SX530 Internal light leakage - General Help and Assistance on using CHDK stable releases - CHDK Forum

SX530 Internal light leakage

  • 10 Replies
  • 527 Views
*

Offline Davo

  • ***
  • 174
SX530 Internal light leakage
« on: 20 / January / 2020, 05:22:17 »
Advertisements
Hi,
I have a SX530HS which I use for astronomy with exposures of up to 5 minutes under CHDK. Each image has a blue pattern on one side as shown in the attached image. This image was taken in a totally dark environment with the LCD off , lens cap on and black tape over the focus light. It is 300 seconds exposure and ISO 1600.


I also shoot 'darks' which are also 300s but with the shutter closed and the same pattern shows up on them. The darks help me subtract the pattern partially from the star shots but it would be good to eliminate it completely.


Any idea what causes this and can it be disabled? I read on another forum that Sony A7s have a similar problem caused by opto-electric sensors bleeding non visible light onto the sensor.

*

Offline srsa_4c

  • ******
  • 4230
Re: SX530 Internal light leakage
« Reply #1 on: 20 / January / 2020, 14:03:54 »
Any idea what causes this and can it be disabled? I read on another forum that Sony A7s have a similar problem caused by opto-electric sensors bleeding non visible light onto the sensor.
I don't know what the cause might be, but the shape of the anomaly does look suspicious (light rays?).
There are a few event procedures that are related to opto parts (there might be more, these are the obvious ones):
EnableFocusPiCircuit
DisableFocusPiCircuit
EnableZoomPiCircuit
DisableZoomPiCircuit
EnableZoomEncoderCircuit
DisableZoomEncoderCircuit

They are registered by Mecha.Create .
But. I seem to recall trying them with no success (i.e. they have no lasting influence on the camera's operation). They are probably meant to be used in factory mode. The underlying firmware functions are referenced from normal firmware routines, so it's possible that a "disabled" photointerrupter or encoder is quickly re-enabled by firmware.
Trying them from certain capt_seq hooks may or may not be more successful, but I can imagine that they can crash the camera due to their complexity.

*

Offline reyalp

  • ******
  • 12595
Re: SX530 Internal light leakage
« Reply #2 on: 20 / January / 2020, 16:24:28 »
You can find previous discussion of this phenomena if you search google with site:chdk.setepontos.com amp glow
The light ray pattern is quite common.

At least one user found a hardware solution on a much older camera: https://chdk.setepontos.com/index.php?topic=8949.0

Whether this is specifically the amplifier or other electronics on modern cameras is unclear. Canon significantly reduced this issue in most cameras around 2012 (it used to be very bad even in one minute exposures) so I suspect they may have already implemented something like the amp off hack.

Don't forget what the H stands for.

*

Offline Davo

  • ***
  • 174
Re: SX530 Internal light leakage
« Reply #3 on: 21 / January / 2020, 08:54:25 »
Thanks both for your help. I did not realise it was so common while reading thru similar forum posts. What I did find was this link https://astronomy-imaging-camera.com/tutorials/what-is-amp-glow.html which describes amp glow in detail. It looks like there are 2 causes - heat and near infrared light. Heat glow is more diffuse while NIR can give specific patterns. I even found one example which looks just like mine (attached) so I guess I probably have NIR glow.
I will try to disable the NIR source using the eventproc suggestion above. If that doesn't work maybe a simple internal baffle would do the trick. Will keep you updated.
ps The article above suggests turning off dark optimisation in Deep Sky Stacker. I checked and it was already off.


*

Offline Davo

  • ***
  • 174
Re: SX530 Internal light leakage
« Reply #4 on: 22 / January / 2020, 10:06:00 »
I tried to disable these calls in turn.
1. DisableFocusPiCircuit
2. DisableZoomPiCircuit
3. DisableZoomEncoderCircuit
For each I then shot a long exposure with manual focus and then enabled them again. 1. gave the same pattern as my original post above while 2. and 3. gave the image attached.
I then tried to disable all 3 and got the same result as 1. Something is changing but not what I was hoping for.
Here is the code I used


Code: [Select]

--[[
********************************
based on this script
https://chdk.fandom.com/wiki/Meteor_Intervalometer_with_Dark_Frame_Management
original script by fudgey modified by waterwingz


CHDK 1.3 r3250 or higher
@chdk_version 1.3
Lua native calls required
********************************
@title NIRblock
@param         s Lights Tv (seconds)
   @default    s 300
   @range       s 1 99999


--]]


shutter = s*1000000


local bi=get_buildinfo()
if bi.os=="vxworks" then
    closeproc="CloseMShutter"
    openproc="OpenMShutter"
    if (call_event_proc("InitializeAdjustmentFunction") == -1) then
      error("InitAdjFunc failed")
    end
  elseif bi.os=="dryos" then
    closeproc="CloseMechaShutter"
    openproc="OpenMechaShutter"
    if (call_event_proc("Mecha.Create") == -1) then
      if (call_event_proc("MechaRegisterEventProcedure") == -1) then
        error("Mecha.Create failed")
      end
    end     
  else
    error("Unknown OS:" .. bi.os)
  end
 
  -- close mechanical shutter
  function closeshutter()
    if (call_event_proc(closeproc) == -1) then
      print("closeshutter failed")
    end
  end
 
  -- open mechanical shutter
  function openshutter()
    if (call_event_proc(openproc) == -1) then
      print("openshutter failed")
    end
  end
 
    -- disable FocusPiCircuit
  function disablefocuspi()
    if (call_event_proc("DisableFocusPiCircuit") == -1) then
      print("disablefocuspi failed")
    end
  end
 
     -- enable FocusPiCircuit
  function enablefocuspi()
    if (call_event_proc("EnableFocusPiCircuit") == -1) then
      print("enablefocuspi failed")
    end
  end
 
      -- disable ZoomPiCircuit
  function disablezoompi()
    if (call_event_proc("DisableZoomPiCircuit") == -1) then
      print("disablezoompi failed")
    end
  end
 
     -- enable ZoomPiCircuit
  function enablezoompi()
    if (call_event_proc("EnableZoomPiCircuit") == -1) then
      print("enablezoompi failed")
    end
  end
 
       -- disable ZoomEncoderCircuit
  function disablezoomencoder()
    if (call_event_proc("DisableZoomEncoderCircuit") == -1) then
      print("disablezoomencoder failed")
    end
  end
 
     -- enable ZoomEncoderCircuit
  function enablezoomencoder()
    if (call_event_proc("EnableZoomEncoderCircuit") == -1) then
      print("enablezoomencoder failed")
    end
  end
 


   --FocusPiCircuit
   closeshutter()
   print("Shooting dark#1")
   set_tv96_direct(usec_to_tv96(shutter))
   disablefocuspi()
   shoot()
   enablefocuspi()
   openshutter()
   
   --ZoomPiCircuit
   closeshutter()
   print("Shooting dark#2")
   set_tv96_direct(usec_to_tv96(shutter))
   disablezoompi()
   shoot()
   enablezoompi()
   openshutter()
   
   --ZoomEncoderCircuit
   closeshutter()
   print("Shooting dark#3")
   set_tv96_direct(usec_to_tv96(shutter))
   disablezoomencoder()
   shoot()
   enablezoomencoder()
   openshutter()
   
   -- disable all 3
   closeshutter()
   print("Shooting dark#4")
   set_tv96_direct(usec_to_tv96(shutter))
   disablefocuspi()
   disablezoompi()
   disablezoomencoder()
   shoot()
   enablefocuspi()
   enablezoompi()
   enablezoomencoder()
   openshutter()
« Last Edit: 22 / January / 2020, 10:07:42 by Davo »

*

Offline reyalp

  • ******
  • 12595
Re: SX530 Internal light leakage
« Reply #5 on: 22 / January / 2020, 13:05:37 »
For each I then shot a long exposure with manual focus and then enabled them again. 1. gave the same pattern as my original post above while 2. and 3. gave the image attached.
I then tried to disable all 3 and got the same result as 1. Something is changing but not what I was hoping for.
Are these images camera jpegs, or from raw?
If they are from raw, are you comparing them at equal levels, or is there some kind of stretch involved?
Don't forget what the H stands for.

*

Offline Davo

  • ***
  • 174
Re: SX530 Internal light leakage
« Reply #6 on: 22 / January / 2020, 13:43:38 »
Are these images camera jpegs, or from raw?
They are all jpegs reduced to 20% of the size of the original jpegs from the camera in Paint.NET. No other processing was done.

*

Offline srsa_4c

  • ******
  • 4230
Re: SX530 Internal light leakage
« Reply #7 on: 22 / January / 2020, 16:35:17 »
Something is changing but not what I was hoping for.
I'm also getting similar results on sx280. I also tried
- different aperture
- switching IS off
- switching the LCD off
with no apparent change in resulting image. I wonder what else could radiate IR light in there.


*

Offline reyalp

  • ******
  • 12595
Re: SX530 Internal light leakage
« Reply #8 on: 22 / January / 2020, 17:04:56 »
They are all jpegs reduced to 20% of the size of the original jpegs from the camera in Paint.NET. No other processing was done.
Very odd that the second image is so much brighter then. I guess messing with the eventprocs makes something misbehave, but I have no idea what.

It might be interesting to see what happens with the settings that produce the second image using a more normal length exposure, like 1s or something.

edit:
https://www.ifixit.com/Guide/Canon+PowerShot+SX530+HS+Lens+Replacement/80533 has images of the internals. The last one (https://d3nevzfk7ii3be.cloudfront.net/igi/T4iKNRAyLOpgrQs5.huge) I think shows the main ribbon going to the sensor. It looks like the motherboard is well away from the sensor stuff, and the LCD display is separated from it by a metal plate. Hard to say what's on the other side though.
« Last Edit: 22 / January / 2020, 17:13:58 by reyalp »
Don't forget what the H stands for.

*

Offline blackhole

  • *****
  • 765
  • A590IS 101b
    • Planetary astrophotography
Re: SX530 Internal light leakage
« Reply #9 on: 22 / January / 2020, 17:38:45 »
 
 Heating the surrounding electronics can cause NIR radiation.
Here is what some of the best experts in the area say.   
   https://astronomy-imaging-camera.com/tutorials/what-is-amp-glow.html

Poslano sa mog GT-I9301I koristeći Tapatalk


 

Related Topics