CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD - page 29 - Creative Uses of CHDK - CHDK Forum  
« previous next »
Pages: 1 ... 24 25 26 27 28 [29] 30 31 32 33 34 ... 71   Go Down

CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD

  • 704 Replies
  • 156455 Views
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #280 on: 10 / December / 2012, 13:45:09 »
Advertisements
QUICK UPDATE on ZOOM MOTOR ENCODER DUMMY

I've got the ENCODER PI POWER dummy state machine functional.  It is the first step and will of course manage the encoder phototransistor drive oscillators.  So far the good news is that there is still no adverse impact all the other dummies.  That is, IS, IRIS, FOCUS and ZOOM hardware remains disconnected and is being simulated OK with the microcontroller (ZOOM MOTOR is still running of course).  So upcoming tests on the oscillators should be possible w/o hardware revival -- I hope we don't get into a revival trap.

That was a bit challenging because of some concurrent interrupts between the MOTION PI POWER and ENCODER PI POWER had to be properly handled.  Now the tricky part ... the oscillators // will there be enough MIPS?  This won't be quick!
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #281 on: 12 / December / 2012, 18:38:46 »
OK FOLKS ... THIS IS LOOKING GOOD .....

Not out of the woods just yet, but I'd say it's 1 yard to go with 3 downs remaining :). The ZOOM ENCODER PIs are hanging out of the motor and are obstructed, with uC electrical override.  It was a huge fight for the last MIPS available, so at full 8 MIPS maxed out we have a solution // beautiful!  The motor spindle is the only item left connected and still runs, so that's next.  Because there I did not code inertia, there is enough slop for FW to accept the 420 pulses, so long as the relative phases and lags between both PIs are correct.

That means out of ~50 tests just completed, PLAY->REC, REC->PLAY,  REC->PDN PUP->PLAY->REC all work perfectly.  This camera is beginning to look very empty.  You will know it's all good when you see a movie advertised with the C code MECHA engine // hopefully within a week.

@srsa_4c
I had to keep your lens error override solution in place.  It gives maximum assurance because there were mistakes that caused the severe lens error, and with AC disconnect and both your solutions, not once did I have reconnect Canon HW to revive the camera.  So the solution is self-reviving too, so far.  That alone is *very* good news because we now know that revival can be done by the microcontroller and we don't need to plug in hardware.  I will only test the worst case (pull the plug in REC) once all is done.  So this is beginning to beg for the auto-full-zoom solution ... if you are still interested.
« Last Edit: 12 / December / 2012, 18:44:33 by SticK »
Logged
*

Offline Microfunguy

  • ****
  • 4037
    • StereoData Maker
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #282 on: 12 / December / 2012, 20:50:13 »
Deadline Xmas  :)

Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #283 on: 12 / December / 2012, 21:14:40 »
I'm shootin' for the same!
Logged
*

Offline srsa_4c

  • ******
  • 4451
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #284 on: 13 / December / 2012, 18:43:35 »
Glad to hear the progress of your hw replacement.

Quote from: SticK on 12 / December / 2012, 18:38:46
auto-full-zoom solution
This version should prevent lens distortion correction in jpegs, but won't affect live view. I hope this is sufficient. As the file name shows, the RTC related code is not enabled, so only try it when the dummy is functioning perfectly.
https://subversion.assembla.com/svn/chdk-s1.bin/files/DISKBOOT_s90_101a_disableisdriveerror_nodistcorr.zip

I have attached the current state of the patch I'm using, in case somebody is curious. The patch is not cleaned up, it includes remnants of previous trials.
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #285 on: 13 / December / 2012, 20:03:16 »
@srsa_4c
Thanks for this.  The way things look right now, the safe solution is to stay with your your lens error RTC override solution, but not sure yet.  Would it be possible to have both versions?  I would most likely start testing the one with the RTC first if you can compile a version.

Quote
won't affect live view.
On the S90 I'm pretty sure liveview is barrel corrected.  For a single random shot of a specimen region, it wouldn't matter much because the barrel-corrected live view is a bit smaller than the full frame CCD (RAW and now JPG with your new solution).  However, if possible, live view matching (disabling barrel correction) would be very useful, probably indispensable, for aligning the CNC stage in controlling tiling tolerances in step-and-repeat mosaic imaging.  What could be the chances?  Any thoughts?
Logged
*

Offline srsa_4c

  • ******
  • 4451
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #286 on: 13 / December / 2012, 21:08:12 »
Quote from: SticK on 13 / December / 2012, 20:03:16
Quote
won't affect live view.
On the S90 I'm pretty sure liveview is barrel corrected.  For a single random shot of a specimen region, it wouldn't matter much because the barrel-corrected live view is a bit smaller than the full frame CCD (RAW and now JPG with your new solution).  However, if possible, live view matching (disabling barrel correction) would be very useful, probably indispensable, for aligning the CNC stage in controlling tiling tolerances in step-and-repeat mosaic imaging.  What could be the chances?  Any thoughts?
Fortunately the Canon developers implemented control over distortion correction for live view.
I believe that
poke(0x5338, 0)
will switch the correction off for live view, and
poke(0x5338, 1)
will switch it on. Try it and report back. You only need this, the below explanation is for others.

The event procedures to control this are
LiveImageTool.Dister (takes one parameter), and
ChangeDistortionSetForLiveImage (takes also one parameter).

Quote
I would most likely start testing the one with the RTC first if you can compile a version.
https://subversion.assembla.com/svn/chdk-s1.bin/files/DISKBOOT_s90_101a_disableisdriveerror_nodistcorr_no_startup_lens_check.ZIP
« Last Edit: 13 / December / 2012, 21:15:10 by srsa_4c »
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #287 on: 13 / December / 2012, 22:15:10 »
@srsa_4c
Quote
Try it and report back.
This is a grand effort ... I can't try it at present unfortunately // camera is down for some minor HW mods // may be over weekend I can get to it but will do as soon as I have the chance.  More on this later.
Logged
*

Offline SticK

  • *****
  • 779
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #288 on: 14 / December / 2012, 20:34:10 »
BIG NEWS:  THE MECHA IS COMPLETELY DISCONNECTED and the S90 ROCKS :D

Totally silent ... just the subtle click of the shutter solenoid.  Not only, but also ... it is completely self-reviving even after POWER FAIL in REC mode !!!   This means PHASE IV of the instrumentation project is about to start (after Christmas :xmas). 

The microcontroller module is so small it can be hidden inside battery compartment!  :).

@srsa_4c
Re: DISKBOOT_s90_101a_disableisdriveerror_nodistcorr_no_startup_lens_check
I won't be able to validate this on the 101a for at least good time.  Could you issue a version for the always-accessible test bed S90_100c?
Quote
I believe that
poke(0x5338, 0)
will switch the correction off for live view, and
poke(0x5338, 1)
will switch it on. Try it and report back. You only need this, the below explanation is for others.
Same for this ... is the address the same on the 100c?

I feel as though I just came out of a dungeon on a rack for 3 months.  Some stress tests coming up over the next couple of days, but because the code is refined and important window ranges established, I expect all is a GO and downhill from now.  If no problems, stay tuned for photos, a movie and code.

Logged
*

Offline Microfunguy

  • ****
  • 4037
    • StereoData Maker
Re: CHDKPTP: S90 Primary Focal Plane Configuration - hacking out the CCD
« Reply #289 on: 14 / December / 2012, 22:04:03 »
CONGRATULATIONS TO ALL CONCERNED !

This is a WORLD First.

I am tempted to buy an S90 just to try it.

3 am and time for bed.


David
Logged
Pages: 1 ... 24 25 26 27 28 [29] 30 31 32 33 34 ... 71   Go Up
« previous next »
 

Related Topics