supplierdeeply

extra long exposures (split from: Hardcore modding on a A530, thanks to CHDK!

  • 183 Replies
  • 44736 Views
*

Offline reyalp

  • ******
  • 11583
Re: extra long exposures (split from: Hardcore modding on a A530, thanks to CHDK!
« Reply #170 on: 06 / November / 2009, 15:53:02 »
Advertisements
Or maybe this is hardware limitation?
No, most likely just needs someone to spend the time/effort to implement it.
Don't forget what the H stands for.

*

Offline reyalp

  • ******
  • 11583
Don't forget what the H stands for.

*

Offline reyalp

  • ******
  • 11583
Re: extra long exposures (split from: Hardcore modding on a A530, thanks to CHDK!
« Reply #172 on: 12 / November / 2009, 22:06:14 »
In changeset 837 I've added extra long exposure for G9. Version 100g is tested, 100d and 100i are not.
Don't forget what the H stands for.

*

Offline reyalp

  • ******
  • 11583
Re: extra long exposures (split from: Hardcore modding on a A530, thanks to CHDK!
« Reply #173 on: 31 / December / 2017, 00:33:31 »
8 year bump  :haha

@ftm discovered that sx700 does > 60 sec exposures (up to at least 10 minutes) without exp_drv task. I verified this is also true on sx710.

Since exp_drv is a fairly large chunk of code, I think it makes sense to leave it out on the cameras that don't need it. CAM_EXT_TV_RANGE still need to be defined to allow longer overrides in the CHDK UI. I already had g7x set up this way, since the stock firmware allows up to 250 sec.

IIRC some cameras also needed exp_drv for short exposures.

I went back and tested elph130, and it does indeed max out at 64 sec if exp_drv is disabled.

It would be interesting to find out if this is digic 6 specific, or just cameras after a certain dryos rev, or something else.

I'll post a test script in a bit.

Edit:
Looking at the camera apex2us code, I think stock sx700, g7x, sx710, g16 and M3 all allow up to 1024s (=APEX96 -960 = 0xFFFFFC40)

sx280 appears to still have the 64s (=APEX96 -576 = 0xFFFFFDC0)  limit

Edit 2:
If I'm interpreting correctly, it looks the the change happened around DryOS r52, with some cameras like sx280 and elph130 having the lower limit, and others like sx510is and g16 having the higher.

Implementing exp_drv would allow ~2000s instead of 1024, but I'm not really convinced this has much value.
« Last Edit: 31 / December / 2017, 01:55:45 by reyalp »
Don't forget what the H stands for.


*

Offline blackhole

  • *****
  • 642
  • A590IS 101b
    • Planetary astrophotography
Re: extra long exposures (split from: Hardcore modding on a A530, thanks to CHDK!
« Reply #174 on: 31 / December / 2017, 01:58:29 »
8 year bump  :haha

@ftm discovered that sx700 does > 60 sec exposures (up to at least 10 minutes) without exp_drv task. I verified this is also true on sx710.

 
It would be interesting to find out if this is digic 6 specific, or just cameras after a certain dryos rev, or something else.

I've seen it before on sx530 and SX410. Both cameras are DIGIC 4+, DryOS r55 + p6.
For this reason both cameras have no exp_drv task.

*

Offline reyalp

  • ******
  • 11583
Re: extra long exposures (split from: Hardcore modding on a A530, thanks to CHDK!
« Reply #175 on: 31 / December / 2017, 15:02:33 »
I've seen it before on sx530 and SX410. Both cameras are DIGIC 4+, DryOS r55 + p6.
For this reason both cameras have no exp_drv task.
Thanks for confirming, that makes sense.

I had also forgotten than CAM_EXT_TV_RANGE is on by default since support was added for essentially all cameras a few years back, so ports will show the extended Tv override options unless specifically undef'd

The override UI does go to 2048
Don't forget what the H stands for.

My SD1200 has a strange limit,
tv96 -893 / 0xFFFFFC83
time 631s/10:31

after that, it acts like a  short exposure, although it takes as long to take. It's too bad people stop at 10 minutes as a nice round number, please test to 11 minutes instead.
Check the resulting picture for the purple amp glow, if it's not there, it's resetting.

I wouldn't assume it's useless either; I see no reason for 34 minute (2048s) exposures not to be useful. At my 10 minute limit, the standard deviation was 97, meaning I have room to take a good pic, in the sense that a thumbnail sized image would have a SD of 9 (not bad, like normal ISO200 pic), after downsizing to 1/10 with averaging. I'd be happy to see a clear nebula at that size.
« Last Edit: 26 / January / 2018, 23:08:43 by jmac698 »

*

Offline reyalp

  • ******
  • 11583
My SD1200 has a strange limit,
tv96 -893 / 0xFFFFFC83
time 631s/10:31
The SD1200 port has exp_drv implemented, so this shouldn't be related to the limits I posted about above.
Quote
after that, it acts like a  short exposure, although it takes as long to take.
That sounds like something is wrapping / overflowing somewhere else in the firmware code.

Quote
I wouldn't assume it's useless either; I see no reason for 34 minute (2048s) exposures not to be useful. At my 10 minute limit, the standard deviation was 97
Would one 30 minute exposure really give you better results than stacking 3x10 minute, or 10x3 for that matter?
Don't forget what the H stands for.


Short answer, there's some logic to longer exposures, it's often debated.
https://www.dpreview.com/forums/thread/3760903

Long answer
http://www.clarkvision.com/articles/astrophotography.and.exposure/
https://www.cloudynights.com/articles/cat/articles/astrophotography/finding-the-optimal-sub-frame-exposure-r1571

Theoretically, the longer the exposure the fainter you can detect, however light pollution limits the total time.  I wasn't limiting myself just to astrophotography though, I'd like to see landscapes lit only by the sky or say the glowing plankton in the ocean and some other ideas, where I just want to see the faintest of signals. You can take that to an extreme with the experiments done to image the inherent faint glow of the human body itself, our 'aura'.
https://www.livescience.com/7799-strange-humans-glow-visible-light.html

*

Offline reyalp

  • ******
  • 11583
https://www.cloudynights.com/articles/cat/articles/astrophotography/finding-the-optimal-sub-frame-exposure-r1571
From the article
Quote
As you can see, it recommends between a 5 and 10 minute optimal sub-frame (60 minutes total in the front, 360 minutes total in the back) except in cases of shooting short total time in extreme light pollution or shooting very long total time in pristine skies with cameras that have exceptionally low camera noise.
CHDK cameras do not have "exceptionally low noise", at least not without serious hardware hacking to cool the sensor.

Quote
Theoretically, the longer the exposure the fainter you can detect, however light pollution limits the total time.
A sum of shorter is exposures is a longer exposure, it just has more read noise. For an extreme example, see http://www.rolfolsenastrophotography.com/Astrophotography/Antennae-Extreme-Deep-Field/i-QZwp6NF

On older CHDK cameras (somewhere up to ~2012), amp glow will start to seriously eat into your signal with just a few minutes of exposure. Later cameras are OK for longer, but the 10 minute test exposure I did on sx710 had a whole lot of hot pixels and generally didn't look great.

Quote
You can take that to an extreme with the experiments done to image the inherent faint glow of the human body itself, our 'aura'.
https://www.livescience.com/7799-strange-humans-glow-visible-light.html
This seems only applicable to extreme low light cameras, not relevant to CHDK cameras.

In practical terms, I'm not seeing a ~10-15 minute max exposure being a major handicap for CHDK cameras. I'm open to being convinced otherwise, but I'd really want to see a real world scenario that makes sense with these cameras.
Don't forget what the H stands for.

 

Related Topics