Night timelapse (was Re: considerations for SX260 external power supply for time lapse situations) - page 4 - Script Writing - CHDK Forum  

Night timelapse (was Re: considerations for SX260 external power supply for time lapse situations)

  • 52 Replies
  • 18030 Views
*

Offline reyalp

  • ******
  • 13457
Advertisements
The readme says:
==meter, meter96==
Average value of the metered area: meter is the raw sensor value, *96 gives this as an APEX96 value where the target exposure is 0. Negative values indicate an underexposure, positive is overexposure.

But I see logging that has a positive meter value with a negative meter96 value; what does that mean?
Sorry this wasn't clear. The part about negative values only refers to meter96. meter is in the same values as raw pixels, which can never be negative since they are proportional to the number of photons that hit the sensor.

meter96 is derived from meter such that 0 is approximately the exposure canon auto exposure would use and each change of 96 is one "stop" of exposure. If you cut the exposure time in half (e.g. going from 1/30 to 1/60), meter96 will go down by 96. If you double it (1/30 to 1/15), it will go up by 96.

Quote
Why does exposure time get shorter when I increase the allowable overexposed fraction?
The over exposure value limits only affect the exposure if the scene is over exposed. In the log from post #27 above, the under exposure was simply due to hitting your chosen shutter and ISO limits.

Without seeing the resulting video, I don't really see any problems. In the attached plot, you can meter96 mostly follows the bv_ev_shift curve until tv96 (shutter) and sv96 (ISO) hit their limits. Then it drops very rapidly, because there is no more exposure control available.

If there is something else you think is wrong, you'll need to describe it more or show an example.

Except for very start over_weight is always 0, so it has no impact on the exposure. under_weight is 0 because you have under exposure protection turned off, but this makes no difference because the shutter and ISO limits are hit.

When you are trying to understand what the script did, the _weight columns tell you what had the most influence on the exposure values. Of course, once hard limits on shutter and ISO are hit, the weights are irrelevant.

Finally, ~25000 shots is by far the most I've seen in a single run with this script. Congrats :)
Don't forget what the H stands for.

*

Offline c_joerg

  • *****
  • 1157
because the shutter and ISO limits are hit.

I tried also to understand what’s going on… but I have also not figured out what’s going on… something to learn about for me ;)

The script goes only to an exposure time from 0.8s. At this time d_ev is always 32. That means the script wants to change exposure to longer values. Something limits it.
tv96_long_limit=-415 that means that tv_max = 20000 (20s)?

Finally, ~25000 shots is by far the most I've seen in a single run with this script. Congrats :)

Well I was never over 10000....
« Last Edit: 01 / April / 2016, 02:27:28 by c_joerg »
M100 100a, M3 101a, G9x II (1.00c), 2*G1x (101a,100e), S110 (103a), SX50 (100c), SX230 (101a), S45,
Flickr https://www.flickr.com/photos/136329431@N06/albums
YouTube https://www.youtube.com/channel/UCrTH0tHy9OYTVDzWIvXEMlw/videos?shelf_id=0&view=0&sort=dd

*

Offline udo

  • ***
  • 123
Quote
Why does exposure time get shorter when I increase the allowable overexposed fraction?
The over exposure value limits only affect the exposure if the scene is over exposed. In the log from post #27 above, the under exposure was simply due to hitting your chosen shutter and ISO limits.
Oh..
I have
#ui_tv_max_s1k=20000 "Max Tv Sec/1000" as I tried the max to be 20 seconds (works OK for SX260 etc for star photography).
#ui_sv_max_mkt=400 "Max ISO" as this is still fairly noiseless.
The ISO goes to 400 but the exposure does not go to 20 seconds.
What am I missing here?

Finally, ~25000 shots is by far the most I've seen in a single run with this script. Congrats :)
That was done using a slightly modded (as posted elsewhere on this forum) Mi.com 10.000mAH powerbank.
The S110 runs off a direct connection to the Li-Ion cells using the fake battery method.
With a bigger card I could get more out of the battery.

To be sure about the exposure limit issue  I tested in a very dark room and the exposure time did not go up as experienced before.
bv96 went to -1255.

So which of my settings is actually limiting the exposure time from going beyond 0.8 seconds to 20 seconds?
« Last Edit: 01 / April / 2016, 11:41:38 by udo »

*

Offline reyalp

  • ******
  • 13457
Oh..
I have
#ui_tv_max_s1k=20000 "Max Tv Sec/1000" as I tried the max to be 20 seconds (works OK for SX260 etc for star photography).
#ui_sv_max_mkt=400 "Max ISO" as this is still fairly noiseless.
The ISO goes to 400 but the exposure does not go to 20 seconds.
What am I missing here?
You are right, this looks like a bug in the script.

In the log, tv96_long_limit=-415 => 20s, but it clearly stops at 33 => ~0.78s

I don't have time to dig into the code right, but now I suspect it has something to do with the ISO adj Tv Sec/1000 value, were if the ISO range runs out before the TV range, it the script doesn't correctly adjust tv beyond that point.
The ISO adj value in your run is 288 (=1/8s) and there are are a bit over 3 stops of ISO (80, 200, 400) so a bit over 1/2 sec looks like it fits.

If this is right, then setting ISO adj value to within 3 stops of your Tv limit should work around it until I can fix the code.

Note if your shutter speed is 20s, your interval will be much slower at night. From the sleep time in log you have about 0.6s margin left, so you could go up to about 1.4s while keeping the same interval.
Don't forget what the H stands for.


*

Offline udo

  • ***
  • 123
I don't have time to dig into the code right, but now I suspect it has something to do with the ISO adj Tv Sec/1000 value, were if the ISO range runs out before the TV range, it the script doesn't correctly adjust tv beyond that point.
The ISO adj value in your run is 288 (=1/8s) and there are are a bit over 3 stops of ISO (80, 200, 400) so a bit over 1/2 sec looks like it fits.

If this is right, then setting ISO adj value to within 3 stops of your Tv limit should work around it until I can fix the code.
To be sure, you mean this setting:
#ui_tv_sv_adj_s1k=125 "ISO adj Tv Sec/1000"
?

I understand the value should go up, but how much? I can test to confirm your theory.
I tested with values 499 and 500 and tv goes up to 3.15 now. Not enough but progress nonetheless.
« Last Edit: 02 / April / 2016, 01:23:44 by udo »

*

Offline reyalp

  • ******
  • 13457
I understand the value should go up, but how much? I can test to confirm your theory.[/font]
If you keep the ISO limit at 400, "ISO adj Tv Sec/1000" should be 3 stops (or less) from "Max Tv Sec/1000"

So if you "Max Tv Sec/1000" = 20000,  "ISO adj Tv Sec/1000" should be around 2500.

edit:
This may be conservative, since exposure is split between ISO and TV in this range.

I am pretty sure this is the problem, and should have a fix for it fairly soon.
« Last Edit: 02 / April / 2016, 01:54:04 by reyalp »
Don't forget what the H stands for.

*

Offline c_joerg

  • *****
  • 1157
I never had this problem because i set always ui_tv_max_s1k  = ui_tv_sv_adj_s1k. That means
that the ISO increases only when I come to the maximum exposure time.

Did you switch the noise reduction off? Otherwise you got an interval on long exposure times from > 2*exposure time.
M100 100a, M3 101a, G9x II (1.00c), 2*G1x (101a,100e), S110 (103a), SX50 (100c), SX230 (101a), S45,
Flickr https://www.flickr.com/photos/136329431@N06/albums
YouTube https://www.youtube.com/channel/UCrTH0tHy9OYTVDzWIvXEMlw/videos?shelf_id=0&view=0&sort=dd

*

Offline udo

  • ***
  • 123
I understand the value should go up, but how much? I can test to confirm your theory.
If you keep the ISO limit at 400, "ISO adj Tv Sec/1000" should be 3 stops (or less) from "Max Tv Sec/1000"

So if you "Max Tv Sec/1000" = 20000,  "ISO adj Tv Sec/1000" should be around 2500.

edit:
This may be conservative, since exposure is split between ISO and TV in this range.

I am pretty sure this is the problem, and should have a fix for it fairly soon.
I tested 2.5 seconds to confirm.
Exposure time goes up a lot but not yet to the max... So that matches your 'conservative'?
Any other ideas why?
Does this help fix the bug?
Please let me know what I can to to help this fixing. (testing, logging..)
A patch would be welcome!

I never had this problem because i set always ui_tv_max_s1k  = ui_tv_sv_adj_s1k. That means
that the ISO increases only when I come to the maximum exposure time.

Did you switch the noise reduction off? Otherwise you got an interval on long exposure times from > 2*exposure time.
I did not switch off the noise reduction as the frame rate at night is not too important.
« Last Edit: 02 / April / 2016, 06:45:30 by udo »


*

Offline c_joerg

  • *****
  • 1157
ISO adjust was at 2.1s...
Who do you made the run?
Meter96 made big jumps from -450 to plus 150 (6EV steps) and back to -260.
That is the reason why it don’t increase anymore...
M100 100a, M3 101a, G9x II (1.00c), 2*G1x (101a,100e), S110 (103a), SX50 (100c), SX230 (101a), S45,
Flickr https://www.flickr.com/photos/136329431@N06/albums
YouTube https://www.youtube.com/channel/UCrTH0tHy9OYTVDzWIvXEMlw/videos?shelf_id=0&view=0&sort=dd

*

Offline udo

  • ***
  • 123
ISO adjust was at 2.1s...
rawopint.lua has:
#ui_tv_sv_adj_s1k=2500 "ISO adj Tv Sec/1000"

Who do you made the run?
Meter96 made big jumps from -450 to plus 150 (6EV steps) and back to -260.
That is the reason why it don’t increase anymore...
Yes the light in the dark chamber can change wildly. (light on/off, door open/close etc)
 14.461714 was the longest exposure thus far.

 

Related Topics