another rare bug - page 2 - General Help and Assistance on using CHDK stable releases - CHDK Forum  

another rare bug

  • 48 Replies
  • 8996 Views
*

Offline srsa_4c

  • ******
  • 4060
Re: another rare bug
« Reply #10 on: 26 / March / 2015, 14:19:41 »
Advertisements
@timgor
It would be useful if you could run the same test on a different model - to decide whether the issue is port specific. Also, if the camera doesn't crash, you could try to recover from this situation in your script (if it has switched to playback mode, switch it back to rec mode).

*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #11 on: 26 / March / 2015, 14:26:34 »
Yes I can try different camera and I can try to switch it back to rec mode.
Can optical or CCD high temperature switch it back to play mode?

Re: another rare bug
« Reply #12 on: 26 / March / 2015, 14:36:12 »
Can optical or CCD high temperature switch it back to play mode?
I would guess that is possible but I tend to think the camera would just shutdown instead.

FWIW I did some experimenting with those temperatures and extended video times.  There was a popular theory that Canon limited the maximum length of video to avoid overheating the sensor.   I found that sensor temperature rose and then stabilizes in the first 10 or 15 minutes (IIRC).  I never did get a shutdown or switch to playback mode.
Ported :   A1200    SD940   G10    Powershot N    G16

*

Offline reyalp

  • ******
  • 12218
Re: another rare bug
« Reply #13 on: 26 / March / 2015, 17:15:03 »
Is there any mode in CHDK when it retract lens when it's idle long enough instead of set_record(0)? I know that this is only a function of normal factory camera operation and menu setting.
No, CHDK does not do this, only the canon firmware.

Quote
Can optical or CCD high temperature switch it back to play mode?
Like waterwingz, I expect it would just shut down if anything. AFAIK no one has ever confirmed that there is actually a thermal shutdown. I think I might have encountered this once with A540, but I don't really want to try to reproduce it. I've never noticed any strings in the firmware that seems related.

It's conceivable there could be a bug which which just causes CHDK's get_mode to falsely report that the camera is in playback. If you can see the camera when this happens, it would be interesting to know what the actual state is: Is the lens retracted, is the canon UI showing the normal record or play mode UI?
Don't forget what the H stands for.


*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #14 on: 27 / March / 2015, 23:10:39 »
I definitely have some new unknown thing that can be related to my study.
If I run the script
Quote
sleep(3000)
print(get_mode() == true)
sleep(3000)
then it gives false if I start camera by clicking the on/off button.
and it gives true if I press and hold the on/off button.
The second case is formally wrong because camera is in the play mode in both cases.


Does it makes any other differences for whole CHDK if I keep the on/off camera button pushed?
Any other information about this issue?
« Last Edit: 27 / March / 2015, 23:37:25 by timgor »

*

Offline reyalp

  • ******
  • 12218
Re: another rare bug
« Reply #15 on: 27 / March / 2015, 23:58:38 »
If I run the script
Quote
sleep(3000)
print(get_mode() == true)
sleep(3000)
then it gives false if I start camera by clicking the on/off button.
and it gives true if I press and hold the on/off button.
The second case is formally wrong because camera is in the play mode in both cases.
Which camera does this happen on? Which CHDK version? Are you using an autobuild, or some custom build?

Many CHDK ports switch to record mode if you power on and hold the power button.
Don't forget what the H stands for.

*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #16 on: 28 / March / 2015, 00:42:55 »
I use elph115 camera, 1.4 CHDK version. The lens is retracted when I switch it on so the camera is not in the record mode.

Re: another rare bug
« Reply #17 on: 28 / March / 2015, 09:40:39 »
According to this code at lines 207-210 in boot.c for the ixus132_elph115 (fw1.00b) :
Code: [Select]
    // look at power-on switch sub_FF031094
    // value and pointer from sub_FF090160
    if ((*(int*) 0xC0220100) & 1)
        *(int*)(0x2F30) = 0x200000; // Playmode
    else
        *(int*)(0x2F30) = 0x100000; // Shootingmode
you should be able to start the camera in "playback" mode with a short press (click) of the On/Off button and you should also be able to start the camera in "shooting" mode with a longer press of the On/Off button (hold until you hear the lens start to move).

Can you confirm that?
Ported :   A1200    SD940   G10    Powershot N    G16


*

Offline timgor

  • ***
  • 150
Re: another rare bug
« Reply #18 on: 28 / March / 2015, 15:10:23 »
Can you confirm that?
Yes it works as you said. Hence this cannot be the reason for my problem.

There is another question. When I monitor camera temperatures
get_temperature(0),get_temperature(1),get_temperature(2) gives the following numbers:
30   35   90   (at the beginning of camera work)
46   71   90   (after few hours of work)
And script performance slows down by factor of 2.
Is it normal that CCD temperature correlates to camera performance?
What would be critical temperature values that should be avoided?

*

Offline reyalp

  • ******
  • 12218
Re: another rare bug
« Reply #19 on: 28 / March / 2015, 15:23:03 »
Yes it works as you said. Hence this cannot be the reason for my problem.
I don't understand. As far as I understood, you previously said that if you hold down the power button, the lens does not extract and script get_mode falsely reports the camera is in record mode. I can't see how both this and your most recent post can be true at the same time...
Quote
And script performance slows down by factor of 2.
How are you measuring this? In general unless you mess with set_yield, script execution time will be dominated by the automatic yields to make sure kbd_task doesn't hog the CPU, and unless your script is extremely complicated or using loops with lots of iterations, actual code execution time is negligible.
Quote
Is it normal that CCD temperature correlates to camera performance?
This isn't something I've heard of or observed before, but I haven't looked for it. Thermal throttling is not unexpected in this kind of device. The camera may also change noise reduction strategies based on temperature, which could affect shooting time. 71 C seems quite high, I don't recall ever seeing much over 50 on my cams.
Quote
What would be critical temperature values that should be avoided?
Assuming your observation is correct, that's something you'll need to determine by experiment.
Don't forget what the H stands for.

 

Related Topics