313 posts
  • Author had a Free File of the Month
  • Referred between 1 and 9 users
  • Sold between 1 000 and 5 000 dollars
  • Bought between 1 and 9 items
  • Exclusive Author
  • Has been a member for 2-3 years
  • United Kingdom
Panda-Bear says


i heard a secret rumor…that ML is actually a covert ‘Black Ops’ Canon research & development team. And if their mission was to kill off Nikon and it’s d800 range, then ML is seriously looking at succeeding.

:-)

Yesterday submited first footage processed from Full HD 14 bit RAW.

Best Regards, Andrey
i bet the post workflow was a nightmare. converting the .dng files.

Would you say the video is remarkably improved over normal H264 ?

can’t wait to see this.

1471 posts
  • Sold between 10 000 and 50 000 dollars
  • Author had a File in an Envato Bundle
  • Most Wanted Bounty Winner
  • Russian Federation
  • Has been a member for 2-3 years
  • Exclusive Author
  • Referred between 10 and 49 users
  • Bought between 10 and 49 items
  • Grew a moustache for the Envato Movember competition
AndrVlad says

i bet the post workflow was a nightmare. converting the .dng files.

Would you say the video is remarkably improved over normal H264 ?

can’t wait to see this.
Having 14 bit RAW video makes easier to shot – no need to setup white balance, brightness, saturation, etc (same as photo shot in RAW). Post Workflow is easy in comparison with how hard sometimes to process H.264 with big blocks noise. Processing H.264 sometimes takes me several hours of After Effects CS 5.5 renderng on core i7 with 16gb mem while rendering from DNGs took me about 30 minutes for resulted 38 second timelapse I submitted.

Resoulution of RAW vs. camera H.264 is visible step-up like comparing 1920×1080p25 vs. 1280×720p50.

The only drawback is files size. 32gb for about 6 minute Full HD RAW.

Best Regards, Andrey

1471 posts
  • Sold between 10 000 and 50 000 dollars
  • Author had a File in an Envato Bundle
  • Most Wanted Bounty Winner
  • Russian Federation
  • Has been a member for 2-3 years
  • Exclusive Author
  • Referred between 10 and 49 users
  • Bought between 10 and 49 items
  • Grew a moustache for the Envato Movember competition
AndrVlad says

While my first items processed from RAW are in the queue here and since anyway FLV previews as well as youtube can’t demonstarate real quality I uploaded two watermarked 5 second pieces in Quicktime Photo Jpeg so anyone can download and check what can be expected from Canon 5D Mark III with 14bit RAW Full HD recording enabled by amazing Magic Lantern team.

https://www.dropbox.com/s/r0ot1cp061mqoiq/sunset_ship_fullhd_preview.mov
https://www.dropbox.com/s/iodxps06hykrkz4/clouds_fullhd_preview.mov
(do not play inside web page, download to preview in full details)

Interesting to hear your opinion how it looks versus any best in camera encoded H.264?

Best Regards, Andrey

1802 posts Videohive author & moderator
  • Elite Author
  • Sold between 100 000 and 250 000 dollars
  • Community Moderator
  • Referred more than 2000 users
  • Author had a File in an Envato Bundle
  • Most Wanted Bounty Winner
  • Contributed a Tutorial to a Tuts+ Site
+7 more
Mocarg Volunteer moderator says

^

Wow! That detail!

313 posts
  • Author had a Free File of the Month
  • Referred between 1 and 9 users
  • Sold between 1 000 and 5 000 dollars
  • Bought between 1 and 9 items
  • Exclusive Author
  • Has been a member for 2-3 years
  • United Kingdom
Panda-Bear says

Definitely see the difference. deeper, richer colour saturation. lots of detail in the sky. nice stuff.

what CF card did you have? 95/mbs? if not, what resolution in ML did you use – say, 1330×720 than scaled to Full HD in post?

Really interested to see if any artefacts occur when scaling up .raw/dng stuff.

Thanks Andrey.

1802 posts Videohive author & moderator
  • Elite Author
  • Sold between 100 000 and 250 000 dollars
  • Community Moderator
  • Referred more than 2000 users
  • Author had a File in an Envato Bundle
  • Most Wanted Bounty Winner
  • Contributed a Tutorial to a Tuts+ Site
+7 more
Mocarg Volunteer moderator says

Definitely see the difference. deeper, richer colour saturation. lots of detail in the sky. nice stuff.

what CF card did you have? 95/mbs? if not, what resolution in ML did you use – say, 1330×720 than scaled to Full HD in post?

Really interested to see if any artefacts occur when scaling up .raw/dng stuff.

Thanks Andrey.

Why would he do that? Shoot in lower res and then up the scale? Then there is no point in this ML hack…

1471 posts
  • Sold between 10 000 and 50 000 dollars
  • Author had a File in an Envato Bundle
  • Most Wanted Bounty Winner
  • Russian Federation
  • Has been a member for 2-3 years
  • Exclusive Author
  • Referred between 10 and 49 users
  • Bought between 10 and 49 items
  • Grew a moustache for the Envato Movember competition
AndrVlad says

Definitely see the difference. deeper, richer colour saturation. lots of detail in the sky. nice stuff.

what CF card did you have? 95/mbs? if not, what resolution in ML did you use – say, 1330×720 than scaled to Full HD in post?

Really interested to see if any artefacts occur when scaling up .raw/dng stuff.

Thanks Andrey.

I am using Lexar 1000x 32gb CF which is fast enough to record 1920×1088 upto 27fps RAW or 1920×1280 – 24fps or even 2K 2048×1152 24 fps works! Magic Lantern also allows to record 3K-4K resolution but accordingly slower like 4-6 fps which is still pretty usefull for timelapses or other slow pictures. It still early stages of RAW recordings and I hope ML developers will find ways to increase capabilities with loseless compression or other tricks.

Best Regards, Andrey

1471 posts
  • Sold between 10 000 and 50 000 dollars
  • Author had a File in an Envato Bundle
  • Most Wanted Bounty Winner
  • Russian Federation
  • Has been a member for 2-3 years
  • Exclusive Author
  • Referred between 10 and 49 users
  • Bought between 10 and 49 items
  • Grew a moustache for the Envato Movember competition
AndrVlad says

Why would he do that? Shoot in lower res and then up the scale? Then there is no point in this ML hack…
It makes sence for slow cards or for other cameras like 5D Mark II, 60D, 50D, etc (they have hardware limits). And per examples I seen image quality after upscaling for many cases is still better than in camera Full HD + H.264.

Best Regards, Andrey
313 posts
  • Author had a Free File of the Month
  • Referred between 1 and 9 users
  • Sold between 1 000 and 5 000 dollars
  • Bought between 1 and 9 items
  • Exclusive Author
  • Has been a member for 2-3 years
  • United Kingdom
Panda-Bear says

Mocarg said Why would he do that? Shoot in lower res and then up the scale? Then there is no point in this ML hack…

Absolutely right. Mocarg. If you have a 5D, and shoot the best footage you can normally in H264. It will not be anywhere near the pristine quality you get with shooting in 14bit Raw ant say, 988×540 and scaling up to 1080p.

THe far superior pristine edge you get with RAW is much worth the extra workflow.

And if you don’t believe me, try shooting at night and compare your native grainy H24 to a scaled-up 14bit RAW composition.

by
by
by
by
by
by