, using the camera control interface (camvc / tab 5) , the images
become
ugly because some de-Bayer process try to use the color
14 matches
Mail list logo
