Description image

Can You Seamlessly Intercut Canon 5D Mark III RAW & RED SCARLET Footage?

RED SCARLET Canon 5D Mark III RAW ComparisonWhile you always want to choose the right camera for the job, sometimes you have to deal with what you have. There are times when you might have a RED SCARLET (or maybe even a RED EPIC), but maybe you need something that’s a little more light-sensitive as a B-cam for specific shots, or you might need dual-coverage and you’ve already got a second camera handy. What if that second camera is a Canon 5D Mark III, and you’ve enabled the Magic Lantern hack so that you can shoot RAW? How would the two cameras stack up against each other?

Ryan Lightbourn has been playing around with the Magic Lantern on a Mark III, and he recently shot a piece where he cut between that and a RED SCARLET (with limited grading). Head on over to Vimeo to download the full 1080p file, as that shows a bit more detail and less compression (thanks to cinema5D for the tip on this one):

Shot back & forth between the two cameras on the same day.

Minor grading done in ACR & RCX to match contrast. No sharpness added. Scarlet footage shot in 4k, downscaled 50% to 1080p.

Lenses used: Tokina 11-16mm & Zeiss 25mm 2.8 (both with Formatt 1.2 ND)

As Ryan noted on Vimeo, what might look like aliasing with the Mark III is actually the power lines painted black and white:

Ryan Lightbourn SCARLET 5D3 RAW power_lines

This is not to say one camera is better than the other, as there are positives and negatives about both. I think it’s important to look at what’s out there and how something might compare in a given situation, and it has become clear from everything I’ve seen and from shooting the Mark III in RAW that the footage absolutely stands up at 1080. Obviously at 4K, the SCARLET will come out on top, but if you’re not shooting something that needs to be in 4K, this is a valid comparison, and the Mark III should fit in nicely.

Ryan also recently shot this music video with the Mark III in RAW, 24fps at ISO 1000 and 60fps at ISO 1600. Slow motion was shot at 60fps at 1728 x 458 and upscaled/stretched to 1080p (some NSFW language):

Another video from Ryan messing with 60fps slow motion Mark III RAW, shot at 1728 x 458 and stretched to 1080p. Exported @ 720p:

And just for kicks, here is a Vimeo Staff Pick of his from a few years ago shot on the Canon 5D Mark II and Nikon primes:


[via cinema5D]


We’re all here for the same reason: to better ourselves as writers, directors, cinematographers, producers, photographers... whatever our creative pursuit. Criticism is valuable as long as it is constructive, but personal attacks are grounds for deletion; you don't have to agree with us to learn something. We’re all here to help each other, so thank you for adding to the conversation!

Description image 61 COMMENTS

  • Seen this on reduser yesterday. The RED fan boys are still pissed about ML.

    • Link?

    • Rodrigo Molinsky on 07.8.13 @ 10:37AM

      I got one 50D, but when I have the conditions, I’ll not hesitate to get one 5D mk III.

    • They must not have known about this:

      “Scarlet footage shot in 4k, downscaled 50% to 1080p.”

    • I’m hardly say “pissed” is accurate. I’ve been doing several tests of the ML RAW module, some of which are posted on – I think it’s amazing the quality of video that is now achievable on the 5D models, but it by no means makes any of the RED owners I now, myself included, regret purchasing from Jim’s outfit. I’m working on this exact same test myself ATM but with a twist, comparing inter-cutting ML RAW vs BlackMagic Pocket alongside RED footage. Right off the bat, if the crop factor and handholding/small ringing is important, a shooter may want to consider dumping their canon cam and picking up a couple BM pockets. Of course ML is free and already here, but i definitely cannot recommend buying a 5D3 to new shooters. Thank You

      • marklondon on 07.8.13 @ 4:09PM

        You make great points about the BMD Pocket, and that looks a very attractive camera (I have pre-ordered one) but I would also love a full-frame RAW video cam in my arsenal. It is hard to consider buying a 5D3 purely for this hack, but as I’ve invested almost 6 years in learning to (successfully) work with DSLRs it seems mad to start over in a new camera system if a decent solution in a form that perfectly fits all my rigs/jibs/sliders and lenses is available over the counter.
        My only real holdback is trying to estimate how long it will be before a version of the 1DC at @$5k pops up. :-) .
        Re the RED thing, there are many benefits to shooting/owning one, although for me the goodness starts at the EPIC level.

        • The problem with using the ML hack for RAW in full liveview is that there is still line skipping issues. To get the best image quality, and no pixel binning / line skipping a shooter will need to use 5x crop mode. In crop mode a lens ends up getting a FOV 5.0~ish crop factor – much larger then the 2.3~ish factor on the BM pocket. Long story short, even though the BM pocket has a smaller sensor, it will provide a larger FOV then the ML hack. Hope this info helps. Thanks

          • marklondon on 07.9.13 @ 9:35PM

            I’ve seen all the various tests so far, and performed my own. Yes (just like the Nikon D4) the zoomed in image is the sharpest. But honestly, the full-frame RAW is still very very good. I would never use it in full zoom – my clients, which include a couple of major broadcasters, would not require it.
            I expect the Pocket to be fantastic. However, I know it will take me a while to get used to new lenses (even if they’re old lenses!) and using it on set, where I won’t miss a beat with the 5D shooting RAW in standard FOV.
            I’ve borrowed a hacked 5D3 for the rest of the week. If it works as I expect it to, then I’ll buy one.

          • Correction, it appears that while the screen may say 5x when in crop mode, it is really recording closer to 3x. Additionally, 10x crop mode also records the exact same 3x view – either way, it is a seriously FOV crop factor.

          • Thanks for sharing Mark. I completely hear what you’re saying. work is work and tools are tools; whatever gets the job done w/in client expectations pays the bills. In terms of lenses though, instead of using a dumb adapters I’m very curious about the metabones speedbooster adapter to help reduce crop factors and increase light going to the sensor.

      • Dan,

        I appreciate the honesty.

        • No problem and thanks for your feedback Gene. I’ll be posting more ML comparison videos on The soonest one that should go up will be focused on FOV. Same camera, Same shot, same lenses but different mode ie, canon standard H264, ML full liveview and ML 5x crop mode – showing specifically how much the 5x crop mode dramatically effects shot composition and side issues that may result in unusable footage. Thank You.

  • Great…. I made video for youtube, I think I dont need RAW video for that job.

  • Here’s a video of a Canon 5D2 with a Panasonic AF100. There’s a difference. The AF100 looks beautiful. The 5D2 looks good too but pastel and flat compared to the AF100:

    • Those kinds off examples are ridiculous. They haven’t even tried to match them and for some reason the 5d stuff is absolutely riddled with aliasing. Even the 5d II doesn’t do it THAT much so something has gone wrong somewhere.

  • It would be more interesting to see the 5D3ML vs GH3, especially when the GH3 hack comes out.

  • Ok, you’ve ALMOST convinced me Joe :-) That 5D3 RAW footage looks fantastic, and I’m not surprised the RED boys/girls are pissed if they’re looking at the rental side. That you can do decent overcrank on it almost seals it for me. Most of the work I shoot personally is very similar to the work above.
    I’ve played with a hacked cam briefly, but I think its time to take a test a little more seriously.

  • I can’t believe at the end how poor the Scarlets colour rendition on the water was. Looked so much browner than the 5d. ;) The MIII looks great. Highlight rolloff, and detail in clouds still preferable on Scarlet. But there’s not much in it at 1080P. Great days to be a shooter.

    Oh and thanks Ryan. This looks a better choice to intercut with Scarlet than my GH3.

  • BTW there are ZERO 5D3s on CL in LA right now.
    And this:

  • hand to God. If I see another video with “Time” from the Inception soundtrack, i will explode in a violent, shitty mess.

  • In “Heroin”, aren’t the shadows severely murky? (esp. in the beginning)

    • Shhhhh, you’re not supposed to point out the man behind the curtain. These are Canon users.

    • 5D has a reputation for only having good quality on the up close objects.

      This is a GH3 video. Look for murky things in the background. And see if it is sharper overall than the 5D3ML.

      • Every single damn message you post is about the shitty ass GH3 and how much it’s “better”. It isn’t.

        • Thankyou, Mikko! This individual sounds like a child. People who are doing well, liking what they have, enjoying themselves or whatever… Have no time, or reason to throw insults or simple ‘hate’ on something else. If the GH was so magnificent, he wouldn’t be so mad about the 5D, he’d simply be out making use of his camera.

  • Having tested the raw video on the 5Dmiii and comparing it to my Scarlet they do cut together very nicely. Its really amazing what the ML team has done! Great job!

    But from a usability stand point, its still a thousand times easier to edit/process Red footage then the uncromprness raw files from the 5Dmiii. The 5Dmiii is great for short burst of filming, but not great as a replacement solution for working with raw footage (yet anyways).

    • Did you try GingerHDR? .RAW files work seamlessly now, just like any other format. I just hope they add the ability to use Camera Raw with the .RAW files…I find it much easier to pre-grade with ACR.

  • pablopablo on 07.8.13 @ 4:23PM

    fuck im getting one ..

    • After clicking the link, make sure you select ‘Original’ for quality

      • marklondon on 07.9.13 @ 2:16AM

        That’s some astonishing footage. Cheers.

        • Can you imagine if just half the people working on improving 5D video had been working on improving GH video what GH video would look like! Nick Driftwood did a fantastic job by himself on the GH2 hack. What if there were 5 more worked on it and would be working on the GH3 hack too?

          • What if, what if. What if these cameras had global shutter. What if we could buy 128GB cf cards with write speeds of 200MB a second for $50

          • What I was saying wasn’t really unreasonable.

            The GH makes better video than the 5D. What if as much was done to improve the GH as was done to the 5D. I don’t see anyone chafing at the upgrades done to the 5D. Some people talk like the introduction of 5D patches are party time. On the other hand, I do see some people getting chafed over the GH. The reason so much is done to patch the 5D is because the video looks inferior and needs it.

          • Are you upset that your precious GH doesn’t have enough interest for a group of hackers to get involved??? CLEARLY this is a topic about the 5D. It amazes me at how people seem to throw themselves in conversations that have NOTHING to do with their gear, and make it a conversation about their gear! If you are that interested in making a hack as such for the GH MAKE one! Otherwise, just keep hoping. Your downplay of the 5D isn’t needed here. The GH certainly isn’t relevant in this conversation either.

  • Question,

    Would be possible to use an SD converter to shoot longer continuous RAW on the 5d Mark II?

  • I would love to see these examples but of peoples profiles and how they compare too. Also is it just the compression from vimeo where there are some weird blocky horizontal lines happening especially on clear skies? eg ?

  • The Canon 5D Mk III is nothing short of amazing in it’s abilities. Here’s a test video I shot showcasing the video capabilities of the camera,

  • The group was known as “Human Individual Metamorphosis,” and then later “Total Overcomers Anonymous,” before becoming “Heaven’s Gate” in the 1990s. Members had to live with the group as well as give up their possessions and familyニューバランス 996 グレー To perfect themselves for salvation, Applewhite encouraged members to detach themselves from emotion and give up sex entirely. Members dressed androgynously with loose clothing and closelyshaved shaved hair.

  • A lot of big name movies with the Red. I haven’t even heard of most of the ARRI ones.

  • Never heard of skyfall?