GH1 Stereoscopic Rig Collaboration

crunchy said:
One solution is by modifying capacitor connected to Quartz oscillator.
Click to expand...

I recall matching two CCTV cameras that drifted, even with blackburst sync, by wrapping copper foil tape around the capacitor with a frequency counter connected to the oscillator. It did work to slightly tweak the frequency. I imagine this trick is only effective with certain caps. In the CCTV camera case, they were electrolytics. The upside is that it's non-destructive and can be removed easily.
 
I'll try to make measurements with NTSC signals when I get some more time. Will report about my findings.

David, have you tried to use pre-focus signal before shooting and then measure disparity with Panasonic lenses?

I agree. Electronics is always better than plain switch. :thumbsup:

Modifying the capacitor is also quite non-destructive. IMO, the largest "destruction" is to open camera's case (Pandora's box).
 
crunchy said:
have you tried to use pre-focus signal before shooting and then measure disparity with Panasonic lenses?
Click to expand...

Not sure exactly what you mean. Are you suggesting that I measure the VSYNC before and after the AF-half-push w/ Panny lenses? I can try that.
 
While we're waiting to decide where to host our more detailed discussions and tests, I'm going to post the details for a simple mechanical controller again. I'm getting LOTS of emails requesting this now, for some reason. Apologies to the seasoned veterans who are WAY beyond this!

I've attached a simple diagram here that shows a basic, mechanically switched sync "controller". This device DOES work well with GH1s and manual lenses, but, may not be the ideal solution. One of its biggest faults is that the manual switches may "bounce" when closed which can cause sync problems. It does work for me. Your milage may vary! Builder beware!

Usage is simple:

1) turn GH1s on (with controller power off)
2) switch controller power on
3) press "shutter release" - cameras will begin recording
4) at end of take, press "shutter release" to stop recording
5) power both cameras off using the GH1 power switches (not the controller power)

fe1f6402c7de07affc9546ca4dcf2e.jpg
 
There is any advated in using this wired switch instead of a wireless remote let say two:

http://gb.suntekstore.com/Wireless-Remote-Shutter-for-Panasonic-GF1-GH1-FZ30K-DMW__RS1.html

Are really necessary two hacked GH13 what about two GH2?

thanks



David Cole said:
While we're waiting to decide where to host our more detailed discussions and tests, I'm going to post the details for a simple mechanical controller again. I'm getting LOTS of emails requesting this now, for some reason. Apologies to the seasoned veterans who are WAY beyond this!

I've attached a simple diagram here that shows a basic, mechanically switched sync "controller". This device DOES work well with GH1s and manual lenses, but, may not be the ideal solution. One of its biggest faults is that the manual switches may "bounce" when closed which can cause sync problems. It does work for me. Your milage may vary! Builder beware!

Usage is simple:

1) turn GH1s on (with controller power off)
2) switch controller power on
3) press "shutter release" - cameras will begin recording
4) at end of take, press "shutter release" to stop recording
5) power both cameras off using the GH1 power switches (not the controller power)

fe1f6402c7de07affc9546ca4dcf2e.jpg
Click to expand...
 
Last edited:
David Cole said:
Not sure exactly what you mean. Are you suggesting that I measure the VSYNC before and after the AF-half-push w/ Panny lenses? I can try that.
Click to expand...
VSYNC should be the same before or after using focus. What I am suggesting is to apply auto-focus on both cameras and after that (or even after locking AF and AE) apply "shoot" signal.
 
crunchy said:
VSYNC should be the same before or after using focus. What I am suggesting is to apply auto-focus on both cameras and after that (or even after locking AF and AE) apply "shoot" signal.
Click to expand...

Let me see if I'm understanding you...

1) Put cameras in autofocus mode,
2) Switch on using sync controller,
3) Push the AF button on sync controller (5K ohms to remote jack),
4) Optional - manually press AF/AE lock button on each camera,
5) Begin recording with shutter release button on sync controller.

Is this what you mean? If so, that's exactly the steps that I took for the Panny lens tests that I reported last week - with 25% of the test runs exhibiting terribly blown sync.

-- David
 
Thanks for the advice Crunchy. When I was at college I signed up for an electronics course, but it got cancelled because not enough people signed up. Still kicks my ass occasionally.

re. wireless remotes, I thought about it, but their timing may not be identical. It would be especially useful for massive hypers, eg. 10m+ apart (but then the trigger delay may be even worse, unless the trigger is equidistant between the bodies).

re. monitoring, are you all ignoring me? Let me ask you direct, David, Crunchy, how are you monitoring your 3D recordings??

And exciting news - Vitaliy has sucessfully dumped the GH2 ROM (by desoldering it)! He's already working on reversing the encryption to enable future patches. The GH2 also stores its video setup params in RAM, not ROM like the GH1. That makes modifying them on-the-fly easier. Details in this thread on his site, everyone, post there that you're interested in a 3D sync solution to get it on his radar!
 
David Cole said:
1) Put cameras in autofocus mode,
2) Switch on using sync controller,
3) Push the AF button on sync controller (5K ohms to remote jack),
4) Optional - manually press AF/AE lock button on each camera,
5) Begin recording with shutter release button on sync controller.

Is this what you mean? If so, that's exactly the steps that I took for the Panny lens tests that I reported last week - with 25% of the test runs exhibiting terribly blown sync.
Click to expand...

Between points 2 and 3 there should be several seconds of pause (let's say more than 5s). Point 5 should be very precisely timed. Are both signals pulled down exactly at the same instant? In my microprocessor solution I have even precisely timed the instant when pulling down the signals within video cycle.

Today I was again shooting and the results were excellent. Reliability was about 90% (slightly more).
 
g.l said:
re. monitoring, are you all ignoring me? Let me ask you direct, David, Crunchy, how are you monitoring your 3D recordings??
Click to expand...

I don't need to monitor it for stereoscopic purposes. I just measure or guess the distance to the nearest subject, check lens focal distance and calculate the required stereo base in my head. And I never make larger mistakes. Just today, among 32 clips, only one had exaggerated depth (it was the first shot which was made without calculating stereo base - it was test shot).
 
I DO like to have a monitor for setting up the rig, determining divergence budget, converging the shot, etc. Here's the poor-man's approach that I've been using quite successfully : http://3dvision-blog.com/dual-camera-rig-live-stereoscopic-3d-preview-and-alignment-method/

The NTSC > USB adapters are $10 on Amazon! I use this setup with an Acer passive-display 3D laptop. Works great! NTSC isn't high enough resolution for focusing or fine convergence, but, it's ok for gross setup.
 
crunchy said:
Between points 2 and 3 there should be several seconds of pause (let's say more than 5s). Point 5 should be very precisely timed. Are both signals pulled down exactly at the same instant? In my microprocessor solution I have even precisely timed the instant when pulling down the signals within video cycle.
Click to expand...

I'll retest - but - I always wait for the SD card busy icon to disappear before I push the AF button (step 3). Regarding the timing of the 5K ohm signal - I'm using a mechanical switch. It could bounce. It works fine with manual lenses, however.

You may have something there with the shutter release timing. How do you do this (e.g. on the left camera blank, on the right camera blank, or independently for each camera)?
 
crunchy said:
...Reliability was about 90% (slightly more).
Click to expand...

Sorry to be redundant in my questioning, but, to be sure I fully understand you....

In the 10% of the shots where you are out of sync, do you determine this in advance of actually shooting by VSYNC? If that is the case then you really never have any recorded footage that is out of sync, because you power-cycle the cameras before recording, correct?

Thanks. I'd REALLY like to get Panny lenses working - there are so many great lens options! Even the little 14-42 makes a great picture and it's very light weight for a hand-held rig.
 
David Cole said:
In the 10% of the shots where you are out of sync, do you determine this in advance of actually shooting by VSYNC? If that is the case then you really never have any recorded footage that is out of sync, because you power-cycle the cameras before recording, correct?
Click to expand...

No, I cannot determine it in advance. I start shooting and if it shows that video signals are not in synch, I stop shooting immediately. Afterwards, when transferring the footage to the PC, I delete those "10%" of bad clips. How do I know that they are bad? They are very short.
 
David Cole said:
You may have something there with the shutter release timing. How do you do this (e.g. on the left camera blank, on the right camera blank, or independently for each camera)?
Click to expand...

I am doing it on the left camera blank (just for now), but the firmware can synch it independently. Up to now I am more than satisfied with results, so I am not doing extensive tests now. I'll just have to finish the project and try it to realise with AVRs as well (namely, I would like to learn how to do it with AVR Studio). Here is a picture of a testing prototype version (larger button is obsolete):

Parallel:
3DSLR_Master_01.jpg

XView:
3DSLR_Master_XView_01.jpg

On the main screen you have battery voltage, frame-rate, synch (in ms), power on/off indication, Focus/shutter released and Time-lapse active indication.
 
Nice controller! What CPU are you using for prototyping? I'm using Arduino - which isn't economical if anyone has a desire to commercialize the controller (not me!).

-- David
 
Just bought a GF1 kit just to get a cheap 20mm pancake lens. Hadn't seen the GF1 in the flesh before, it's gorgeous, aluminium body feels great & very well designed. When I first handled the GH2 body, it felt like a cheap toy! (the pancake lenses still do, though they're great). Saying that I love it now, they went all out for image quality and usability - but the GF1 build really is much nicer. Finished my circuit design and will build shortly, then onto sync tests ... (hey, why can't I have line breaks anymore?? have the admins changed the forum software?)
 
You must log in or register to reply here.
Back
Top