sync'ing 3D video - page 8 - Creative Uses of CHDK - CHDK Forum supplierdeeply

sync'ing 3D video

  • 103 Replies
  • 37731 Views
Re: sync'ing 3D video
« Reply #70 on: 01 / July / 2013, 13:57:32 »
Advertisements
I should mention that I have also been informally discussing this with two other interested parties.


50% duty cycle?

For my S95 in 1280x720 mode, 181:180 for one waveform and 181:181 for the other.

Quote
Should be no problem for even the cheapest ( sub $10) Aurdino clone.

I was thinking of something so small and with no external components required that it was just part of the USB cables.
Maybe an ATtiny ?


Quote
I wouldn't bother with interrupts for something this simple.  Just poll both signals in a tight loop and  count the time between when one goes from 0 -> 1 and the other does the same.

If synchronised. that time would be less than 100 microseconds.

Quote
But more interesting to me is automating this.

It is extremely simple.
In response to momentary press of a button, the controller sets USB to +5V in usual way.
When synch is achieved, the controller sets USB to 0V

Initially, I will assume drift is negligible for the duration of a typical clip.

Re: sync'ing 3D video
« Reply #71 on: 01 / July / 2013, 19:24:11 »
I should mention that I have also been informally discussing this with two other interested parties.
By all means,  please feel free to share those discussions in this forum.


Quote
I was thinking of something so small and with no external components required that it was just part of the USB cables.
Still need a power source - one of those tiny LiPo batteries used for the little RC helicopters is the smallest battery I've found.

Quote
Maybe an ATtiny ?
Not as familiar with how the I/O lines can be configured on these but getting a "through hole" package might make "dead bug" construction easier.  A quick google tells me it should be fine.

Quote
Quote
I wouldn't bother with interrupts for something this simple.  Just poll both signals in a tight loop and  count the time between when one goes from 0 -> 1 and the other does the same.
If synchronised. that time would be less than 100 microseconds.
Lots of time - I would not even bother coding in assembler.  C will be just fine.


Quote
It is extremely simple.  In response to momentary press of a button, the controller sets USB to +5V in usual way.
When synch is achieved, the controller sets USB to 0V Initially, I will assume drift is negligible for the duration of a typical clip.
Maybe I'm missing something here but why use a button push?  And I'd still go with a PWM communication - it will be easier to tune the feedback loop that way.
Ported :   A1200    SD940   G10    Powershot N    G16

Re: sync'ing 3D video
« Reply #72 on: 01 / July / 2013, 21:07:59 »
please feel free to share those discussions in this forum.
I mean discussions regarding if they would be willing to produce the device commercially if it works.

Anyway, let us see how the outdoor tests with my analogue version go.

The weak point is the mounting of the photo detector, especially on the S95 where the AF led is so near the lens barrel.

Today I should receive a faster phototransistor, it is just 2mm x 1mm in size.

Re: sync'ing 3D video
« Reply #73 on: 01 / July / 2013, 21:20:49 »
please feel free to share those discussions in this forum.
I mean discussions regarding if they would be willing to produce the device commercially if it works.
Ah,  your friend James at gentles I assume.  Good idea.

Quote
The weak point is the mounting of the photo detector, especially on the S95 where the AF led is so near the lens barrel. Today I should receive a faster phototransistor, it is just 2mm x 1mm in size.
I'll admit I'm totally lost here. Somewhere in the hardware-hacker stream of consciousnesses posts you seem to have switched from using sync from the video output to actually generating a sync signal?  If so,  it should be possible to feed that into the other camera and skip the external microcontroller completely. (too slow)

I'll go back and see if I can figure out where I missed that step ....

Update :
In video mode, with SDM, while the USB switch is pressed I now flash the AF led in phase with LiveView period counter.
and
At 640x480 I get a 30Hz square wave and at 1280x720 I get a 24Hz square wave.
I need to make a second photo sensor assembly tomorrow.
Now, by half pressing on one camera I could increase the LiveView period slightly at intervals for one cycle and maybe drift the two waveforms into synch.
So does just half pressing move the sync or are you watching that press in a script and doing something ?
« Last Edit: 01 / July / 2013, 21:31:34 by waterwingz »
Ported :   A1200    SD940   G10    Powershot N    G16


Re: sync'ing 3D video
« Reply #74 on: 02 / July / 2013, 05:33:54 »
you seem to have switched from using sync from the video output to actually generating a sync signal?

The circuitry would be a little more complex, the screen is blanked while the A/V cable is attached (though that can probably be software overriden, tedious for so many cameras) and only gives an indication (maybe) of synch .. does not allow you to affect it.

Quote
So does just half pressing move the sync


Yes.

Re: sync'ing 3D video
« Reply #75 on: 02 / July / 2013, 10:30:47 »
Quote
So does just half pressing move the sync
Yes.
Very interesting - so all we need is a short script that does click "half_press" while video is running and the signal from the uP says there is no sync?
Ported :   A1200    SD940   G10    Powershot N    G16

Re: sync'ing 3D video
« Reply #76 on: 02 / July / 2013, 11:54:25 »
so all we need is a short script that does click "half_press"

The current implementation uses USB signal.

Quote
while video is running

No, that has never been possible and maybe never will be.
It does not seem to be a requirement, at least with my S95 tests.

Anyway, I have the new phototransistors, have to decide how to mount them so that position is not critical.

*

Offline ahull

  • *****
  • 634
Re: sync'ing 3D video
« Reply #77 on: 02 / July / 2013, 19:17:52 »
Quote from: Microfunguy
Anyway, I have the new phototransistors, have to decide how to mount them so that position is not critical.

I don't suppose you have some short lengths of plastic fibre optic light guide, that way you can mount the photo transistors on the pcb, and simply route the fibre optics from there to the LED. 

« Last Edit: 02 / July / 2013, 19:29:55 by ahull »


Re: sync'ing 3D video
« Reply #78 on: 02 / July / 2013, 19:25:39 »
Well, I have just finished making some nice housings for the sensors but they are not really commercially suitable, requires some precision manual work.

Fibre optic would not be suitable for same reason.

The new phototransistor gives a very clean, stable waveform, not 50:50 mark-space though.

I will investigate if that is a real situation or caused by my code.

*

Offline ahull

  • *****
  • 634
Re: sync'ing 3D video
« Reply #79 on: 02 / July / 2013, 19:36:36 »
Well, I have just finished making some nice housings for the sensors but they are not really commercially suitable, requires some precision manual work.

I presume clipping the sensor to the camera is the issue.

Quote
Fibre optic would not be suitable for same reason.

For fibre-optics I was thinking of something like a cheap toslink cable, but you would still have a mounting issue.

Quote
The new phototransistor gives a very clean, stable waveform, not 50:50 mark-space though.

I will investigate if that is a real situation or caused by my code.

If the signal is derived from the vsync or hsync signal, I wouldn't expect it to be 50:50, I would expect a pulse, and a long leadout, repeated at regular intervals
« Last Edit: 02 / July / 2013, 19:41:39 by ahull »

 

Related Topics


SimplePortal 2.3.6 © 2008-2014, SimplePortal