SX410 IS porting thread

  • 86 Replies
  • 42954 Views
*

Offline nafraf

  • *****
  • 1303
Re: SX410 IS porting thread
« Reply #10 on: 21 / August / 2016, 21:45:48 »
Advertisements
... ,cam has iris failed 2.
I think sx410 does not have iris, add this to platform_camera.h:
Code: [Select]
#undef  CAM_HAS_IRIS_DIAPHRAGM

*

Offline blackhole

  • *****
  • 642
  • A590IS 101b
    • Planetary astrophotography
Re: SX410 IS porting thread
« Reply #11 on: 22 / August / 2016, 03:55:34 »
Quote
Regardless, if ALT mode doesn't block certain buttons from the Canon UI, there's something wrong with the keyboard...

alvm reported keyboard related issues a while ago on recent PowerShots. If you're unlucky, you're experiencing the same.
To tell you the truth, I have no idea what's going on. I can not solve this problem. I will continue to work on other things that have not yet been made.
Quote
I think sx410 does not have iris, add this to platform_camera.h:
No camera has Iris, no ND filter. SX420 also has Iris. I found this information only on the official website of Canon India. Iris is clearly visible through the lens at a larger zoom.
http://www.canon.co.in/personal/products/compact-cameras/powershot/powershot-sx410is?languageCode=EN#specificationAnchor

*

Offline reyalp

  • ******
  • 11591
Re: SX410 IS porting thread
« Reply #12 on: 22 / August / 2016, 13:26:53 »
No camera has Iris, no ND filter. SX420 also has Iris. I found this information only on the official website of Canon India. Iris is clearly visible through the lens at a larger zoom.
http://www.canon.co.in/personal/products/compact-cameras/powershot/powershot-sx410is?languageCode=EN#specificationAnchor
I suspect this is an error on the Canon India site (copy /paste from another model, perhaps?)

In the past, all powershots with a real iris have Av and M modes. It's hard to see the point of including the hardware without that.
Don't forget what the H stands for.

*

Offline srsa_4c

  • ******
  • 3728
Re: SX410 IS porting thread
« Reply #13 on: 22 / August / 2016, 15:34:18 »
I suspect this is an error on the Canon India site (copy /paste from another model, perhaps?)

In the past, all powershots with a real iris have Av and M modes. It's hard to see the point of including the hardware without that.
It might be an error. However, some Ixus models (for example Ixus300, 310) appear to have an adjustable iris (there are forum discussions about it, and of course the ports are configured accordingly). AFAIK none of the Ixus models have manual modes.

Blackhole could prove it with either a photo that shows the diaphragm blades or a wide angle photo taken with the camera where the recorded aperture is at some intermediate value.

To tell you the truth, I have no idea what's going on. I can not solve this problem.
Unfortunately, it's something we have not experienced before: the shutter buttons (half, full) seem to be evaluated elsewhere, their bits in physw_status are ignored.


*

Offline blackhole

  • *****
  • 642
  • A590IS 101b
    • Planetary astrophotography
Re: SX410 IS porting thread
« Reply #14 on: 23 / August / 2016, 09:07:59 »
Here's 9 images of the same frame with a fixed TV and ISO values @24mm.
AV values ranging from F 3.5 to F 11.0 Minimum declared aperture on the camera's F 6.3
F numbers over F11.0 have no effect on the appearance of the image.
https://drive.google.com/open?id=0B2NRqqO9AOURUUJMQWdPZG5ia1E   6MB

EDIT:Added video
« Last Edit: 28 / August / 2016, 12:33:05 by blackhole »

*

Offline blackhole

  • *****
  • 642
  • A590IS 101b
    • Planetary astrophotography
Re: SX410 IS porting thread
« Reply #15 on: 28 / August / 2016, 12:33:41 »
Video features (bitrate and quality overrides) and additional AF scan, while recording a video are now supported.
Link to the source code in the first post is replaced with a new one.

*

Offline srsa_4c

  • ******
  • 3728
Re: SX410 IS porting thread
« Reply #16 on: 10 / September / 2016, 16:07:38 »
Video features (bitrate and quality overrides) and additional AF scan, while recording a video are now supported.
Link to the source code in the first post is replaced with a new one.
If the port is usable (not counting the keyboard issue), you may want to consider submitting it for inclusion in the official source - so it doesn't get lost.
I probably would not enable autobuild for it until the keyboard related issue is solved.

*

Offline blackhole

  • *****
  • 642
  • A590IS 101b
    • Planetary astrophotography
Re: SX410 IS porting thread
« Reply #17 on: 11 / September / 2016, 05:14:50 »
Quote
If the port is usable (not counting the keyboard issue), you may want to consider submitting it for inclusion in the official source - so it doesn't get lost.
The port is fully usable, except for the running of scripts.
I agree with adding code to the repository, if sometimes finding a solution for a problem with a keyboard that can be easily added later.
Patch is posted to the usual place
Quote
I probably would not enable autobuild for it until the keyboard related issue is solved.
This is understandable, if someone asks for build, it is not a problem to do.


*

Offline srsa_4c

  • ******
  • 3728
Re: SX410 IS porting thread
« Reply #18 on: 11 / September / 2016, 16:04:49 »
Some notes (reyalp may decide to commit the patch any time, so I'm not attempting to check it in).
- the readme should mention the major obstacle, the shutter button behaviour
- it could also mention that the cam has a real iris diaphragm
- from a quick look, the 100b firmware seems compatible
- the LED addresses in lib.c do not match the sigfinder's suggestion AND the code uses the legacy switching method which is unsuitable for 0xc022dXXX GPIOs

*

Offline reyalp

  • ******
  • 11591
Re: SX410 IS porting thread
« Reply #19 on: 11 / September / 2016, 19:23:22 »
A added a link for the shutter problem and mention of iris in notes.txt

In addition to the things srsa_4c mentioned, I noticed that the included movie_rec.c does match what is generated by codegen, because of the lines
Code: [Select]
//"    STMFD   SP!, {R4,R5} \n"
//"    LDR     R5, =0xFFC0CC8C \n"
I don't think this is significant, since the preceding instruction is a B, it looks like sub_FF9154D4_my should just be 2 instructions shorter.

Have the propcases been verified? platform_camera.h has PROPSET 6, but that's quite old compared to the camera release date.

Regarding the button problems
Quote
Other keys I could not check because the value of the PS1 and PS2 are constantly changing.
Probably only some unrelated bits are changing. You can read the values with rmem and convert to binary to see if any additional bits really change in response to the buttons. e.g.
Code: [Select]
rmem -i32 0x12224 3

If the shutter button is truly being detected outside of the normal physw_task code, the camera log might provide some insight into what code is doing it. E.g.

Code: [Select]
!require'extras/devutil'.init_cli()
dlstart
...  half press shutter in alt mode
dlgetcam
Look through the messages in dbg.log for anything that looks related to the half press, if you find them, look for the strings in rom code etc.
Don't forget what the H stands for.

 

Related Topics