My DIY cinematography camera project at DVinfo.net

Daniel Lipats · August 10th, 2009, 10:59 PM

Just thought I would share my progress in case anybody is interested.

This is a software DVR for the Sumix SMX-12A2C head, and the brains of my DIY camera project. I borrowed a few ideas I liked from a number of cameras and the design is still evolving.

The software is multithreaded and has a number of safety features to guarantee reliable video acquisition. Output is Adobe Digital Negative (DNG) 1080p 12bit 24p at about ~100 megabytes a second. This is captured to a hotswap interchangeable 1TB raid 0 magazine.

Video sequence names are determined by the digital slate accessed from the [slate] button.

Adobe products provide the backbone for the post-production RAW workflow.

http://dl.getdropbox.com/u/305574/alpha.jpg
This is a screenshot with the emulation enabled. It's in a development mode I use to avoid messing with the camera head.

I'm considering rewriting the interface in OpenGL and take advantage of hardware acceleration... There may be no advantage but I figure the only way to know for sure is try.

August 10th, 2009, 10:59 PM	#1
Daniel Lipats Major Player Join Date: Feb 2006 Posts: 260	My DIY cinematography camera project Just thought I would share my progress in case anybody is interested. This is a software DVR for the Sumix SMX-12A2C head, and the brains of my DIY camera project. I borrowed a few ideas I liked from a number of cameras and the design is still evolving. The software is multithreaded and has a number of safety features to guarantee reliable video acquisition. Output is Adobe Digital Negative (DNG) 1080p 12bit 24p at about ~100 megabytes a second. This is captured to a hotswap interchangeable 1TB raid 0 magazine. Video sequence names are determined by the digital slate accessed from the [slate] button. Adobe products provide the backbone for the post-production RAW workflow. http://dl.getdropbox.com/u/305574/alpha.jpg This is a screenshot with the emulation enabled. It's in a development mode I use to avoid messing with the camera head. I'm considering rewriting the interface in OpenGL and take advantage of hardware acceleration... There may be no advantage but I figure the only way to know for sure is try.