testing sync among more cameras - General Discussion and Assistance - CHDK Forum

testing sync among more cameras

  • 121 Replies
  • 29589 Views
testing sync among more cameras
« on: 14 / July / 2012, 18:22:12 »
Advertisements
I have 60 cams to synch perfectly
I made really a lot of tests but
probably im missing something...

If I shoot a an lcd chronometer with 1/100th of seconds I have perfect sync, all cameras takes exatcly the same 1/100th of second (but maybe the chronometer doesnt display all the 1/100th...I dont know) but if I shoot with a flash triggered by the same triggers I use for the cam or if I use camera's flashes to check synchronism, I got problems...
Why that? I assume my sync precision is 1/100th of a second because of the chrono, but why with flash is so much worse? Is flash much faster than 1/100?

Does anybody know any other good way to test synchronism in the most accurate way?
(1/100 is ok but I woudl prefer 1/1000)

thx


Re: testing sync among more cameras
« Reply #1 on: 14 / July / 2012, 18:37:20 »
another thread on this ?
Ported :   A1200    SD940   G10    Powershot N    G16

Re: testing sync among more cameras
« Reply #2 on: 14 / July / 2012, 18:53:41 »
Not sure what you mean, this is another thread.

Do you mean a different title ?


*

Offline reyalp

  • ******
  • 12155
Re: testing sync among more cameras
« Reply #3 on: 14 / July / 2012, 18:58:31 »
The tool mentioned by Microfunguy in this post may be useful http://chdk.setepontos.com/index.php?topic=6231.msg81365#msg81365
Don't forget what the H stands for.


Re: testing sync among more cameras
« Reply #4 on: 14 / July / 2012, 18:58:47 »
I have 60 cams to synch perfectly

You cannot synch 60 cams to 1/1000 sec.

Quote
I have perfect sync, all cameras takes exatcly the same 1/100th of second

So, you have two definitions of 'perfect synch', 1/100sec and 1/1000sec !

Quote
if I shoot with a flash triggered by the same triggers I use for the cam

What does that mean ?

Quote
if I use camera's flashes to check synchronism, I got problems...

describe the problems.

Quote
Does anybody know any other good way to test synchronism in the most accurate way?


http://stereo.jpn.org/eng/sdm/crt_synch.htm

Re: testing sync among more cameras
« Reply #5 on: 15 / July / 2012, 06:51:24 »
The tool mentioned by Microfunguy in this post may be useful http://chdk.setepontos.com/index.php?topic=6231.msg81365#msg81365

That's very intersting,
Microfunguy I saw in that post that you say
Quote
A friend tested the S95 a few days ago :-
a good result except 25% of the time other tasks interrupt the capture sequence, I assume.
He has asked if we can disable interrupts until after capture.

That's probably the core problem I'm facing... was it possible, in that case, to disable interrupts until after capture?
thx

Re: testing sync among more cameras
« Reply #6 on: 15 / July / 2012, 07:26:48 »
was it possible, in that case, to disable interrupts until after capture?

 I did not try.

I would need reyalp or Philmoz to advise on how to do that.

However, even if that can be done I cannot believe you can force other tasks to 'yield' for the duration of image capture.

The capture task may also require to interact with other tasks.

Let me put it this way  ..  apart from CHDK/SDM how else could you synch 60 cameras to within about 1/100 sec ?

It is the best you can do (assuming even that can be achieved the majority of the time).


David

Re: testing sync among more cameras
« Reply #7 on: 15 / July / 2012, 07:51:49 »
you are right and I'm not complaining at all,
I'm very happy with chdk and very gratefull to all of you, really.

Since I need most precision possible I was just trying to understand
if my cams (sx130is) can be pushed further or if I'll have to consider
to buy other cams (I would prefer not... of course).

I'm of course available, in the case, to test whatever ideas/build/patch,etc...
come to your minds...

thx


Re: testing sync among more cameras
« Reply #8 on: 15 / July / 2012, 08:53:25 »
Has anyone determined for sure that the DryOS camera task management actually uses interrupts ?  (i.e context switching without the cooperation of the individual tasks rather that cooperative multi-tasking).
Did a simple little test to try and answer my own question.  While the "capt_seq_task" is waiting in a tight loop for the sync signal (via removal of USB power) the "blinker task" that I added to the G10 port continues to run and blink the power button LED briefly once per second.  So other things are clearly happening while waiting for "sync"

I then had a look at the sync code (both CHDK versions and SDM do essentially the same thing there).  While waiting for the USB voltage to go away,  the loop makes two system calls - one to get the tick time (for a timeout) and the other to get the USB power status.  Both of these are an opportunity for a context switch.

So I set the sync delay timeout value to the maximum and watched the power button LED during the sync delay period.  It was still blinking - even in a tight loop with no system calls,  task switching was a happening.

So my conclusion is that context switching under interrupt control is taking place.  Not good new for simple methods of getting better sync.

« Last Edit: 15 / July / 2012, 11:41:16 by waterwingz »
Ported :   A1200    SD940   G10    Powershot N    G16

Re: testing sync among more cameras
« Reply #9 on: 15 / July / 2012, 09:06:21 »
Has anyone determined for sure that the DryOS camera task management actually uses interrupts ?  (i.e context switching without the cooperation of the individual tasks rather that cooperative multi-tasking).

No, that is just me  using words loosely without knowing in detail what I am talking about  ... really.
I do not know of anyone who has investigated this.

Quote
  A better option is probably to do a deeper dive into the code and find somewhere closer to when the actual shutter release occurs.

I did this with the SX230HS and a VERY competent and knowlegeable tester.
It took ages and we 'reached the end of the line'.
As it happens, interrupts were involved.
For the sake of putting it on the record, I can post more details if you wish.

The tester sold his newly-bought SX230's and bought a pair of S95's instead.


 

Related Topics