Video thumbnail for vimeo video Canon 50D Couldn't Shoot Video in 2008. Now in 2013 It Has Reached 1080P RAW - No Film SchoolMagic Lantern progresses more every single day, and this one is particularly impressive considering the camera couldn't even shoot video when it was released in 2008. A little less than a month ago, those working on ML managed to get the Canon 50D shooting not just video, but RAW video, and now it's possible to get full 1080p, even on some slower cards.

Andy600 posted this video to the ML forum, here's what he said on Vimeo about the test (1080p at 3X zoom possible for 6-10 seconds):

This is a quick test of the EOS 50D recording 1920x1080 24p in Magic Lantern 14bit raw video.

Download the original to see in full HD or check out some frame grabs here:

Shot on Canon EOS 50D (Yes, the one that came out before the 5d MkII and the one that didn't have video LOL)

Nikon 50mm F2 (Vintage 1974) - Shot at F5.6 - F11 (The lens is effectively 240mm because of APS-C crop + 3x video crop)
Transcend 600x 16GB CF card

Thanks to a1ex from Magic Lantern for the buffer sorting hack that made this possible on a relatively slow CF card.

Raw video files converted to Cinema DNG with raw2cdng app (Thanks Chmee!)

Quick color balance in Davinci Resolve using Black Magic Film settings and Black Magic Film LUT.

Output to H.264, edited in Premier Pro

This particular video was made possible by some experimenting a1ex has been doing involving something he's calling Variable Buffering (thanks to cinema5D for the heads-up on this):

If you ever looked in the comments from raw_rec.c, you have noticed that I've stated a little goal: 1920x1080 on 1000x cards (of course, 5D3 at 24p). Goal achieved and exceeded - even got reports of 1920x1280 continuous.

During the last few days I took a closer look at the buffering strategy. While it was near-optimal for continuous recording (large contiguous chunks = faster write speed), there was (and still is) room for improvement for those cases when you want to push the recording past the sustained write speed, and squeeze as many frames as possible.

So, I've designed a new buffering strategy (I'll call it variable buffering)

Even though this isn't using the full sensor, you'd have to be pretty cynical not think that more developments are coming. It's not going to work any miracles, but this particular development will provide slight improvements when you need to squeeze just a little more write speed out of a card.

Things are definitely getting more and more interesting, and the sky is the limit when we've got cameras like the Canon 50D now recording full 1080.