Adding new cameras, applying patches into trunk (with source code prepared) - page 111 - General Discussion and Assistance - CHDK Forum  

Adding new cameras, applying patches into trunk (with source code prepared)

  • 1679 Replies
  • 787854 Views
*

Offline philmoz

  • *****
  • 3450
    • Photos
Advertisements
CHDK for PowerShot G15 100b

Enabled BETA autobuild since it seems there's no issue
Since i already posted a big diff, here's a quick fix for CFAPattern and DNG exposure bias
edit: I now see active area is in converted EXIF data, I will use that value... patch soon

edit2: patch attached to match native raw active area and cfapattern

Added in revision 2725/2726 (trunk) and 2727 (release-1.1).

Phil.
CHDK ports:
  sx30is (1.00c, 1.00h, 1.00l, 1.00n & 1.00p)
  g12 (1.00c, 1.00e, 1.00f & 1.00g)
  sx130is (1.01d & 1.01f)
  ixus310hs (1.00a & 1.01a)
  sx40hs (1.00d, 1.00g & 1.00i)
  g1x (1.00e, 1.00f & 1.00g)
  g5x (1.00c, 1.01a, 1.01b)
  g7x2 (1.01a, 1.01b, 1.10b)

*

Online reyalp

  • ******
  • 14082
Added in revision 2725/2726 (trunk) and 2727 (release-1.1).
Note this broke the release autobuild, so I added skip_autobuild. Since I was on my lunch break, I didn't look closely, but I think it was related to changes in the trunk core that aren't in the release branch.

I encourage building before every checkin, even if it seems trivial.
Don't forget what the H stands for.

*

Offline philmoz

  • *****
  • 3450
    • Photos
Added in revision 2725/2726 (trunk) and 2727 (release-1.1).
Note this broke the release autobuild, so I added skip_autobuild. Since I was on my lunch break, I didn't look closely, but I think it was related to changes in the trunk core that aren't in the release branch.

I encourage building before every checkin, even if it seems trivial.

My bad - normally I do build before commit; will fix it tonight if I get time.

Phil.
CHDK ports:
  sx30is (1.00c, 1.00h, 1.00l, 1.00n & 1.00p)
  g12 (1.00c, 1.00e, 1.00f & 1.00g)
  sx130is (1.01d & 1.01f)
  ixus310hs (1.00a & 1.01a)
  sx40hs (1.00d, 1.00g & 1.00i)
  g1x (1.00e, 1.00f & 1.00g)
  g5x (1.00c, 1.01a, 1.01b)
  g7x2 (1.01a, 1.01b, 1.10b)

*

Offline c10ud

  • ***
  • 245
Added in revision 2725/2726 (trunk) and 2727 (release-1.1).
Note this broke the release autobuild, so I added skip_autobuild. Since I was on my lunch break, I didn't look closely, but I think it was related to changes in the trunk core that aren't in the release branch.

I encourage building before every checkin, even if it seems trivial.

My bad - normally I do build before commit; will fix it tonight if I get time.

Phil.
sorry, I forgot to say I tested in trunk only :)

Anyway, here's another load of changes (yes, for trunk), this time for S100 (tested with s100 101b):
hacked loader, removed useless comments, fixed platform_camera

with the hacked loader now booting with power button short press works
fixed defines in platform_camera for correct dng generation
tested movie_rec and it works

link to patch: https://github.com/c10ud/CHDK/commit/bee5ee88f8c94a9fb6d708bde2db3f7fea9cb36c.patch


*

Offline philmoz

  • *****
  • 3450
    • Photos
Added in revision 2725/2726 (trunk) and 2727 (release-1.1).
Note this broke the release autobuild, so I added skip_autobuild. Since I was on my lunch break, I didn't look closely, but I think it was related to changes in the trunk core that aren't in the release branch.

I encourage building before every checkin, even if it seems trivial.

My bad - normally I do build before commit; will fix it tonight if I get time.

Phil.
sorry, I forgot to say I tested in trunk only :)

Anyway, here's another load of changes (yes, for trunk), this time for S100 (tested with s100 101b):
hacked loader, removed useless comments, fixed platform_camera

with the hacked loader now booting with power button short press works
fixed defines in platform_camera for correct dng generation
tested movie_rec and it works

link to patch: https://github.com/c10ud/CHDK/commit/bee5ee88f8c94a9fb6d708bde2db3f7fea9cb36c.patch

I've added this to the trunk in revision 2730.

However the S100 platform_camera.h file contains the active area and color matrices from your G15 port - is this correct?

Once you've confirmed (or fixed) platform_camera.h I'll update release-1.1.

Phil.

Edit: Added to release-1.1 in revision 2733.

« Last Edit: 28 / April / 2013, 03:08:55 by philmoz »
CHDK ports:
  sx30is (1.00c, 1.00h, 1.00l, 1.00n & 1.00p)
  g12 (1.00c, 1.00e, 1.00f & 1.00g)
  sx130is (1.01d & 1.01f)
  ixus310hs (1.00a & 1.01a)
  sx40hs (1.00d, 1.00g & 1.00i)
  g1x (1.00e, 1.00f & 1.00g)
  g5x (1.00c, 1.01a, 1.01b)
  g7x2 (1.01a, 1.01b, 1.10b)

*

Offline c10ud

  • ***
  • 245
Added in revision 2725/2726 (trunk) and 2727 (release-1.1).
Note this broke the release autobuild, so I added skip_autobuild. Since I was on my lunch break, I didn't look closely, but I think it was related to changes in the trunk core that aren't in the release branch.

I encourage building before every checkin, even if it seems trivial.

My bad - normally I do build before commit; will fix it tonight if I get time.

Phil.
sorry, I forgot to say I tested in trunk only :)

Anyway, here's another load of changes (yes, for trunk), this time for S100 (tested with s100 101b):
hacked loader, removed useless comments, fixed platform_camera

with the hacked loader now booting with power button short press works
fixed defines in platform_camera for correct dng generation
tested movie_rec and it works

link to patch: https://github.com/c10ud/CHDK/commit/bee5ee88f8c94a9fb6d708bde2db3f7fea9cb36c.patch

I've added this to the trunk in revision 2730.

However the S100 platform_camera.h file contains the active area and color matrices from your G15 port - is this correct?

Once you've confirmed (or fixed) platform_camera.h I'll update release-1.1.

Phil.
I'm not at home (so I cannot check if I did something wrong) but I got the active area and color matrices from the CR2->DNG Adobe conversion of a S100 raw file

Looking at the diff with the previous S100 platform_camera active area looks the same, just with numbers explicited (I guess S100 and G15 probably have the same sensor or something?)

The color matrices are different, I also remember the S100 DNG missing the calibration info, probably you looked at the wrong diff?

Anyway once I'm home I'll double check this just to be sure

p.s. While we're at it, I didn't really get what the CAM_JPEG_{WIDTH, HEIGHT} are for and which value should the hold (e.g. 4000x3000 for 12Mpix? or CAM_ACTIVE_AREA_X2-CAM_ACTIVE_AREA_X1? or else?)

*

Offline nafraf

  • *****
  • 1308
a2400: 100c, 100d, and 100e blind port ALPHA

100c firmware dump was posted here

*

Offline philmoz

  • *****
  • 3450
    • Photos
a2400: 100c, 100d, and 100e blind port ALPHA

100c firmware dump was posted here

Added to trunk in revision 2734.

Phil.
CHDK ports:
  sx30is (1.00c, 1.00h, 1.00l, 1.00n & 1.00p)
  g12 (1.00c, 1.00e, 1.00f & 1.00g)
  sx130is (1.01d & 1.01f)
  ixus310hs (1.00a & 1.01a)
  sx40hs (1.00d, 1.00g & 1.00i)
  g1x (1.00e, 1.00f & 1.00g)
  g5x (1.00c, 1.01a, 1.01b)
  g7x2 (1.01a, 1.01b, 1.10b)


*

Offline c10ud

  • ***
  • 245
Added in revision 2725/2726 (trunk) and 2727 (release-1.1).
Note this broke the release autobuild, so I added skip_autobuild. Since I was on my lunch break, I didn't look closely, but I think it was related to changes in the trunk core that aren't in the release branch.

I encourage building before every checkin, even if it seems trivial.

My bad - normally I do build before commit; will fix it tonight if I get time.

Phil.
sorry, I forgot to say I tested in trunk only :)

Anyway, here's another load of changes (yes, for trunk), this time for S100 (tested with s100 101b):
hacked loader, removed useless comments, fixed platform_camera

with the hacked loader now booting with power button short press works
fixed defines in platform_camera for correct dng generation
tested movie_rec and it works

link to patch: https://github.com/c10ud/CHDK/commit/bee5ee88f8c94a9fb6d708bde2db3f7fea9cb36c.patch

I've added this to the trunk in revision 2730.

However the S100 platform_camera.h file contains the active area and color matrices from your G15 port - is this correct?

Once you've confirmed (or fixed) platform_camera.h I'll update release-1.1.

Phil.
I'm not at home (so I cannot check if I did something wrong) but I got the active area and color matrices from the CR2->DNG Adobe conversion of a S100 raw file

Looking at the diff with the previous S100 platform_camera active area looks the same, just with numbers explicited (I guess S100 and G15 probably have the same sensor or something?)

The color matrices are different, I also remember the S100 DNG missing the calibration info, probably you looked at the wrong diff?

Anyway once I'm home I'll double check this just to be sure

p.s. While we're at it, I didn't really get what the CAM_JPEG_{WIDTH, HEIGHT} are for and which value should the hold (e.g. 4000x3000 for 12Mpix? or CAM_ACTIVE_AREA_X2-CAM_ACTIVE_AREA_X1? or else?)
Just tried current vanilla trunk, everything's good ;)

It probably appropriate to disable the hooks for capt_seq_hook_set_nr in capt_seq.c for the SX50HS - at least for now.

In the current build, they do not work  (i.e. you can not enable or disable dark frame subtraction)  and they cause an immediate crash when the camera is used in "Sports" mode  (continuous shooting).

I'll keep looking at it but in the meantime, best it's disabled.

Patch works for both 1.1.0 &  1.2.0 branches.
Ported :   A1200    SD940   G10    Powershot N    G16

 

Related Topics