DV Info Net

DV Info Net (https://www.dvinfo.net/forum/)
-   3D Stereoscopic Production & Delivery (https://www.dvinfo.net/forum/3d-stereoscopic-production-delivery/)
-   -   Using Component inputs for Real-Time 3D Preview (https://www.dvinfo.net/forum/3d-stereoscopic-production-delivery/144753-using-component-inputs-real-time-3d-preview.html)

Jack Aaronson February 27th, 2009 12:21 PM

Using Component inputs for Real-Time 3D Preview
So, I had an idea that didn't work. And I don't know enough to know why it didn't work, and I was hoping someone could explain it to me.

I thought that I could simulate a real-time 3D preview on an HD TV by doing the following.

I use two video cameras side-by-side to record in 3D (I post-process and create anaglyph, and it works perfectly).

I thought I would use the component outs from each video camera. I would take the RED component from the left and plug it into the HD TV, and take the BLUE AND GREEN components from the right camera and plug it into the HD TV.

Therefore, the TV is getting a signal that automatically merges the two images and colors them due to the input signals it is getting from each camera.

I really thought this would work, but it didn't. The red image floated across the screen like there was timing problem (like when you used to adjust old TV's so the image didn't migrate up the screen).

I imagine there is some kind of synchronization that each signal must have (that the red channel is missing because it is coming from a different channel)?

Just curious if others have tried this, and if there is any other way? I saw the posting about the new box that will do this, but at a ~$3,000 price point, I was hoping for a DIY method.


Alister Chapman February 27th, 2009 03:33 PM

You have two problems.

The first as you have guessed is sync. Unless the cameras are genlocked together or thier outputs are accurately synced in some other manner it will be impossible to merge the images.

The other problem is that the component output is not RGB but YUV so you cannot get a standard Anaglyph by simply mixing Y+U from one and V from the other.

What might work is to use a vision mixer with frame synchronisers (an old Panasonic MX type mixer) and use that to sync the cameras and then apply a colour effect to the cameras before half mixing them.

Better still take a look at the Stereoscopic multiplexer from 3dtv.
3dtv.at - Stereoscopic Multiplexer Overview
If you have two firewire cards in a reasonable PC you should be able to get really good real time optimised anaglyphs or with the right hardware feed a 3D monitor such as the Zalman Trimon.

Jack Aaronson March 1st, 2009 09:00 PM

Thanks, Alister. I'm Mac based, so the product you mentioned probably won't work. I will look into a vision mixer and report back if I have any success.


All times are GMT -6. The time now is 07:56 AM.

DV Info Net -- Real Names, Real People, Real Info!
1998-2021 The Digital Video Information Network