SX 10 - Extended AV!!! - page 2 - General Discussion and Assistance - CHDK Forum

SX 10 - Extended AV!!!

  • 34 Replies
  • 10082 Views
Re: SX 10 - Extended AV!!!
« Reply #10 on: 21 / October / 2009, 11:55:30 »
Advertisements
I did a quick check of the capt_seq.c for the SX10 and SX1 builds.
The function address I've been using is:
SX10 103a: FF93B440

for SX10 version 101b and 102b, the corresponding address appears to be identical.  I'll have to look in the FW to verify if all addresses in the branches are identical as well.

For all other versions of the SX10 and SX1 the reference addresses are different, but the general structure looks the same.  Would likely only need updating of the reference addresses within th branches:

SX10 100c: FF93B398
SX10 101a: FF93B3A0
SX1  200h & 201a:FF94EAF0

Note that these are only a surface check within the available code in the capt_seq.c files to do a preliminary check on addresses and structure, not from firmware disassembly.

-tgq

Re: SX 10 - Extended AV!!!
« Reply #11 on: 21 / October / 2009, 14:42:29 »
Code: [Select]
ffa568e4: e1510000 cmp r1, r0
ffa568e8: 11d120f0 ldrshne r2, [r1]
ffa568ec: 11520004 cmpne r2, r4
This appears to be just an error check as R1 should always equal R0 as it was set at ffa568dc.
Sorry, I misspoke here, this actually gets triggered when an AV apex96 value is not on the table and ends up doing a linear approximation between the value above and below it.

I tried stepping through for a 1/6EV interval and used F3.36=APEX 336. (half between F3.5 and F3.2)

The apex value 336 was not in the lookup reference so it approximated between the upper and lower values on the table:
335 (F3.35) = 538
338 (F3.39) = 550
to come up with the value 542.

Very slick for something that's not supposed to be possible (camera only has 1/3EV aperture stops)

Re: SX 10 - Extended AV!!!
« Reply #12 on: 21 / October / 2009, 16:06:39 »
Hallo
Please take a look at the SX1 Firmware (201a) so we can test it on the SX1.

Smartkiller
Build for SX1 201a
This includes both the extended shutter speeds fix and extended AV fix.

The branches within the FW were identical to the SX10 except for the address references.


Re: SX 10 - Extended AV!!!
« Reply #13 on: 21 / October / 2009, 18:30:59 »
Hallo
Please take a look at the SX1 Firmware (201a) so we can test it on the SX1.

Smartkiller
Build for SX1 201a
This includes both the extended shutter speeds fix and extended AV fix.

The branches within the FW were identical to the SX10 except for the address references.


Hallo
The Cam starts and it works. I have made an AV Test with 11.31.

Smartkiller
SX200 IS Firm. 100c
SX1 IS Firm. 201a


Re: SX 10 - Extended AV!!!
« Reply #14 on: 21 / October / 2009, 18:56:26 »
Hallo
The Cam starts and it works. I have made an AV Test with 11.31.

Smartkiller
On the SX10, the AV override already worked up to F11 without the fix, it was beyond this up to 16 that didn't.  I would assume the same for the SX1.

-tgq

Re: SX 10 - Extended AV!!!
« Reply #15 on: 22 / October / 2009, 08:55:11 »
Hallo
The Cam starts and it works. I have made an AV Test with 11.31.

Smartkiller
On the SX10, the AV override already worked up to F11 without the fix, it was beyond this up to 16 that didn't.  I would assume the same for the SX1.

-tgq

What tests should I do? Make
What are the results I have to write?

The best would be a test script with the log file.

Smartkiller
SX200 IS Firm. 100c
SX1 IS Firm. 201a

Re: SX 10 - Extended AV!!!
« Reply #16 on: 22 / October / 2009, 10:11:09 »
What tests should I do? Make
What are the results I have to write?

The best would be a test script with the log file.

Smartkiller

There is nothing that a log file would help with.  What you need to do is to take a series of identical photos (preferably with a constant light source) at the different AV settings either using the overrides or through a script.
Make sure:
-manual mode
-fixed ISO (suggest 80)
-fixed shutter speed
-and that any other auto exposure settings are off or fixed
I would suggest only 3 would be needed: F8.0, F11, F16.
You can then look at the photos and compare the exposure.  If it works, each should be about half as bright as the previous (as these are full stop setting for aperture).  If F16 isn't working, it will have the same exposure as the F11 one.  (Ignore the exif, as it will still record it as F11 for anything above F11)
-tgq

Re: SX 10 - Extended AV!!!
« Reply #17 on: 23 / October / 2009, 12:51:24 »
I did a quick check of the capt_seq.c for the SX10 and SX1 builds.
The function address I've been using is:
SX10 103a: FF93B440

for SX10 version 101b and 102b, the corresponding address appears to be identical.  I'll have to look in the FW to verify if all addresses in the branches are identical as well.

For all other versions of the SX10 and SX1 the reference addresses are different, but the general structure looks the same.  Would likely only need updating of the reference addresses within th branches:

SX10 100c: FF93B398
SX10 101a: FF93B3A0
SX1  200h & 201a:FF94EAF0

Note that these are only a surface check within the available code in the capt_seq.c files to do a preliminary check on addresses and structure, not from firmware disassembly.

-tgq

I went through each of the FW dump disassemblies.  All structures are identical, just differing location addresses and reference addresses.  (The only two that are completely identical are the SX10 101b & 102b.)
I have inserted the changes into the respective capt_seq.c files for each of the 7 builds and can post if anyone is interested in compiling and testing.


Re: SX 10 - Extended AV!!!
« Reply #18 on: 13 / November / 2009, 22:33:25 »
I tried the av and tv override on sx10 103 and it's working fine...the av lower limit at full zoom is software limit but if you bypass it you have a lot of purple fringing...

Re: SX 10 - Extended AV!!!
« Reply #19 on: 13 / November / 2009, 23:19:48 »
I tried the av and tv override on sx10 103 and it's working fine...the av lower limit at full zoom is software limit but if you bypass it you have a lot of purple fringing...
How did you bypass the lower limit? The area I've been modifying seems to be as close as I can get to setting the aperture without getting actuall access to the aperture mechanism code.
I've tried doing it directly, but my results are the same results as 2.8 giving me the impression that F2.8 is the limit and represents true aperture full open (can't physically get larger). 

 

Related Topics