Jump to content

Photo
- - - - -

New 5D Raw developments - plus my low light comparison with Blackmagic Cinema Camera


  • Please log in to reply
52 replies to this topic

#1 EOSHD

EOSHD

    Andrew Reid - British Filmmaker - Editor EOSHD

  • Administrators
  • 3,737 posts

Posted 17 May 2013 - 12:51 AM

5D Mark III raw at ISO 1600

 

First the news. Magic Lantern have playback of raw files working with early stage code. Black and white and slow frame rate but works - already. Experimental audio recording has been added. Testers and early adopters will need to see if it in sync.

Future nightly builds of the Magic Lantern 5D Mark III Alpha firmware will see raw recording start / stop added to the physical movie record button as well as many other usability improvements.

Read the full article here

#2 Oscar M.

Oscar M.

    Advanced Member

  • Members
  • PipPipPip
  • 69 posts

Posted 17 May 2013 - 02:03 AM

Oh yeah. Let's give ML 2-3 months and 5D3 is going to be a BEAST!



#3 Alex Carignan

Alex Carignan

    Newbie

  • Members
  • Pip
  • 6 posts
  • LocationMontréal, QC, Canada

Posted 17 May 2013 - 02:05 AM

For fucking fuck's fucking sake that is making me jealous... Damnit. 

Andy, do you think something like this could become possible for the Panasonic GH3?



#4 Oscar M.

Oscar M.

    Advanced Member

  • Members
  • PipPipPip
  • 69 posts

Posted 17 May 2013 - 02:09 AM

Andrew, just wondering have these been your temperature settings for all your other RAW sample videos? Just wondering if this would be the offset for all RAW frames coming out of the sensor . . .

5d3-raw-ps-iso-1600.png



#5 EOSHD

EOSHD

    Andrew Reid - British Filmmaker - Editor EOSHD

  • Administrators
  • 3,737 posts

Posted 17 May 2013 - 02:11 AM

Andy, do you think something like this could become possible for the Panasonic GH3?

 

Very unlikely

 

How to run C code on the GH3? Vitaliy can't do it yet.



#6 nahua

nahua

    Advanced Member

  • Members
  • PipPipPip
  • 408 posts
  • LocationHawaii

Posted 17 May 2013 - 02:33 AM

All the improvements are really good.  I wish the B/W playback could be paused or quit, hard to watch those frames at maybe 5fps?  And for the entire duration of the recording?  But I'm asking too much.  There's so much changing so quickly and the improvements are astounding.  I still can't get 1080P w/o dropped frames but I know they're doing their best to improve it.  This is really awesome!!!  :-)



#7 Germy1979

Germy1979

    Advanced Member

  • Members
  • PipPipPip
  • 575 posts

Posted 17 May 2013 - 02:52 AM

I always liked the Mark 2's image over the 3's.. But that was the compressed stuff. I wonder if there's a big difference here. (Sans moire, blah blah.)

#8 gabbassov

gabbassov

    Newbie

  • Members
  • Pip
  • 3 posts

Posted 17 May 2013 - 04:19 AM

If there is no real time playback... the whole thing is kind of useless for TV spot production work.

I hear people say, that shooting raw is expensive and they count the  dollar per minute cost, but I would buy most expensive cards on the planet and have my money back in a couple of days of shooting, because instead of renting Red Epic (600$ per shift) or Alexa(1000$ per shift),paying camera technition (200$ per shift) and SDI field monitor (60$ per shift) I just bring my raw 5d, shoot the commercial and go back home to make family pictures and videos with the same tool. Of course there is a BIG risk to shoot the commissioned work on the hacked 5d, so I would test it couple of month in a different conditions before I ever take it to any serious shoot.

I really hope the ML team will get it to the point when it is reliable and solid. For now 5d's H264 and Pro cameras are the save option. Although I would never rent or buy any of canon's C line cameras, just because I don't want to give any profit to the gritty bstrds canon.



#9 RobertJ

RobertJ

    Newbie

  • Members
  • Pip
  • 1 posts

Posted 17 May 2013 - 04:29 AM

Just wondering, was the lighting exactly the same for both images?

 

If you look at that strong highlight on the round base of the desklamp, you'll see that the BMCC can completely recover it, while the 5D3 cannot.  This with the "Whites" and "Highlights" slider pulled all the way down to recover as much as possible. 

 

I'd rather shoot 5D3 RAW because of the sensor size, but if the lighting was the same for both images, it really shows the tremendous dynamic range that the BMCC is capable of.



#10 jlev23

jlev23

    Advanced Member

  • Members
  • PipPipPip
  • 82 posts

Posted 17 May 2013 - 05:32 AM

thank you so much for this andrew, since this is the first ML hack ill ever try, i have two quick questions i can not find the answers too.

So, i read that you have to use the CF card that the firmware was installed with, is that correct? so the firmware boots up correctly, and if you go back to a regular card the hack will be gone?

if that being the case, can you NOT format the card after you transfer the raw files off of it?

secondly, i wonder, does the uncompressed clean hdmi out still work while you are recording raw to the card?

if so that would be great if you can be getting a prores file to edit with right away and then go back to the raw files for the final grading and out put. please let us know,

thanks again for everything.



#11 ike007

ike007

    Advanced Member

  • Members
  • PipPipPip
  • 39 posts

Posted 17 May 2013 - 05:55 AM

 

 

 

If there is no real time playback... the whole thing is kind of useless for TV spot production work.

I hear people say, that shooting raw is expensive and they count the  dollar per minute cost, but I would buy most expensive cards on the planet and have my money back in a couple of days of shooting, because instead of renting Red Epic (600$ per shift) or Alexa(1000$ per shift),paying camera technition (200$ per shift) and SDI field monitor (60$ per shift) I just bring my raw 5d, shoot the commercial and go back home to make family pictures and videos with the same tool. Of course there is a BIG risk to shoot the commissioned work on the hacked 5d, so I would test it couple of month in a different conditions before I ever take it to any serious shoot.

I really hope the ML team will get it to the point when it is reliable and solid. For now 5d's H264 and Pro cameras are the save option. Although I would never rent or buy any of canon's C line cameras, just because I don't want to give any profit to the gritty bstrds canon.

Of course if you are shooting a commercial gig with between 20k to 100k of productions cost per day, you will bother to save few hundred of dollars for rent over taking an Alexa or an Epic and shoot it with the thingy.

BTW, testing something for a couple of months, is worth a lot of money and time from your life that you will never get back.



#12 sugartown

sugartown

    Member

  • Members
  • PipPip
  • 28 posts

Posted 17 May 2013 - 06:03 AM

BTW, testing something for a couple of months, is worth a lot of money and time from your life that you will never get back.

 

I think they just mean shooting it a lot for their own personal use, not formal tests, and just waiting until they know it like the back of their hand before using it on the clients dime.

 

If you're with a budget to afford a better camera out of the box, there's no reason to use this unless it gives you some superior function/form. If you have access to your own gear, it is empowering.



#13 gabbassov

gabbassov

    Newbie

  • Members
  • Pip
  • 3 posts

Posted 17 May 2013 - 07:21 AM

Of course if you are shooting a commercial gig with between 20k to 100k of productions cost per day, you will bother to save few hundred of dollars for rent over taking an Alexa or an Epic and shoot it with the thingy.

BTW, testing something for a couple of months, is worth a lot of money and time from your life that you will never get back.

True.

But not all the time our client commissions high budget TV spots, sometimes it is just small promo, and client asks us to give something under $10k. For this kind of the job 5d is the best tool for us. And now imagine with 5d raw the low end commercial can look like high budget spot. At the end everyone is happy. If the raw will be rock solid I will take it to high end shoot (I think raw 5d deserve it :) )

I don't mind get out somewhere in mountains or seaside with my close friends and shoot some raw stuff. If the 5d raw can handle some high temperature on a beach and the Buffer will not  run out after we finish couple of beers the test will be considered successful  :)



#14 cerdt

cerdt

    Newbie

  • Members
  • Pip
  • 1 posts

Posted 17 May 2013 - 07:39 AM

Thanks for comparsion. Canon metadata shows exposure time 1s. Is it a photo or a video frame? What is shutter speed of both frames for comparsion? I suppose it should be max. 1/25s from video.



#15 MatthewP

MatthewP

    Advanced Member

  • Members
  • PipPipPip
  • 63 posts

Posted 17 May 2013 - 07:42 AM

Andrew, have you seen the suggestions in this thread regarding storage? http://www.eoshd.com...kiii-raw-video/

 

 

 

Might be worth looking into if you can't find a good fast CF card.



#16 EOSHD

EOSHD

    Andrew Reid - British Filmmaker - Editor EOSHD

  • Administrators
  • 3,737 posts

Posted 17 May 2013 - 09:43 AM

Thanks for comparsion. Canon metadata shows exposure time 1s. Is it a photo or a video frame? What is shutter speed of both frames for comparsion? I suppose it should be max. 1/25s from video.

 

Ignore the metadata. It isn't right yet in the 5D raws.



#17 mtheory

mtheory

    Advanced Member

  • Moderators
  • 254 posts

Posted 17 May 2013 - 09:44 AM

I always liked the Mark 2's image over the 3's.. But that was the compressed stuff. I wonder if there's a big difference here. (Sans moire, blah blah.)

 

Is the rolling shutter still terrible on 5D2 raw?



#18 EOSHD

EOSHD

    Andrew Reid - British Filmmaker - Editor EOSHD

  • Administrators
  • 3,737 posts

Posted 17 May 2013 - 09:44 AM

Just wondering, was the lighting exactly the same for both images?

 

If you look at that strong highlight on the round base of the desklamp, you'll see that the BMCC can completely recover it, while the 5D3 cannot.  This with the "Whites" and "Highlights" slider pulled all the way down to recover as much as possible. 

 

I'd rather shoot 5D3 RAW because of the sensor size, but if the lighting was the same for both images, it really shows the tremendous dynamic range that the BMCC is capable of.

 

Yes lighting the same.

 

Need to do a proper dynamic range test. I think they have they distributed differently. You can pull the blacks up more on the 5D Mark III but recover more in the highlights on the BMCC.



#19 odie

odie

    Advanced Member

  • Members
  • PipPipPip
  • 106 posts

Posted 17 May 2013 - 10:28 AM

Since I own the 5dmlll ...there is no reason to purchase any of the black magic cameras??



#20 EOSHD

EOSHD

    Andrew Reid - British Filmmaker - Editor EOSHD

  • Administrators
  • 3,737 posts

Posted 17 May 2013 - 10:44 AM

Since I own the 5dmlll ...there is no reason to purchase any of the black magic cameras??

 

Not unless you might need -

 

4K

Global shutter

Micro Four Thirds mount

SSD media

HD-SDI

Resolve

 

Still a lot to offer.






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users