wait_for_event causes a picture to be taken? #358
Comments
|
This happens when wait_for_event runs and you press the shutter button on the camera. The shutter button press is reported via event 0xc00e and that in turn will run the canon_capture. |
|
Thanks @msmeissn - this clarifies matters a bit:
... but the thing is, I do not press the shutter button on the camera physically in any way - I just leave the camera as is, connected to USB, and run the python script. And considering I do Could there be any other suggestions as to what might cause a shutter button press event to occur, even if there isn't a physical button press? Maybe I should "drain" the event queue somehow (if possible?) before issuing |
|
hmm. perhaps it is reporting to us the shutterbutton is not pressed... can you actually try to wait longer and press it and see what events arrive? |
Sure - just tried it; first changed the relevant code section to this: ...
camera_config = camera.get_config() # must have this, else effectuated values are not read back in
print("{} about to wait_for_event".format(time.time()))
sys.stdout.flush()
a, b = camera.wait_for_event(10000)
OK, gpprop = gp.gp_widget_get_child_by_name( camera_config, "exposurecompensation" )
...... so the timeout is now 10000 ms - but that makes no difference, because as soon as I get the There could be something in the previous Python code that triggers this event, but I simply cannot see what... Any other suggestions? |
|
Can you get a log output where you actually press the shutter button? I need to see what event is sent and what parameters it gets. I tetsed with my SX100IS, it does not have this specific problem |
|
Many thanks, @msmeissn : First, just wanted to mention that I posted a way to wait for
Thanks for confirming that - good to know its not a general problem, at least: maybe it's because the Canon S3 IS is an old model...
Oh, I see - sorry I misunderstood that! :) I wasn't sure at first how to obtain such a log, but then I just used ... first the camera "itself" took one image, and then i pressed the shutter button, it took the second picture, and then command exited. After re-running it, I came to the point where the command just waits, I press the shutter button, picture is taken, and command exits: Here is that (gzipped) log, attached: Let me know if I should provide anything else. |
|
that is fine, thanks! I changed the code to ignore the 0xc00e event now for capture triggering, only the 0xc00b sems to be the shutter button press. |
|
Many thanks again, @msmeissn :
Fantastic - I just tested, and it seems that fixes the This is what I did (Ubuntu 18.04): This process ends succesfully, and I get files like Then, in order to use these, I need to have this massive preamble for ... or as one line: Note that when you call ... you will still get the old libraries listed. However - using This works on python-gphoto2 too; so I've ran the script in the OP again: ... and there are no more mentions of I also tried the This result I get, if Live View has not been exercised yet: ... is one with no spurious photos taken - and where the exposurecompensation ro state does not change neither when switching to Manual, nor when switching back to Auto. And this is the result when having exercised live view: ... where again no spurious photos are taken - and exposurecompensation ro state changes only when switching to Manual (but not when switching back to Auto) So while the problem this issue was originally posted for is fixed -- there are no more spurious photos taken when using Thanks again - will close this issue now, as the problem has been solved! |
I first started with this issue at jim-easterbrook/python-gphoto2#75 - but it seems this may be more related to libgphoto2.
The problem is, on a Canon S3 IS, that I cannot get
exposurecompensation's read-only property to change (immediately after) in response to a change ofshootingmode, so I need to either arbitrarilysleepfor 2 seconds, or wait in a different way. I tried waiting withwait_for_event- but that command initiates taking of a picture!?So, I'd like to know what causes
wait_for_eventto take a picture in this case - because if I can prevent that, I could likely wait for theexposurecompensation's read-only property to change in response to a change ofshootingmode, without guessing arbitrary sleep times.Anyways, I tried to debug more, and this is the version of the script that I used here:
I call this with:
The
test-sm-ec.logis attached:test-sm-ec.log
... and the relevant part is:
So after this printout I call the equivalent of:
gp.gp_widget_get_child_by_name( camera_config, "shootingmode" ).set_value( "Auto" ), which is indicated next in the log:... and then
camera.set_config(camera_config), which is also indicated as the log continues:... and once that - and
camera_config = camera.get_config()is done:... I call
a, b = camera.wait_for_event(10)which is indicated in the log:... but immediately after that, there is
camera_prepare_captureandcamera_canon_capture- where did those come from? The next line aftera, b = camera.wait_for_event(10)isOK, gpprop = gp.gp_widget_get_child_by_name( camera_config, "exposurecompensation" )- which is first mentioned in the log some 5000 lines after this location!?So, to sum up - what could possibly cause a picture to be taken on
wait_for_event, and how to prevent it, if possible?The text was updated successfully, but these errors were encountered: