5D Mark III with Continuous RAW Video Recording

[LetTheRightLensIn]

I am always hesitant to post new ML features from the dev stage here because general users aren't advised to use the stuff yet - but since the news ended up here anyway:

Another great feature of this discovery is the raw histogram which will let you actually see if anything is under- or overexposed in raw and not just in jpeg like with the vanilla Canon fw.

Wait. Is the Trammel Hudson heavily involved with all of this the very same famed hacker of old Atari 8 bit days Tram Hudson??

[cayenne]

I've been poking around. Apparently the Lexar 1000x cf 64GB cards are 'on sale' now for about $299.

Hmm...I wonder if it might be a good idea to pick up one of these in anticipation of ML raw video coming online soon.

It isn't like I couldn't be using it right now for normal video/stills photography.....

Hm......


cayenne

I bought mine for $278 last fall from Adorama, so $299 is not much of a sale.  I'd wait for a better price.

HEY!!!

Thanks for the heads up!! 
I'll definitely be on the lookout for price drops on these then more to that ballpark!!


cayenne

[bp]

I keep seeing occasional posts about how Canon should put these guys on the payroll.

Frankly, at the rate they're going in the direction of high bitrate recording, the company who should seriously consider subsidizing ML is LEXAR

 

[Marsu42]

I keep seeing occasional posts about how Canon should put these guys on the payroll.

Canon fw and Magic Lantern have completely different approaches (market stratification & ultra-conservative stable fw vs. customization, features & backports). Any of the ml devs would probably go crazy after working one month for Canon :-) so let's hope that Canon at least stays out of ml's way if they decided not to help.

[@!ex]

Not sure if anyone has posted this yet but I had to share.  Check out the difference in dynamic range from the 5D mk3 with canon firmware output (i-frame):

Canon 5D Mark III i-frame video test

Now here is the same scene shot with magic lantern firmware using the new 14 bit RAW output.  He is able to completely recover all the shadows and highlights.  Crazy.

Canon 5D Mark III i-frame video test

You'd probably need a lot of 1000x CF cards to hold all that data though.

[mrmarks]

I have a basic question: Which software can process the 5D3 RAW movie files? Can Adobe Premiere do this? Thanks

[risc32]

the videos look the same to me.

[matt2491]

Yea he accidentally posted the same video twice. Here is the raw video:

Magic Lantern 1920 x 1080 full 1080P raw test Canon 5D Mark III Edit 2

[LetTheRightLensIn]

I have a basic question: Which software can process the 5D3 RAW movie files? Can Adobe Premiere do this? Thanks

first raw2dng then into After Effects as a batch load and then written out as an uncompressed video file and then into Premiere  . There are probably other possible steps to take after raw2dng.

[westr70]

the videos look the same to me.

+1.  I was seriously underimpressed.  The "new" one is great. 

[peederj]

The A/B test to do is the RAW version vs. the HDMI out using Cinestyle, recorded to the Ninja 2 in 220Mbps ProRes HQ and graded in post with the proper LUT. Because anyone who cares about getting the most IQ out of the camera will be using that setup (or something very similar) rather than internal...of course the RAW is going to kill the miserable internal codec.

And BTW, all my testing has repeatedly shown ALL-I and IPB are 100% identical IQ on the 5D3 internal. Haven't seen anything credible to refute that...I think it's just Mbps marketing to counter the GH2 hack.

That is alot of misinfo.

1. The internal codec isn't what does much damage. Point at a static scene and record with Ninja 2 at ProRes HQ and record internally 1.2.1 and there is NOT much difference you can see at all. A little but it is all very subtle to be honest. It's vastly smaller compared to the difference between what this ML recorded stuff looks like compared to normally internally recorded video.

2. For static scenes or ones with just bits moving around in the frame all-i and IPB are pretty much the same and all-i is just a horrible waste of space, perhaps worse if anything. If you pan around or the entire scene is changing frame to frame then ipb totally falls apart and all-i holds up much better (as does say pro res on ninja 2).

Please. I'm not the source of misinfo, I post to keep people honest. I have done quite scientific tests with the 5d3 and Ninja 2 (and the C100) to know very well what I am talking about, compared to the legions of people who have no idea how to properly ask these questions much less answer them. And dear Andrew is too animated in his disdain for Canon to claim any credibility on the matter if that's your authority.

For the hundredth time, video desperately needs an independent lab to publish fully controlled test charts and the like for us to work off objectively. A bunch of well-meaning enthusiasts and blowhards does not suffice. I will test the ML implementations when I get a fast card to do so and I will report back. While not a common net.idiot I am not a substitute for a lab either.

[LetTheRightLensIn]

The A/B test to do is the RAW version vs. the HDMI out using Cinestyle, recorded to the Ninja 2 in 220Mbps ProRes HQ and graded in post with the proper LUT. Because anyone who cares about getting the most IQ out of the camera will be using that setup (or something very similar) rather than internal...of course the RAW is going to kill the miserable internal codec.

And BTW, all my testing has repeatedly shown ALL-I and IPB are 100% identical IQ on the 5D3 internal. Haven't seen anything credible to refute that...I think it's just Mbps marketing to counter the GH2 hack.

That is alot of misinfo.

1. The internal codec isn't what does much damage. Point at a static scene and record with Ninja 2 at ProRes HQ and record internally 1.2.1 and there is NOT much difference you can see at all. A little but it is all very subtle to be honest. It's vastly smaller compared to the difference between what this ML recorded stuff looks like compared to normally internally recorded video.

2. For static scenes or ones with just bits moving around in the frame all-i and IPB are pretty much the same and all-i is just a horrible waste of space, perhaps worse if anything. If you pan around or the entire scene is changing frame to frame then ipb totally falls apart and all-i holds up much better (as does say pro res on ninja 2).

Please. I'm not the source of misinfo, I post to keep people honest. I have done quite scientific tests with the 5d3 and Ninja 2 (and the C100) to know very well what I am talking about, compared to the legions of people who have no idea how to properly ask these questions much less answer them. And dear Andrew is too animated in his disdain for Canon to claim any credibility on the matter if that's your authority.

For the hundredth time, video desperately needs an independent lab to publish fully controlled test charts and the like for us to work off objectively. A bunch of well-meaning enthusiasts and blowhards does not suffice. I will test the ML implementations when I get a fast card to do so and I will report back. While not a common net.idiot I am not a substitute for a lab either.

I compared them all myself. All-i vs IPB is hugely different if there are major frame to frame changes (granted during such motions the eye isn't as adept at noticing things are wrong at full speed) although for static yeah not much. Ninja 2 did way better than IPB for frames that huge changes as well (didn't get to compare to all-i for that yet) but for static stuff it doesn't look so different and even for most grading it doesn't seem super different yet, better but it's modest. These ML RAW video samples I mean wow the difference leaps out at you and hits you over the head.

[wickidwombat]

They are claiming that heat is not really an issue, in reality the 1DX "should" be hackable and not have the heat issues too.
The thing stopping them there is Canons legal team.

I wonder how long it will take before Canon come down on ML for releasing this.
It could put a large dent in C100/C300 sales.

Well, if ML is doing a clean room reverse engineer of the camera, that should be perfectly legal. He isn't doing anything that helps 'circumvent' safe guards....like people have come down on gaming mods for. This hack doesn't allow you to circumvent copyrighted material.

One is allowed to do what they wish with hardware they purchase (at least still for now, although the corps are trying more and more to infringe on that)....so, this should be perfectly legal.

However, Canon can still throw legal threats around if they wish, whether they have merit or not. These days, it doesn't matter if you have a legal case or not...it is deep pockets that often win.

However, if Canon were to do this, the "Streisand Effect" might take place, especially if he's released the code open source and other people take the code and join in the efforts. Canon might get more spread and publicity of this type code rather than stiffling it.

But generally, if you're doing a clean room reverse engineer on hardware you bought and own, with no NDA's signed, etc...you should be perfectly legal.

If Canon legal comes knocking on ML's front door, I'd say first thing he might do, is contact the EFF and try to get them on his side with their legal help/advice.

My $0.02,

cayenne

I think Canon are secretly happy with this.  ML has been around for quite sometime and Canon knows how much ML helped them sell a lot of 5D2s.  Who knows, they might also profit by adopting those free codes to their new cameras thereby raising raising their value without even doing anything... Good times!

I think you are right they are probably reading the news of this and going right o boys better crank up the L glass production sales volumes are gonna go UP! sure some people will go with alot of good 3rd party glass but at the end of the day everyone loves some nice L glass

if anything this is gonna help propel 5Dmk3 sales well ahead of the nikons so I would think Canon should be very happy and they didnt have to lift a finger

[risc32]

they should be. i hope they don't get any stupid ideas of shutting it down. It's a real + for them to have people like ML doing this.

[dslrdummy]

So why is he wearing a helmet?