Hi zcream

with stereoscopy you should be very careful to keep the cameras genlocked as mush as possible. With CMOS this is more strict due to the rolling shutter. I don't know if the USB camera you propose can be genlocked at least at frame level, I suggest you make some research about it because conforming two images that are not synchronized is very difficult. Inter axis and convergence of the cameras are both important, and while you can play a bit with the inter axis in post, I think that your trick of keeping the camera steady and do the rest in post should be used more for reframing the two cameras for conforming reasons. Also cameras are 50% of your job. A good rig (with mirror...) and a conforming software are equally very important. Depending on your budget, I would suggest to hire an equipment right for your project. A couple of Scarlet and a Genius stereo rig are a good combination to hire. A recent firmware update lets you control the focus of both the lens with just one remote controller. I know you are missing the fun of writing your own encoder and experimenting with cameras, but stereoscopy has enough a lot of inherent problems without adding all the technological stuff that can goes wrong.

@kino The 8-pin GPIO Hirose connector has Sync in and Sync out. The cameras can be genlocked. I confirmed this before buying the cameras. The cameras are small enough to be used side-by-side There will not be any toe-in - I will film parallel. Convergence will be decided in post. I will keep a strict depth budget. Nothing for 3m (10ft) in front of the camera.

@fatpig large DOF is an advantage in 3D. You want to see everything in frame - ideally nothing should be out of focus.

Hi zcream happy to know you can sync the cameras. Convergence cannot be decided in post, it would require rebuilding part of the scene, only inter axis can be simulated in post. The factors to decide the inter axis distance is not only related to the closer object, but mainly on how much "depth" you would like to give to your scene. However I really suggest you to add a convergence feature to your rig. Sensors are small, so you probably will use a wide lens which helps you to keep a fixed focus. About encoding: I read you are looking for a lossy compression. If you are able to stand a raw format (maybe with some simple compression as RLE) I think it would be important to get a good stereoscopy. Compressing two different images could bring a different result that create disparity in the stereoscopic view. If you are interested in jpeg based comression, take a look at http://www.libjpeg-turbo.org/. It is optimized to use CPU vector instruction sets. This can be enough for your goal, considering also that jpeg encoding and decoding can only be in part parallelized using a GPU. This would avoid the need for a big GPU, just letting you to use a simpler embedded one.

@zcream

All you need it raw compression. Good and fast one. I think you know the answer.

A summary of where we are today - in terms of recording options.

8-bit 4:2:0 - QuickSync

8-bit 4:2:2 or 4:4:4 - libjpeg-turbo, AMV2MT or Utvideo or ffv1

10-bit 4:2:2 - Cineform High or ffv1 PIX_FMT_YUV422P10

RAW - Cineform RAW or Compressed DNG

I got the cameras. I used 2x 2 way Macro Shot Focusing Focus Rail Sliders from ebay ~ 10$ each. Now I have a sliding calibrated 3d rig that allows the cameras to move from IA of 32mm to 84mm. I reckon this covers interiors and exteriors quite well. I have no need for 3d macro shots or hyper-stereo. For comparison there is an ipod touch in front.

I'm quite looking forward to seeing this progress. I looked at the flea cameras in hopes of building a rig capable of some higher frame rates. alas, life got in the way. I hope it works out for you.

The elphel project has code to create DNG - and the people at the Field project came up with a makeDNG utility.

http://openendedgroup.com/field/images/makeDNG.zip

raw2dng - Magic Lantern as well - http://www.magiclantern.fm/forum/index.php?topic=5404.0

ATM, I am trying to get my head around this.

As for storing the image data - I am currently gravitating towards usage of lz4 compression of the 12-bit raw image data. I am not sure if I want to pay for another Cineform license (my current license is on OSX). The Point Grey SDK has a simple example program that writes out the raw image data. It would be an easy job to link with the lz4 library - http://code.google.com/p/lz4/source/checkout and save it as compressed RAW.

I will also build a batch CLI that will take this raw compressed data after filming and convert it to a compressed DNG after the video has been saved. The speed and scalability of lz4 is phenomenal!!

If anyone here has used the Adobe DNG API, please help me! Their documentation is not the best, and there is little in terms of example programs.

Virtualdub has problems recording to Cineform. I cannot capture no matter what - the camera does get recognized if I register the DirectShow driver.

Next option - tried to build the Flycapture SDK. The GUI is useless.

Some notes :-

Install

Win 7.1 SDK .Net 4.0 VC++ Express 2010

File -> New Project and select “Win32 Console Application”.

“Win32 Application Wizard” -> Next ...next page select “Empty Project” under the “Additional options”.

“Configuration Manager” -> “Active solution platform” -> “New”: x64 “Copy settings from” “Win32”.

“Configuration Properties” -> “General”. In there we select as “Platform Toolset” “Windows7.1SDK”

"Options" > "Tools" > "Projects and Solutions" > "VC++ Directories" > "Include files" C:\WinDDK\7600.16385.1\inc\mfc42;C:\WinDDK\7600.16385.1\inc\atl71;

"Options" > "Tools" > "Projects and Solutions" > "VC++ Directories" > "Library files" C:\WinDDK\7600.16385.1\lib\mfc\ia64;:\WinDDK\7600.16385.1\lib\atl\ia64;

Linker -> General c:\Program Files\Point Grey Research\FlyCapture2\lib64; C/C++ -> General c:\Program Files\Point Grey Research\FlyCapture2\include;

“Configuration Properties > General > Output directory” $(SolutionDir)$(PlatformName)\$(Configuration).

“Configuration Properties > General> Intermediate Directory” $(PlatformName)\$(Configuration).

“Configuration Properties > Linker > Debugging > Generate Program Database File” $(ProjectDir)$(PlatformName)\$(ConfigurationName)\$(TargetName).pdb.

Compile the project. If there are compile errors you may need to add the $(PlatformName) macro to other settings as well. For example, if there are user library files the macro may need to be added to “Configuration Properties > Linker > General > Additional Library Directories”.

I renamed the FlyCapture2.lib to FlyCapture2_Windows7.1SDK.lib as the build needs this file.

Status - Hello World compiled with FlyCapture SDK.

Just kidding, but on a serious note, I thought 3d was kind of becoming a thing of the past...again...?

I doubt that it is "thing of the past". It is great tech, especially for amateurs. In fact all idea to ditch it in consumer cameras came from pros and from sponsored reviewers after companies realized that you need much more powerful LSI, 2x sensors and 2x lenses and still can't make 2x price :-)

Turns out that one can only access RAW12 by storing a sequence of RAW images within FlyCapture2 or paying for a Streampix 6 license with 1-7 cameras and Cineform RAW module.

Streampix 6 with these options is worse than a ripoff, and no other version can access USB3 cameras.

So we move to CF again - my previous approach failed as I was trying to modify the DPX2CF converter within the SDK.

Turns out David Newman used RawSpeed some years ago as a frontend to a RAW2CF utility. http://cineform.blogspot.com.au/2010/11/camera-raw-to-cineform-raw.html

However, the source file is linked to the old CF site, and I cannot locate it.

http://cineform.com/downloads-other/rawspeedDLLsrc101028.zip

If I can see this source, its a simple matter to integrate this within FlyCapture2 for a free USB3 to Cineform RAW solution. No one seems to post on this thread - but if you are one of the few who downloaded this, can you post this again ?

I have also asked Cineform for this.