chdk zoom vs canon ring zoom accuracy - Creative Uses of CHDK - CHDK Forum
« previous next »
Pages: [1]   Go Down

chdk zoom vs canon ring zoom accuracy

  • 8 Replies
  • 2386 Views
*

Offline Mlapse

  • *****
  • 584
  • S95 S110
chdk zoom vs canon ring zoom accuracy
« on: 15 / October / 2021, 15:27:55 »
Advertisements
I was just talking with another long time user about zoom and chdk and that a lua script can use the preset zoom functions, but that i found it wasn't always accurate between shoots.
the same can happens if you use buttons to alter the zoom.

however, on cams that have a ring around the lens set to zoom (steps) in CF (..., 28, 35, 50 ,etc). You can then save it in a [C]ustom setting.

I have the experience with the s95 and s110 it will always (re)boot at exactly the same zoom distance.
(on some types of shoots with a slightly different barrel distortion)

I do not know if this is something that is known or that it is only my experience with the limited models i have at my disposal.

but after discussing we thought sharing this might help somebody with finding better hooks for G, S and some SX, M models?
« Last Edit: 15 / October / 2021, 15:42:37 by Mlapse »
Logged
frustration is a key ingredient in progress
*

Offline c_joerg

  • *****
  • 1251
Re: chdk zoom vs canon ring zoom accuracy
« Reply #1 on: 16 / October / 2021, 02:28:02 »
Quote from: Mlapse on 15 / October / 2021, 15:27:55
I was just talking with another long time user about zoom and chdk and that a lua script can use the preset zoom functions, but that i found it wasn't always accurate between shoots.
Have you ever tried to read the value with get_zoom after setting? Were they the same then?
With Set_zoom and get_zoom I never have any differences. With the focus, yes.

Quote from: Mlapse on 15 / October / 2021, 15:27:55
however, on cams that have a ring around the lens set to zoom (steps) in CF (..., 28, 35, 50 ,etc). You can then save it in a [C]ustom setting.
I think the camera does the same internally. She will use Set_zoom.

Quote from: Mlapse on 15 / October / 2021, 15:27:55

(on some types of shoots with a slightly different barrel distortion)
I don't think it's because of the zoom. Slight changes in focus are likely to cause the deviation.

I would write a script that does a rebbot a few times with set_zoom and then in C afterwards. I would focus on one on infinity and one on the smallest distance.
I don't think you will see any differences then.
Logged
M100 100a, M3 121a, 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 Mlapse

  • *****
  • 584
  • S95 S110
Re: chdk zoom vs canon ring zoom accuracy
« Reply #2 on: 16 / October / 2021, 03:16:47 »
no, i've never tried to read out the values.....most of that is still beyond me, i'm still struggeling with the lua syntax due to a lot of depricated examples i find while searching for solutions. maybe i've grown a bit after the winter, that is when i usually find time for scripting.

i always assumed that it was the same value if you set it by ring or via buttons..but is it really?

since i use most cams the same way i never looked at it again after i found that setting zoom to the ring and then saving it solved it for me.
it might be a timing problem.
and maybe since reyalp created the set_clock that I use at reboot the problem is no longer there.
« Last Edit: 16 / October / 2021, 03:25:19 by Mlapse »
Logged
frustration is a key ingredient in progress
*

Offline c_joerg

  • *****
  • 1251
Re: chdk zoom vs canon ring zoom accuracy
« Reply #3 on: 16 / October / 2021, 04:16:54 »
Quote from: Mlapse on 16 / October / 2021, 03:16:47
i always assumed that it was the same value if you set it by ring or via buttons..but is it really?

You can have the zoom step displayed in
CHDK Settings => OSD Settings => Miscellaneous values
Then you could compare ring and buttons...
Logged
M100 100a, M3 121a, 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 Mlapse

  • *****
  • 584
  • S95 S110
Re: chdk zoom vs canon ring zoom accuracy
« Reply #4 on: 16 / October / 2021, 05:05:58 »
S95: with the toggle switch on my 3rd try i created 10 distinct zoom steps, and i can repeat that now by very short taps to the switch
however usually i only get 9..i thought that was the maximum of the s95. (28 31 34 39 44 49 59 69 84 105)
what i do see with the toggle and not with the ring is values in between, like 41 when moving between 39 and 44.
but haven't got it to stay stuck on 41 yet.
a few times a tap did not lead to a change, but that might be me tapping too lightly.

and with the ring i only have 5 to choose from. (28, 34, 49, 84, 105) i do see different values during zoom, but too quick changing to read while the time it zooms per step is longer.
they appear to be not halfway but closer to the setting it's going for.
with a lua script that sets zoom per 10% (and calls a zoom step) i see the next value before it zooms back to the set value, so not 41 when moving between 39 and 44, but 34 or 49 before it zooms back to the chosen value.

well, al least it looks a bit different when using a script or the toggle switch.
after an hour of testing different ways to zoom i have no added info, just a thought: maybe canon uses a non linear motor drive for the zoom.
« Last Edit: 16 / October / 2021, 06:11:17 by Mlapse »
Logged
frustration is a key ingredient in progress
*

Offline c_joerg

  • *****
  • 1251
Re: chdk zoom vs canon ring zoom accuracy
« Reply #5 on: 16 / October / 2021, 07:14:50 »
You shouldn't concentrate on the focal length (EFL) but on the zoom level (X).
Different levels sometimes give the same focal length. The focal lengths are coarser than the steps.
Logged
M100 100a, M3 121a, 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 Mlapse

  • *****
  • 584
  • S95 S110
Re: chdk zoom vs canon ring zoom accuracy
« Reply #6 on: 16 / October / 2021, 08:14:06 »
right, set to x:
ring: (I use the Canon 'official' figures, so 35 instead of 34
105=9
85=8
50=5
35=2
28=0

with button:
105=9
85=8
...=7
...=6
50=5
...=4
...=3
35=2
...=1
28=0

btw. the lens itself is a 6.0-22.5mm with 1/1.7inch sensor, where 50mm=11mm
« Last Edit: 16 / October / 2021, 08:23:54 by Mlapse »
Logged
frustration is a key ingredient in progress
*

Offline c_joerg

  • *****
  • 1251
Re: chdk zoom vs canon ring zoom accuracy
« Reply #7 on: 16 / October / 2021, 08:56:32 »
Now i understand. Only 9 steps on S95
S110 => 24 bis 120 mm => 120 steps
G1x => 28-112 mm => 100 steps
So these cams has more steps than mm focal length
Logged
M100 100a, M3 121a, 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 Mlapse

  • *****
  • 584
  • S95 S110
Re: chdk zoom vs canon ring zoom accuracy
« Reply #8 on: 16 / October / 2021, 09:32:42 »
Yes, the S95 has only 10 steps(0 to 9), while the S110 has 121 steps(0 to 120), but that was a lot to write out ;)
the S95 has 5 'presets' under the ring, the S110 has 7. (24, 105=>100 and 120)

I'd say a lot more than one step per mm, since your g1x lens isn't 24-112, but 12.5-62.5 if i'm right, and the s110 with 5.2-26mm is doing even less.
so all in all very accurate, i mean 0.2mm play on the total of the lens would be able to create a 1 step difference on the s110.

I am interested in some other models than the s95 and s110 that have a stepped ring setting, if somebody has a cam that supports that and wants to post them, thanks.
So what are the fixed steps of the?
S110:
24=step0
28=step4
35=step13
50=step33
85=step79
100=step98
120=step120
« Last Edit: 16 / October / 2021, 16:55:21 by Mlapse »
Logged
frustration is a key ingredient in progress
Pages: [1]   Go Up
« previous next »
 

Related Topics


SimplePortal © 2008-2014, SimplePortal