A590IS porting...

[PhyrePhoX]

hm, now that you mention it - i dont know. thats odd.

[ewavr]

Try for PT_Playsound 0xFFC4C8B8 for both 1.00E and 1.01B

sub_FFC4CEDC have different argument:

int sub_FFC4CEDC(some_struct_from_two_int *)

[PhyrePhoX]

ok, now we wait and see.

[whim]

@phox

PT_Playsound now works on 101b.

wim

[gho]

Hi, I just downloaded the newest version. I can confirm that the metronome sound is working fine.

I just want to add that I am deeply impressed with this software. I do use the A590 IS mostly for low light black and white photography. The raw capabilities give me the possibility to produce pictures with a much better quality than the standard jpegs. Especially in terms of noise and granularity at high ISO settings. Bravo!

[whim]

@gho

i suppose you've got a 100E ? metronome script is still crashing on 101B (maybe the LEDs ?)

wim

[ewavr]

metronome script is still crashing on 101B (maybe the LEDs ?)

Yes, because PostLedMessage() not exists in a590 (not only string "postledmessage" absent, but function also).

[whim]

@ewavr

does that mean that this is not correct 

(from trunk620 platform/a590/sub/101b/stubs_entry_2.S)

--snip--
NSTUB(PhySw_testgpio,   0xFFC356EC)
NSTUB(PostLEDMessage,   0xFFC34B78)
NHSTUB(PT_PlaySound,      0xFFC4C8B8)
--snip--

wim

[Coutts]

i put the latest build on, and tried to use the dng feature. when i try to enable it, it just says "cannot find chdk/badpixel.bin"

i tired making a badpixel.bin file (just a blank file) and putting it there but it still says it.

ugh :-s ideas?

[ewavr]

does that mean that this is not correct 

NSTUB(PostLEDMessage,   0xFFC34B78)

Yes, sub_FFC34B78 is other function, it has (int, int) arguments, but PostLedMessage(struct *)
For example, at comparing SX100IS 1.00C and A590 1.01B:

sub_FFC37280(int, int) == sub_FFC34B78(int, int)
sub_FFC3734C(int, int) == sub_FFC34C14(int, int)
sub_FFC373E8(int, int) == sub_FFC34CE0 == LedDrive(int, int)
sub_FFC37430 == (not present in a590) == AFAssistBeam()
sub_FFC37468== (not present in a590) == PostLEDMessage()
sub_FFC37508 == sub_FFC34D18
usw...

Maybe for A590 we must use LedDrive(int ?, int ?) or direct LED control via 0xC022nnnn ports for ubasic_set_led().

[Coutts]

nvm i found the program to generate the bad pixel file.

but the feature doesn't seem to work on my a590. i have it enabled in the menu but it still saves crw files.

on build #620 with an A590IS:
Things That Work:
- start up sound
- metronome script
- countdown script
- input script
- intervalometer script (one on chdk wiki)
- lotto script
- stopwatch script
- universal tv mode lua script (from this forum, in the scripts section)
- motion detection script (latest version from the scripts section on here)


Things that Don't Work:
- DNG saving (it's set to save a dng but still saves crw files)
- egg timer script (camera crashes when you select which kind of egg you're cooking (hard and soft boiled etc etc)
- dark frame subraction (set it to off but still shoots a dark frame)
- paramdump script (i run it and it says:
 

Code: [Select]

*** STARTED ***
 :9: malloc failed in luaC
 B_get_parameter_data
 PRESS SHUTTER TO CLOSE
 *** INTERRUPTED ***
 optical zoom in video
autofocus button in video settings (camera crashes when you try to use it while shooting a video)
video quality settings (with the display enabled to show remaining video time, it says i have more time remaining shooting at quality setting 99 vs 84, some sort of bug?)
bad pixel removal under raw settings (i don't notice anything, hows it work exactly?)
battery meter (it skips around a lot, is it working correctly?)


i hope you guys can use my data i keep giving you to further develop and debug this firmware! keep up the good work, you guys are doing an excellent job! 

[ewavr]

but it still saves crw files.

The special extension for dng files it is not provided at present. DNG file is 37 kbytes larger than usual RAW file .

[Coutts]

i edited my post to be easier to read, please read it over =D

edit:
ewavr i tried it again.

i shot one shot with raw and shot the same picture again with dng save enabled.

the second shot (the should be dng one) is 38,400bytes larger (37.5kb).

i renamed it to .dng and after doing the registry tweak in windowsxp sp2 (there was a link in some thread on here about dng saving linking me to it), i can see the thumbnail, so the thumbnail works, the saturation just is really low in the thumbnail, either that or the white balance is different in it, idk lol.

i try to open the new .dng file that i just renamed in photoshop cs4 to bring up the raw editor dialog, but no luck. photoshop complains: "Could not complete your request because the file-format module cannot parse the file".

err, do i have to run it through dng4ps still? i thought this whole dng on camera thing was supposed to eliminate that step? :-s

[ewavr]

i can see the thumbnail, so the thumbnail works, the saturation just is really low in the thumbnail, either that or the white balance is different in it, idk lol.

It is normal.

i try to open the new .dng file that i just renamed in photoshop cs4 to bring up the raw editor dialog, but no luck. photoshop complains: "Could not complete your request because the file-format module cannot parse the file".

For developers:

#define CAM_COLORMATRIX1                               \
0.647380, 1000000, -0.169846, 1000000, -0.115337, 1000000, \
-0.011566, 1000000, 0.451223, 1000000, 0.013279, 1000000, \
0.062541, 1000000, 0.054140, 1000000, 0.161148, 1000000

Red numbers are converted to integer numbers, and we obtain:

{0,0,0,
0,0,0,
0,0,0}

color matrix.

Maybe A590 must be unported and then ported again from the beginning.   

[Coutts]

reported??