February 28, 2013

Blackmagic Cinema Camera, RED EPIC, and Arri Alexa RAW Camera Test Part 2: Low Light

This is a guest post by Cinematographer Ryan E. Walters.

Welcome to Part 02 of Some Like It RAW, where I am comparing the Arri AlexaBlackmagic Cinema Camera, and the RED EPIC. My goal for these tests is to explore how each of these cameras handles real world shooting environments. Part 01 explored how these cameras handle IR pollution. In Part 02, I test the limits of low light levels, or underexposure. Continue on to watch the 11 minute video, read my summary, and get the downloadable RAW frames from each camera.

I'd again like to thank the following people who helped to make this test possible:

Picture This Productions
(Provided the Blackmagic Cinema Camera)
Shawn Nelson
(Provided his EPIC)
Patrick Eggert
(Provided his Alexa)
Isaac Marchionna
(Assisted for the shoot)
Laurie Slater
(Was our Talent for the day)
180 Films
(Provided the Cooke 20-100mm Zoom)
Letus
(Provided the cage for the Blackmagic Cinema Camera)
Schneider Optics
(Provided the IR Tuner Kit)

The more I explore and experiment in my craft, the more interested I am in what happens at the limits of our recording mediums. I am definitely pushing these cameras to their limits in Part 02 & 03 of this series. :) The greats that have gone before us pushed their technology beyond where it was intended to go and broke with conventions to make their projects happen. (A great example of this is Barry Lyndon). As I am always desiring to push myself, I thought it was time to explore the limits of today's technology and evaluate the results. By testing in this manner, I'll be better equipped to experiment on a project, or push beyond conventions and hopefully create my own compelling imagery. If you are looking for clear winners and losers, then you've come to the wrong place, as that is not the point of test. The point is to see how these tools perform when they are pushed beyond what is normal, or considered safe. A better understanding of what happens at the limits not only ensures better images when I play it safe, but it will empower me to make more informed choices when I don't play it safe.

Here is the overhead diagram of how I lit these shots. The lighting did not change in-between cameras; it only changed slightly in-between setups. (I flagged off part of the background, and brought the key over a touch).

Below the 11 minute video you will find details of how I determined the underexposure values, a summary of my observations, and the link to the downloadable RAW frames from each camera.

How I Determined Underexposure Values:

I ran two tests to determine the underexposure values for these cameras. For the first test, I took the Sekonic Profile Target II exposures and loaded them into the Lite version of Resolve. Once there, I pulled up the waveform and looked for the last discernible chip that was viewable in the waveform. For the most part, this gave me my basis for my real world ratings.

Waveform from Resolve

Close up of Waveform Reading

For the second test, I loaded the files into Photoshop, and I selected each patch individually to acquire its Median value. As soon as this number stopped decreasing, I knew that I had reached the absolute bottom in exposure. These values tended to go much further than what the waveform in Resolve was telling me. While these values are interesting, practically speaking, they are packed too tightly together, and by the time the images are graded, they will end up as one value and essentially count as one stop. So I don't count them as being of any real world value. It is just a technical exercise.

Selected chip of a -8 EV exposed Sekonic chart in Photoshop

Results of selection

You can download the Profile Targets for all of the cameras from my Sekonic Profile Page. And yes, I freely acknowledge that the Xyla chart from DSC labs is the best chart to use for testing dynamic range. However, its price is outside of my personal ownership. I have found the Sekonic Profile Target II to be adequate for all practical purposes in camera profiling, and for calibrating my 758Cine & 478D.

Observations & Recommendations:

  • Lighting at high EI's is more about controlling spill and fill levels than it is about "lighting."
  • The more these cameras are pushed to their limits, the more expertise, and the greater care that is needed in the grade.
  • Each camera has its own look which may or may not suit a particular project.
  • My preference is to shoot wide open at higher EI's, which will reduce the contrast of the image, allowing for more subtle gradations in the shadows. (This is good news, as I am typically shooting wide open at high EI's anyway). :)

Alexa Underexposure Limits:

  • Maximum: -8 5/6
  • Usable: -5 5/6

Blackmagic Underexposure Limits:

  • Maximum: -7 1/6
  • Usable: -5

EPIC Underexposure Limits:

  • Maximum: -8 1/3
  • Usable: -5 1/6

The following is my personal taste, which may not align with yours. :)

When shooting on the Alexa:

  • Work with the camera at EI 800 to get the least amount of noise, even in an underexposed image. An underexposure of 1 stop still yields acceptable noise levels.
  • With an underexposed image at EI 1600, cleanup needs to be done, but it looks to be salvageable. (By someone more skilled than myself.)
  • An underexposed image at EI 3200 is unusable.

When shooting on the Blackmagic:

  • I feel comfortable working with an underexposed image at EI 800 or 1600.
  • An underexposed image at EI 3200 is unusable.
  • I think that my perception of this camera having less noise than the Alexa has to do with the desaturated image of the camera file, and the use of a wider lens. (By using a wider lens to get a similar field of view, the parts of the image that show noise are magnified less than their longer lens counterparts on the other cameras).
  • [Editor's Note: the BMCC does not have an OLPF.] The camera natively produces sharper images than the default files of the Alexa, or the Epic. (And the Epic is down sampled from 5k). The sharpness of the Alexa and Epic can be adjusted using the post tools provided by each company. (ArriRAW Converter & REDCINE-X Pro)
  • Some kind of diffusion is needed when filming talent to make the image more pleasing and forgiving.
  • The "overly sharp" images may also contribute to a "video" look if not addressed.

When shooting on the EPIC:

  • I feel comfortable underexposing the image at EI 800.
  • At EI 1600, and 3200, the results of the underexposed images are unacceptable and I don't think that they are salvageable.
  • Stay away from low color temperature light sources when shooting with the EPIC; it only exacerbates the color science of the EPIC. (Better colors and less noise comes from the light sources closer to 5,000k).

Downloadable RAW Frames (93MB): SLIR-LowLightTest.zip

What do you think? Do these results surprise you? Have you experimented with low light or underexposed images? Have you noticed the overly "sharp" look that the Blackmagic produces?

This post originally appeared on Ryan’s Blog.


Ryan E. Walters is an award-winning Oregon-based cinematographer. His work has allowed him the opportunity to travel worldwide in the pursuit of telling stories that are visually compelling. His experience includes feature films, documentaries, commercials, and shooting for Comcast, TLC, Oxygen, and the Discovery Channel.

Your Comment

44 Comments

Thanks again for a great post, Ryan.

This might be a whole 'nother post, but when you say "by someone more skilled than myself," where would one turn to look for such a noise ninja? I'm versed in Neat Video but beyond that... I would wonder where to go.

February 28, 2013

0
Reply
avatar
Ryan Koo
Founder
Writer/Director

Your welcome. Thanks for posting. :) I would turn to a professional colorist. I have seen them do thinks to the footage that I have shot that has really impressed me. Their skill and resources go way beyond what I can do even when using a similar tool set. They also know what high end tools to use to get a particular job done, where I would just be frustrating myself through the use of plugins ...

February 28, 2013

2
Reply

Yes, would second that regarding noise and FCP X, be interesting to know your preferred approach, ta

February 28, 2013

1
Reply
Jim

I do not have a preferred approach yet- at least not at the affordable level anyway. I have used the built in plug ins in apps, and programs like Neat Video, which work great when there are small amounts of noise, but make the image too soft and plastic looking for my liking when there are these levels of noise. The few times I have had the opportunity to work with a colorist and a Pablo system (ultra high end) I have seen amazing things done with the footage I shot - like real time wire removal from 4k footage.

February 28, 2013

-1
Reply

Ryan, thanks for doing this!

February 28, 2013

-1
Reply

Anytime- glad to help / put it out there.

February 28, 2013

1
Reply

Just to check i'm not being an idiothole or in case someone makes the same mistake this isn't a test of low light sensitivity right? More purposely underexposing all three to compare how they handle a rescue attempt in post, right? (my coffee this morning hasn't completely kicked in yet) As the BMC faired up a bit better than I'd expect it too.

I got hands on with it at BVE yesterday and have gone from being pumped for it to being a bit torn, on the BMD stand they had one set at iso 1600 f2.8 shutter about 240degree's and the exposure still looked quite dark was some truely horrific strobing during pan's on the CVP stand external display, not the inbuilt lcd but I don't know if maybe that was the monitor setup (none of the other camera's bar maybe the pmw-200 even approached it on their monitors) lovely image though hoping it was just a bad setup as it was nails on a chalkboard to my eyes

February 28, 2013

-1
Reply
Chris Lambert

Anytime glad to help and put it out there. :)

February 28, 2013

-1
Reply

You are not an idiot hole. :) The purpose was to see what happened at low light levels and underexposure. And then to see what happened in the grade.

The BMC is not a low light performer for sure- the 5D MKIII and C300 do a lot better in that department in my opinion. But I was pleasantly surprised with how the BMC did perform in such bad conditions. That is unfortunate the they had the shutter open to 240 at the booth you were at- they will produce some "interesting" motion rendition...

February 28, 2013

-1
Reply

Thanks Ryan good to know.

That was only on the BMD stand where the slow shutter demo was occurring where the camera was locked off unable to play with and lit by mostly incidental light of the arena with a couple of prop lights. iso 1600 f.28 and slow shutter is a little worrying compared to the fact I was in the same arena working a couple of weeks back with Sony EX3 F1.9 1/50 0db for broadcast and although it wasn't ideally lit I wasn't worried about applying gain was hoping the BMC would be more of a step up from my EX3 in lowlight.

I got hands on and did the pans at the CVP store stand and had a play and tried everything I could think of too reduce the effect. Changing the frame rate from 25-30-24p in case it was a monitor matching source issue, slowing the shutter, double checking the lens didn't have some kind of IS on it to resist it (50mm t2 prime so pretty unlikely lol) I'm hoping it was just a display issue as it didn't seem apparent on the tiny lcd on the camera

February 28, 2013

0
Reply
Chris Lambert

I'm confused- you slowed the shutter more then the 240 degrees? That would have made the problem worse (motion rendition). I would have thought that you wanted to make the shutter faster- bring it back to 180 degrees to get rid of what you were seeing. (But then that would have reduced the amount of light the camera was seeing.)

One thing I am learning with this camera especially, is that the RAW files make a big difference, and what you see on the monitor is not always the best representative of what is being recorded ...

March 1, 2013

0
Reply

Saying that the BMC is sharper than the Epic at 2k is odd. Epic is 5k so the sharpness at 2k is completely dependent on your scaling. BMC has aliasing at edges or sharp lines so that does give it a subjective "boost" in sharpness but it's only aliasing. The same way 5dmarkII looks sharper than the mkIII but it's only the aliasing giving that impression.

February 28, 2013

0
Reply
mikko löppönen

I was surprised how out of the box the BMCC is perceivably sharper then the Epic or the Alexa- not in a good way. Especially as the 5k is down sampled to 2k. I would have thought that the 5k would be sharper. Especially as I have not been happy in the past with shooting the Red camera at lower resolutions, as they get to soft for my tastes- especially when compared to their 4k image down sampled to 2k. Regardless of the reason for the sharpness, I think it is worth nothing the real difference so that it can be addressed in camera or in post. I have not like The "sharpness" of the footage that I have seen from this camera and I'm interested in exploring ways around that...

February 28, 2013

0
Reply

"My preference is to shoot wide open at higher EI’s, which will reduce the contrast of the image, allowing for more subtle gradations in the shadows."

How does shooting wide open reduce the contrast of the image? NEVER heard that before!

Also, this continuing fetish for "low-light" is pretty offensive to the art of cinematography.

February 28, 2013

0
Reply
James

Contrast and iris diameter are linked to one another as it is an "artifact" of lens design. It happens on all lenses, but it is more pronounced on older lens designs. It is less of an issue on modern glass. Check out the Portland Lens Test I did with a bunch of mid range modern lenses and you can see it there too.

I disagree, Low light is not an offense to cinematography- it is only an offense if it used as a way to excuse for the lack of lighting / planning. Good lighting is good lighting at EI 200 or at EI 2000. At EI 200 it is more about adding light, and at EI 2000 it is more about removing light. But in either case it is about control, purpose, and thought. Unfortunately, people think that you don't have to light at low light levels- and that thinking is an offense to cinematography- at least in my opinion anyway...

February 28, 2013

2
Reply

Gordon Willis, ASC. 'Nuff said.

February 28, 2013

0
Reply
alex

Good point! :)

March 1, 2013

0
Reply

No it's USA and Europe, you don't need to travel half of the world to feel oppressed. Learn the history of your own homeland, pal. You would be surprised.

February 28, 2013

0
Reply
Natt

Thanks, interesting.

At first I really liked the BMC image but the more and more footage came out, I too started to dislike it's special kind of sharpness. I guess there are always ways to reduce it though.
Overall 3 great cameras, if you can't do any good with either of them, then it's time to change occupation :)

February 28, 2013

0
Reply
hansd

How true and well said. :)

March 1, 2013

0
Reply

Ryan, have you messed around with a KineRAW S35 at all? I realize it's not a common camera yet, but considering its features and price point, I would be curious to see how it compares to the other cameras.

February 28, 2013

-2
Reply
Andrew Swan

I would love to play with that camera, but I have not had the opportunity to do so yet. :( I just don't have the right connections ... If anybody wants to hook me up, I'd love to put it through its paces. That camera does interest me a lot ...

March 1, 2013

1
Reply

Ryan, I heard Denver Riddle from ColorGradingCenteral will get a KineRAW S35 soon.

March 1, 2013

0
Reply
Jason

Yeah, seriously. You think the US doesn't oppress people? Look at native americans, black people, women, gays, the uneducated, the poor, the mentally handicapped (is that the proper term this year?)...pretty much anyone who isn't a white educated male. The US likes to trick people into thinking it doesn't oppress, and that we have free speech, but how free can our speech be when all of the major news outlets are owned by the same group?

February 28, 2013

2
Reply

I downloaded the raw files and tried comparing the three cameras but found it really difficult. When loaded into photoshop the BMC looks so much brighter then the Epic that it made me think something was off, like the BMC had more light than the Epic or something.

When loaded into Resolve things don't get much better because each camera has a completely different starting point in raw. But still the Epic looked a lot darker than the BMC, with the Alexa somewhere in between.

February 28, 2013

1
Reply
Sander Kamp

I'd stay away from Photoshop- from my experience, it is hard to use / get the most out of the RAW files. Your best bet would be to use each of the programs supplied by each camera manufacture, or bring them into Resolve, or Adobe Premiere. You are correct- that each camera has a different starting point in RAW. That is where you will need to go in and work with the metadata of each camera. Switch them all over to log, and then get the white balance as close to 1600k as possible. (I believe that in Resolve, the Arri file will only go to 2000k.) Each camera will still look different- which is due to their own color science - and why I HIGHLY recommend choosing the camera based on the look you want / need for a project ...

March 1, 2013

2
Reply

Once again, great work.

I'm not surprised you had trouble matching the Epic to the others. My guess is that it is the spectral response of the Epic CFA, and not your ability.

I am a bit surprised by your comments on sharpness. There is a wide choice of lens filters available (and widely used) to soften close shots, which can be removed for wide angles. But a soft image is hard to fix.

Cheers.

February 28, 2013

0
Reply
Noel

Thanks Noel. :)

As for the soft image, that came from the default settings in both the Alexa and Epic. In their software, they have options to increase sharpness, so I know that there is a lot of room to play with that in post. From what I understand, their thinking is to reduce aliasing and moire problems in the original image, and then let us dial in the amount of preferred sharpness in post. (And there is a lot to work with in those files.) The BMCC on the other hand, doesn't allow for that kind of flexibility ... and is "overly sharp" right out of the gate. (At least for my tastes anyway.)

March 1, 2013

0
Reply

I find it interesting that we have to blow up an image and inspect it that closely to see the difference between a $3K camera and a $100K camera.

When the client's paying for the rental, rent an Alexa. When you need a camera for your own projects, and it's your own cash on the line, I can't imagine why anyone would pay that much more for that little of a difference. It's not gonna make your script any better.

I'd say the Alexa looks maybe $2K better than the BMCC, definitely not $97K better.

February 28, 2013

-1
Reply

Horses for courses. :) There are options in the Alexa and Epic that you cannot get in the BMCC. (Like variable frame rates- over or under.) And there is some big differences when it comes to how these cameras handle over exposure- which will directly relate to the lighting package needed ...

I do agree with you about renting and owning. :) I'd much rather own a BMCC and rent the other two when I need them for a project. :) ( http://www.ryanewalters.com/Blog/blog.php?id=6306591485559567041 )

March 1, 2013

0
Reply

Thanks for all your tests Ryan. We've done a couple unscientific tests with the BMCC and overexposed it by 5 stops, then pulled all the details back in, in post. And that was shooting in ProRes.

I read the post you linked to in your comment. The c/y lenses sound like a great way to put together a lens package for a reasonable amount of money. Any further tips regarding the c/y lenses? Specific brands, specific lenses, what not to buy, etc.?

Just FYI, I don't believe the Sony battery you recommended for the BMCC will work, as it's only 7.4v, and the BMCC needs 12-30v.

Again, thanks for all you do for the filmmaking community. Very informative. I'm looking forward to Part 3. How many parts will there be?

March 2, 2013

0
Reply

really informative, thanks Ryan!

February 28, 2013

1
Reply

Anytime- glad to help. :)

March 1, 2013

0
Reply

I am waiting on NAB 2013 to make definite decision on what camera to buy. Now I am leaning in to buy a bmc, but perhaps good things will happen at NAB.

February 28, 2013

0
Reply
Ano Nymous

NAB 2013 will be interesting for sure- who knows what people have up their sleeves. I am VERY interested to see what the new Red Dragon sensor brings to the Epic ...

March 1, 2013

0
Reply

So the BMCC is SHARPER AND LESS NOISY THAN THE ALEXA AND EPIC?!?!?! - Now that's a headline! Thanks for the test..another BMCC justification video for me.:D

February 28, 2013

0
Reply
Quobetah

Depends on your tastes, and your needs. Personally, I don't like the overly sharp images on the BMCC, and I'm looking into ways to reduce that artifact. In that aspect, I like the Alexa and the Epic better ...

March 1, 2013

-1
Reply

Yeah I guess...the more I pixel peep movies shot on film right now..the more I see their edges beeing soft..which is part of their look. Recently saw Flight though..and my jaw dropped. Its just so sharp and clean! (Shot on RED Epic) I now can confim that I abosolutely love the look of digital!:D

March 2, 2013

0
Reply
quobetah

Well, he did say that because of the lens he used, that could have contributed to the less noise factor being seen. Can I suggest using the same focal length lens and just move the BMCC back further from the subject, so the shoot can match with the other cameras.

March 4, 2013

1
Reply
Sidmar Holloman

Ok guys, all the 3 cameras are very good cameras. If you are a really pro cinematographer you should know that no matter what you will always need some form of light to do your shooting, weather is a feature film, documentary or music video. It is up to you and the story and how you want to lit your scean. Sometimes you will see that it seems that a movie is shot in the dark! but it is not, they have to use some light and it also depend on the man behind the camera to bring down his f stop to make it look like it is been shot at night and then the rest is done on post or by a colorist. All pro cameras are good, but it is up to you to know what kind of shot you want to archieve. You can get the best (low light) camera out there but if you don't use some form of light you will always get camera noise! All of you should know that video and light go together in order to do a good job or tell a story by manupleting with light. Thanks everyone, I hope I did not offen anyone.

March 1, 2013

-1
Reply

Very true statements. Low light DOES NOT MEAN NO LIGHT. I can't agree more. :) I am interested in seeing how much light I do need, and at what light levels and exposure levels are acceptable to me- which is the point of this test for me. From there, it is all about shaping the light (though adding or taking away). Great cinematography is about control, not about just getting an exposure. :)

March 1, 2013

0
Reply

Hi, sir I would like to know which is best colour correction software ?

March 5, 2013

0
Reply

Nice use of the Shot Designer app Ryan. Been a godsend on a feature at the moment, hotsyncing with dropbox, designing all the shots and blocking and taking to set on the ipad. Ace.

March 2, 2013

0
Reply

Ryan, great info, can u explain the figures onyour underexposure limits please, not understand .

Many thanks
Chris

July 28, 2014

-1
Reply
Chris White