yet another personal build (archive)

[Jucifer]

Hm. Strange.
As a first aid you could try

Code:

<snip>
do
z=get_exp_count
until z>=y
<snip>

or even

Code:

<snip>
do
until get_exp_count>=y
<snip>

.

Does the get_exp_count freeze or skip, or is it uBASIC-related (= also try Lua)?
You can also try my yet another accurate intervalometer (uBASICversion, Lua version), and see if they have the same problem. (I don't think I have tested them that extensively.) (And (note to self) I really should update these scripts to be more "universal"...)
I can try to take a look at the problem some time this week, if you can't pinpoint the problem before me.
I don't have much enthusiasm now, as I just put together the first version of

e-build – now with scriptless remote support

changelog for 080603:
- added scriptless remote support from SDM (there's a new submenu for remote parameters).

(Remotes should work as with SDM. Don't know of every camera, there were some issues...)

[rawhead]

Hi Jucifer, thanks for the reply. fbonomi inquired about the trouble I'm having over at

Bracketing series (for example for timelapse HDR)

this thread, so I have some info there.  Basically, after 90 shots are taken, get_exp_count *both* jumps (ahead by 16,387 or thereabouts) *and* freezes (doesn't increase as the shots are taken).


I'll see if the first aid would work.

[rawhead]

All right, here's the result.  I simply used

Code:

until z>=y

this instead, and as you'd expect, the 19th cycle (i.e., the 91st shot) ended in just one shot, because z was way larger than y.  Good news is that with the 20th cycle, the script continued to go as expected, with both z and y reset to the much higher number.


This is workable for me, because I know that all I need to do is to remove every 91st shot before I batch process the images.


In the mean time, I would greatly appreciate it if you can look into the issue.  I will also try your script and see what happens, and get back to you.

[rawhead]

Little update. It turns out the problematic shot *isn't* always the 91st shot.

With the band-aided script, I ran a test of 100 cycles of 5 shot brackets, and the following were the problem areas (i.e., the Solo shot taken whenever Z went way high):

Shot #86, #187, #288, and #384


So at first it looked like the problem was going to be every 100 shots after the first one, but the 4th one threw that off as well.  I don't know what, if any, the pattern is here...

[Jucifer]

updated the e-build

changelog for 080604:
- get_exp_count now returns the (up to) 4-digit number as in the last file saved by cam. This behavior may be pretty much final, unless a better counter is found for every cam. NOTE: when scripting using get_exp_count, the possible counter reset should be taken into account.
(might be something along these lines:)

Code:

T=get_exp_count+N
if T>9999 then T=T-9999
press "shoot_half"
press "shoot_full"
do
until get_exp_count=T
release "shoot_full"

I noticed a dramatic speed decrease when taking shots in continuous mode with scripts, especially uBASIC. I'll have to look into this.

[rawhead]

K, Jucifer,

Just tried running a 300 cycle test and it worked beautifully.  Took exactly 900 photos without skipping a beat.  The only change I'd need to make is the counter check for >9999.


THANK YOU SO MUCH!

[fe50]

Hi "Jucifox" 
found your new branch (nice synonym) and asked myself whether LUA  is on the way to wonderland now...will there come some news reffering to this (i hope so) ?

[Jucifer]

Phox created a new branch for us.
It will become the ultimate chdk build.



But seriously, it will contain lots of tweaks and mods floating around.
Lua should definitely be one of those.

[PhyrePhoX]

yeah we probably are going to open a new thread about this, we dont have a real "strategy" as of now - yet. we will let you know, thats for sure.

[pethead]

How do I own branch on your SVN with my timestamp feature? or join to your branch?

[Jucifer]

I guess your timestamp feature could be included in our branch. Can you attach the patch (somew)here?

[Jucifer]

(the speed decrease was probably somehow related to chdk.cfg)

updated

[lukg]

Moved to the other thread!