January 17, 2013

Blackmagic Cinema Camera, Sony FS100, and Canon 5D Mark II Face off in a Hair and Skin Tone Test

Color is just about the most subjective aspect of any visual creation. Everyone sees color a little differently, so it's no surprise that we talk endlessly about color science and about which cameras we prefer. Certain looks are too much for some people, and others are not enough. Blackmagic spent a great deal of time developing their color science with Australian Director of Photography John Brawley, and I think working with an actual shooter in developing their camera has made a significant difference in the visuals of the final product. Adam Roberts got a hold of the BMCC and performed a thorough test to compare the camera's skin tones to that of the FS100 and the Mark II. Click through to check it out.

Here is his process, thanks to Michael at Notes On Video for the find:

A simple Hair and Skintone test.

We wanted to test the new Blackmagic Cinema Camera to see how it performed with skintones and hair. We used the Sony NEX-FS100 and Canon 5D MK2 as known refrences for comparison.

Sample DNGs can be downloaded here:
dl.dropbox.com/u/6460491/BMCC_skintone_test.zip

Notes on workflow:
Both the Sony NEX-FS100 and the Canon 5DMK2 footage was transcoded to ProRes HQ with 5DtoRGB. I then colour corrected the footage in Final Cut Pro X.

The Blackmagic footage was exported from Resolve in 3 ways as I wanted to test different work flows.
• 1st option I applied a Rec709 LUT to the RAW files and exported ProRes HQ 1080p.
• 2nd option I applied the BMC Film LUT to the RAW files and exported ProRes HQ 1080p.
Both methods involved not colour correction in Resolve. All that was do was a LUT was applied to the RAW files. I then colour corrected the footage in Final Cut Pro X.
• In the 3rd option I graded the footage in Resolve with a single node and exported ProRes HQ 1080p

All footage was then edited in ProRes HQ in Final Cut Pro X.

The video is in full 1080, but you can also download the original file to see it in higher quality.

I think there is definitely something to like about each of the images, and obviously different skin and hair colors may respond differently than we're seeing here. From the very beginning, the Cinema Camera was exciting just because of the sheer amount of performance for the money, but as things progressed, it's clear this thing was made to produce the most pleasing images possible from a cinema standpoint. It's not necessarily about reproducing reality as closely as possible, but finding the balance between what looks good subjectively and what the camera is actually capable of doing given the hardware inside.

I'm still partial to the Cinema Camera here, but if you work them enough I'm sure you could come close to matching them. What you can't match is resolution and clarity, and that's definitely where the DSLRs have been lacking from Canon compared to the more fully-featured large sensor video cameras (though the 1D C may finally be the one exception to that -- except at $12K it's not really in typical DSLR territory anymore).

I think it's also helpful now to start looking at these tests not to see which camera is "better," but to look at which images might work best for a particular shoot. While I say you can probably come close to matching them, they are all still biased in one way or another from the start, so rather than fighting the camera for the look you want, it might benefit you to find the one you like for one shoot, not necessarily all of them. Different cameras also produce different images under different lighting scenarios, so one may actually produce a better result than another under daylight, tungsten, or mixed lighting situations.

Which one do you prefer? Do you like all of them in their own way? Have you had the opportunity to choose a camera based on the look you liked for a particular production?

Link: Adam Roberts -- Vimeo

[via Notes On Video]

Your Comment

83 Comments

Is it just me or does the BMCC have a really strong greenish tint?
I've noticed that across every single BMCC clip I have seen so far and am wondering why nobody else complained about that earlier.

January 17, 2013

0
Reply

I actually prefer the non corrected versions. I was surprised with the look out of the FS100, I liked it. Which is unusual. The blackmagic is a treat though. If they ever make a s35 version that isn't so awkward...watch out.

January 17, 2013

0
Reply

do we even need an S35 version with the speed booster coming out in M4/3? if focusing the light from a FF to apsc produces 1 more stop, focusing that light onto M4/3 should produce even more right? BMCC just needs an active M4/3 mount, and its done IMHO.

January 18, 2013

1
Reply
kevin

It's already said that the m4/3 Version won't have a different reduction-factor.
Getting this on a BMCC, you receive the Crop-Factor you would get from a Canon APS-C DSLR (slightly more, it's 1.633). So also the same DoF as on one of that Canons, which brings somehow the same look of a S35 Camera (at least it's close enough).

January 18, 2013

1
Reply
CameraRick

ah ok, so thats good right? would a S35 version costing 5-7 grand really be more attractive than a $3000 ~M4/3 with a 600 dollar adapter? i guess if they updated frame rates and ergonomics...

January 18, 2013

0
Reply
kevin

Would the crop-factor problems go away with using the Metabones Speedbooster adapter with this camera? You also gain a stop with that adapter.

January 18, 2013

1
Reply

Yes it will make the camera behave as if it had an aps-c ish sensor.

January 20, 2013

0
Reply
Ola

Me too, I preferred the uncorrected versions. The Sony looks bad though.

George - tshit.de/freshdailies

January 19, 2013

0
Reply

I am asking this not as a race thing but a color thing. Why aren't people doing skin tone test on darker skinned people hispanics, blacks. There might be videos but I've never seen them. We all have different shades and I always see white skin used in test can some tell me why? (Please leave the race debate out of it)

January 17, 2013

-2
Reply
Ty Stone

I'm not saying this video is bad I think it's a great video I just wanted to know why we don't see other skin tones.

January 17, 2013

0
Reply
Ty Stone

It probably happens for all sorts of reasons, but I was literally thinking the same thing when I started watching. I tried to do it on my last test session and it just fell through, but yes, you're right, it's definitely important to see how all different shades of skin react to a particular camera.

Zacuto did a good job in their last test mixing it up, and I believe that they had another test previously that did the same.

January 17, 2013

1
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

Thank you for understanding what I was asking. It clearly has nothing to do with race. As I said before these test are about TONEAL range not RACE range. With or without different races it's not going to change my mind on how great these cameras are it's a simple question that I know people will read with blinders on and miss the real question at hand.

January 17, 2013

1
Reply
Ty Stone

I totally agree, there is an absolute lack of supporting information (or even best practices or styles) for grading skin tones of different ethnic origins. You'll see what I mean if you ever get to watch TV here in Kenya, people generally look anywhere between orange and blue.

January 18, 2013

0
Reply

It's more of an issue of proximity, and White people tend to know more White people in their social and circles than non-whites to call on for all sorts of things, including camera testing. I do think redhead provides an interesting example with the contrast of her pale skin and fiery mane, but it would be helpful to see a larger variety of skin tones in future tests.

January 18, 2013

-1
Reply
Marc B.

It's a problem, but not a huge problem:
http://prolost.com/blog/2007/10/22/hue-are-you.html

What I would have loved is to see Flaat in that 5D2: I think it gives me the best skin tones the Canons can deliver:
http://www.similaar.com/foto/flaat-picture-styles/index.html

January 18, 2013

0
Reply

The grade is different in each with skin changing. I wish all test had a ten step grey scale so that neutral was a true neutral. Correct for grey and you can compare apples to apples.

January 17, 2013

0
Reply
ThunderBolt

I think the crackas pose more of a threat than darker skin tones do.. Lol.

January 18, 2013

0
Reply
F.M.L.

I do this every time I do a skintone test if i can, with either cameras or lights. I must start posting more of them. Doing a skintone/general test on the BMDCC tomorrow, but sadly, my black talent dropped out last minute. But its crucial for me. So, there are other people thinking about it!

January 17, 2013

1
Reply
marklondon

Despite the slight tint the bmcc owns imo. Great test thanks for sharing.

January 17, 2013

1
Reply
jordan carr

Ty, don't know that I can tell you why specifically, [though I have literally no black folks available as actors in my area here] but I will say that Darker skin folks are certainly easier to light. They're pigment absorb's so much nicer than Caucasians. IMO Caucasians reflect light hence you'll be better able to see what kind of 'crap' you got. HTH

January 17, 2013

0
Reply

This is actually a really interesting, if delicate, topic. Exposure can be an issue too.

January 18, 2013

0
Reply
Tom

White/Caucasian skin is by far the most difficult skin type to get pleasing results through digital cameras.
I found that dark skin is much better and easier, same for Hispanic and Asian.
There's this pinky tone to Caucasian skin that challenges most digital cameras (and that makes the Asian call Europeans/Americans 'Pig Faces' ;) )

January 18, 2013

0
Reply
hansd

I agree. Black skintones respond differently to light and when I get a chance I'll be doing more tests. It is a race thing... but not in horrid way we have come to know. :-)

January 18, 2013

0
Reply

Great question. Zacuto did a fantastic job of exposing for various skin tones. Hopefully, we'll start seeing more tests that will provide comparisons for noise, hues and exposures in the future.

January 18, 2013

0
Reply
Jerome

"I think it’s also helpful now to start looking at these tests not to see which camera is “better,” but to look at which images might work best for a particular shoot."

People seem to keep forgetting this. Never mind the peeping. Go with what pleases your eye the most and what fits your needs.

January 17, 2013

0
Reply
Chris H

@Ty
I have watched quite a few BMCC reviews and have never once thought about what color the person was, if there even was one. I dig the picture quality of the camera and love seeing it compared to other cameras. I understand you don't want this to be a "race thing", but you made it a race thing. Why does that matter?

January 17, 2013

0
Reply
Weldon

Yes TY, you managed to make it a race thing. Instead of looking at the different cameras, her not being black is what you thought of. Oh man oh man.

January 17, 2013

0
Reply
mike

How did TY "make it a race thing"? All he said was it would be nice to see tests for talent of different, perhaps darker complexions. He said it has less to do with race than it does with complexion. That is a totally fair, and interesting observation.

January 19, 2013

0
Reply
Markus

Here here I notice and appreciated the same thing in that student film with the new f5.

January 19, 2013

1
Reply
Anthony Marino

@Weldon, The reason I was there are people that would like to see how cameras work on different shades of skin. It's not about race like I said before. The reason I said it wasn't about race, because that would be the first thing people would assume, Which is clearly not the fact. I would like to see test done on skin TONES that's why I ask. You don't see that and I thought it would be a different thing to ask.

I think it's sad that people think it's about race when someone asks to see something different. It's a test of TONES not a test of RACE.

January 17, 2013

0
Reply
Ty Stone

@Weldon Sorry for the fat fingers when I was writing, but as I said it matters with TONE range and I've just never seen it done thats all. I'm not pissed or upset about it, I just wanted to see different test done with different skin TONES. If you don't ask the questions then nobody ever know to try it

January 17, 2013

1
Reply
Ty Stone

It's all good...

January 17, 2013

0
Reply
Weldon

I think one reason is because pretty much everyone knows what a fair skinned red head's skin tones are. Depending upon the camera profiles, you could be fooled into thinking that a hispanic person could be from southeast Asia. They have different skin tones. Brown isn't just brown. I know what you are asking for, but then you are shooting a multitude of people to get an across the board spectrum. Sometimes, time does not allow for that. The new F5 test that was put up a few days ago was heavily directed towards the black man that was in it.

January 17, 2013

-2
Reply
Chris H

Had no idea the FS100 can look that good.

January 17, 2013

0
Reply
john jeffreys

same here! Really a surprise!

January 18, 2013

0
Reply
Alex Mand

Same too. Seems really nice for me !

January 18, 2013

0
Reply
antoinelp

Don't sweat it @Ty I think it's a extremely valid point. I'm glad you brought it up.

January 17, 2013

2
Reply

The BMCC looked the best uncorrected and the most subtle when corrected - the other two cameras I thought looked horrible in the "corrected" close-ups - Hated the grade done on it which was totally unflattering to the model. Pale skinned women with red hair don't need this kind of heavy-handed grade. Of course, this is just what I find aesthetically pleasing. Like others above, I was surprised at how well the FS100 turned out.

January 17, 2013

-1
Reply
Neil

Maybe its just me but I'm not a fan of the "corrected" shots

January 17, 2013

0
Reply
Creep

All three cameras looked great. Hats off to Adam Roberts, I do believe he got the best out of each camera in this test.

January 17, 2013

0
Reply
Anthony Marino

why has no one compared the dynamic range of the BMCC to the RED or Alexa?

January 18, 2013

0
Reply
kevin

Just trying to put that test together in-house right now. Still bitchingly difficult to get a BMDCC here in LA to use.
We want to strap one to the top of an Alexa first.

January 18, 2013

0
Reply
marklondon

Send that test to no film school if you dont mind, i would like to see the results : )

January 18, 2013

0
Reply
kevin

Should have more then a RED, less then an Alexa; at least from the whitepapers (and w/o HDRx).
DR from the Alexa is insane, I dig the V3LogC.

January 18, 2013

0
Reply
CameraRick

cause allot of people won't be happy... :)

Kidding, but it'll probably be better than Red MX and still a bit less than an Alexa. Looks to be on par with native Epic DR... which brings up the point... does anyone even use HDRX? And are there actually videos of it in use?

January 18, 2013

1
Reply
bwhitz

FS100 + G-LOG Ultimate looks fantastic. Seeing some great stuff here and on dvxuser. This plus rumored significant firmware updates coupled with the metabones speedbooster have me leaning toward a FS100 purchase next.

January 18, 2013

0
Reply
Lane

Honestly thought this was a horrible comparison. I enjoyed the 5D MK2 with the Cinestyle Picture Style better than everything else I saw, and I know that not to be the case with the actual BMCC. The BMCC has a greater dynamic range that I feel like we didn't get to see at all, and it could have went EVEN FLATTER. There's 13 stops of Dynamic Range... Come on? That was the best that could've been done? And why only lighter skin tones? I don't create movies with just all white casts all the time. I need to see how these comparisons fair to other darker skins of other ethnic groups also.

As nice of an idea I thought the video was, I wanted to see a little bit more. That dynamic range wasn't the flattest it could've gone on the BMCC, and I really didn't like the FS100 look at all, which is surprising to me.

January 18, 2013

1
Reply

Well RAW isnt necessarily flat in it's curves. You can apply BMC Film Lut and get a very flat picture if thats your prefered look. This was not a DR test, but a skintone test ;)

January 18, 2013

0
Reply
Kandre

LOL

You are far to tense dude... relax. This was not a scientific test to show off what camera is best. It was a simple test for our own reference to compare 2 cameras we have shot on regularly with a new one that we are about to shoot on. I just shared it as I thought it might be useful to others and is seems it has... just not useful to you.

Emma happened to be available. If you wanna pay for a bunch of other models I'll reshoot the test for you....

January 18, 2013

-1
Reply

Yea, it disgusting that people make this about race when it's clearly just the odds of how things worked out at the time...

January 18, 2013

1
Reply
bwhitz

nice test, but don't understand why to compare the bmcc with a 5 year old camera like the 5D MK2 is and not
with the 5D MK3 or canon 1D X.

January 18, 2013

0
Reply

The test was for our own reference. Those are 2 cameras we have been shooting on regularly over the last few years so we used them as a point of reference. It was not some scientific lab test. I only shared the results as I thought it might be useful to others.

January 18, 2013

0
Reply

Since not the same lens was used for all cameras, conclusions are rather limited. The tonality of the lenses varies surprisingly among manufacturers, and maybe even among lens lines.

January 18, 2013

0
Reply
Thyl Engelhardt

True. However only the Canon footage was not shot with a Zeiss lens. As I said before it was never intended to be a scientific lab test.

January 18, 2013

0
Reply

While the BMCC footage is at a higher resolution and has much more dine detail, it has lead to some rather noticeable moire in the hair of the model in the close up shot.

Why bother going for the extra resolution if it's just going to cause you problems?

January 18, 2013

0
Reply

I really wouldn't call that noticeable. If you're looking for it you might just about see it. Anyone not pixel-peeping will never spot it IMO.

January 18, 2013

0
Reply
Luke

The BMCC and fs-100 where much under exposed comared to the 5d. That fact alone kills this test for me.

January 18, 2013

0
Reply

Hi, firstly sorry if this sounds like a silly question but please keep in mind I still regard myself a student and thus still learning - ty has reminded me of something I've wondered about, when lighting scenes, or even grading shots with actors with drastically different skin tones to each other (say one character with really dark tones and another with milder tones and one more perhaps even paler) would you need to give extra thought to camera profiles and lighting? Would it make grading or colour correction more troubling? Or am I over thinking?

January 18, 2013

0
Reply
C

Hey all.

This was my test. It seems to have caught a lot of peoples eye, I guess because everyone is so eager to get their hands on the BMCC.

It was just a quick test. I was in the studio and Emma happend to be available. I will be doing more rests and happy to share them.

One thing I have found is that the BMCC can do with about a stop more light than the 5D / FS100. I guess I've gotten used to slightly underexposing those 2 cameras as they don't handle highlight clipping very well.

All shots were based on the same light metered exposure. So the variance in exposure comes down to the cameras internal processing.

As for black talent... Emma just happened to be available at the time. I'll see what I can do on the next one. :-)

January 18, 2013

0
Reply

Thanks Adam for the tests! This is one of the first FS100 to BMCC comparisons I've ever seen...and honestly, I'm very impressed with how the FS100 holds up, so thanks for saving me some money ;)

Great to know about the exposure difference with the BMCC—As a 5DMK3/FS100 shooter, I too have become far too used to the 1-stop under exposure as the norm.

January 18, 2013

0
Reply

Pleasure. Glad it was helpful.

I do love shooting on the FS100 and you can get some amazing images out of it. It just does not handle highlights very well. I wish Sony would update the firmware with CineGamma (like the FS700) as I think that would help address this.

January 18, 2013

0
Reply

Oh, how I'd love the CineGamma option! And maybe an S-Log (I know, 8-bit 4:2:0 codec, but still!)? It's a fantastic, completely under-appreciated camera—but you're right, highlights (and banding) have always been the blemish of the FS100.

January 18, 2013

0
Reply

Nice, thanks. Would like to see a the BMCC and the Scarlet in such a battle.

January 18, 2013

0
Reply
Nanno

Each camera looked much better before the grade .

January 18, 2013

0
Reply
Tim

This might have to do with the DOF, but with the 5D you can't see the seam in the wall over her left shoulder, you can kind of see it with the Sony and it' very clear with the BMCC. But that probably has more to do with DOF than anything else.

I'm also not a fan of the color grading as it looks like it was trying to make the images look more real instead of cinematic.

January 18, 2013

0
Reply
Julian

I mean right shoulder, her right shoulder.

January 18, 2013

0
Reply
Julian

http://www.youtube.com/watch?v=mbJ3vacGOhw

Guess the camera and film stock...

January 19, 2013

0
Reply
Juhan-i

scarlet?

January 19, 2013

0
Reply
gerald

Ooozes Canon 5D all over it...

January 19, 2013

0
Reply
mike

If the people who find these tests lacking would do some testing of their own to satisfy their own curiosities and cover whatever they find lacking, then put them up for others to view (and potentially criticize), we would all benefit. Having done numerous tests for my self, I appreciate the efforts that go into the physical set-up and mental organization.

Hopefully when the bmcc is readily available we will see more variables covered.

Thanks for the tests Adam.

January 24, 2013

0
Reply
Granville

Seems to me a "test" like this is really rather useless. First there is no standard chip chart included in the shots. Second, there was never a side by side comparison, really hard to judge anything that way. Third, how accurate are the viewers' monitors? Is everyone set up to the same standard? That's why we have bars, scopes, and standard colour charts. Without these, it really comes down to your individual perception of the images. Sorry not trying to flame anybody here, but without standards, there is no accuracy to the comparison.

January 24, 2013

0
Reply
Alola

I haven't noticed it in other tests, but I definitely noticed it here. I normally really like the BMCC shots, but here I thought they all looked decent across the board. If this was the only test I saw, I don't think the BMCC would be on the list of cameras I would want.

January 24, 2013

0
Reply
Michael

The canon i preferred the not corrected one ! the close-up

Sony to much red for me but like the skin how look more soft .... maybe it's the lens.

Blackmagic i ptefered the raw but the Prores is nice also

January 24, 2013

0
Reply
Pierre Samuel Rioux

yes - it´s always green and a bit of blue!

January 25, 2013

0
Reply

Ok so, it's 2013 and I am about to buy a second 5d mk2 body, before they have all gone.

The Blackmagic isn't on the shelf yet and for the money nothing else comes close to the Mk2.

Get a Mk2, install magic lantern OS, stick a mosaic engineering moire filter in it and get the Lightform picture styles. Bingo, now go to work and earn the money you will need when they make something that really is a big step forward.

January 25, 2013

0
Reply
Chris K Jones

I think the grading in each case makes the image less attractive.

The 5D, uncorrected, was lovely
The FS100 footage was underexposed, but again nicer uncorrected.
The BMCC was nicer before the over-saturated correction too.

January 25, 2013

0
Reply
Lucas Adamson

Hmmm. It's a complexion issue not an ethnic or race issue. I've filmed dark, light and medium completed people. While lighting comes into play skin tone has almost nothing to do with how the camera performs in relation to the final color look.

I actually asked the same question in a post production class. The proff pulled up footage white, black, Asian and Indian people and put it through the vector scope. Everyone fell within the normal range for skin tone from the darkest to the lightest complexion. And I frequently work with a model whose skin appears to have a yellow tint (as opposed to red). Her image never has to be adjusted. The human eye is still more perceptive (or maybe overreceptive would be a better term) than the camera sensor.

Just a thought.

January 27, 2013

0
Reply
Armor1138

I can't believe anyone prefers the uncorrected shots. So you like footage to be flat and bland?

February 12, 2013

0
Reply
Muh

The speed boosters all have soft edges ALL OF THEM. The Chinese one is worse but the meta bones isn't going to have sharp edges.

July 5, 2013

0
Reply
RandyS.

Ok, how do you compare skin tones from a RAW image to anything other than RAW? Come on people ... let get real here!

August 24, 2013

0
Reply
Craig

which is best camera ( black magic 4k, sony fs 100, canon 5D!!!)

September 11, 2013

0
Reply
Bobby

One of those is not for sale yet, and no images recorded with it are public...

September 11, 2013

0
Reply

I will instantly grasp your own feed when i can't find your contact membership web page link and also e-newsletter assistance. Do you've got just about any? Remember to permit me to acknowledge in order that I could truthfully sign up. Thanks a lot.

October 11, 2013

0
Reply
HP