Superproducer Ted Hope on...

December 24, 2013

Digital Bolex D16 Review Part 2: First Footage & D16 vs. Blackmagic Pocket vs. SI-2K Mini Test

I shared my initial thoughts on the camera last week in Part 1, and since then I've been able to spend a little more time with the camera. The firmware I had when first shooting with it was slightly more beta than the firmware the camera currently has, but even in that time many more things have improved. I will elaborate on those in future posts, but for now I thought it would be a good idea to post some footage and talk about the image. With help from Adam and Dylan from Rule Boston Camera, we compared the D16 to the Blackmagic Pocket Cinema Camera and the SI-2K Mini (which is in a housing known as Black Betty). Check out that test below, as well as some of the first footage we've shot near the bottom of the post.

Digital Bolex D16, Blackmagic Pocket Cinema Camera, & SI-2K "Black Betty"

Here is the comparison test, again thanks to Rule for the help putting this together -- and of course thanks to Joe and Elle from Digital Bolex for providing the camera. Our lighting setup was daylight LEDs in the front, and a Tungsten head coming in from the back, so this test was a mixed color temperature scenario:

All of the cameras were shot RAW, and everything was kept in DaVinci Resolve, even titling, to make the comparison as fair as possible. Each shot was put in BMD Film and then graded to match. It should be noted that while 800 ISO was in the firmware at the time of this testing, it was pulled out when I sent the camera back for a newer version. As of right now the Digital Bolex team is still working on getting 800 to look a bit cleaner, but it was fairly noisy when I played around with it.

What's striking to me is just how flat the log is from the Blackmagic Pocket Cinema Camera. It takes quite a bit of pushing to get that camera back to something usable, whereas the other two cameras didn't go quite as flat when set to BMD Film. The Blackmagic is the cleanest of the bunch, even at 800 ISO, but it's also the newest sensor tech of any of these cameras. We had to move the frame quite a bit to get the Blackmagic to match as that sensor is bigger than the D16, and much bigger than that of the SI-2K.

The moire on the spinning camera in the scene was very apparent with the Blackmagic Pocket. While some of that could be attributed to moving the camera and slight changes in focus, it was noticeably bad on that camera even in the monitoring.

Dynamic range wise, the Pocket also looked like it had the edge on the Bolex, and a significant edge on the SI-2K. I don't think you'll have problems with either the Pocket or the Bolex for even uncontrolled environments.

D16 Footage

We also were shooting a quick short film with the camera, which has seen all sorts of scheduling issues, but at least for the time being I thought I would put up some flat and un-flat footage so you can see what the D16 is doing. At some point I will have some color graded stuff handy, but I know people always seem to be asking to look at log footage with new cameras, so it made sense to me to provide some. Here is Megan Phelps standing in for us, with our own Dave Kendricken also operating camera (everything in these clips is daylight balanced shot at 200 ISO, with Zeiss Super Speeds). Vimeo did a pretty good job killing the sharpness so it's probably not a bad idea to download the original:

To me the image looks a bit less digital than some other cameras, and I think part of that is due to the CCD sensor. The texture is different from CMOS sensors, and certainly at lower ISOs this can be a great look if the project you're working on benefits from it. I have always been partial to CCDs and the way they produce an image, so I'm probably a little biased. Either way, you can do so much with these images when you bring them in post, and it's pretty easy to keep everything in a gradable range on set -- not blowing out highlights or losing shadows.

I think the image is much closer to Kodak Super 16mm than anything digital I've used so far, and that makes a lot of sense considering Kodak developed the technology for the sensor (the sensors are now made by Truesense Imaging). The life to the image is very noticeable when you're playing with it in post, but I do prefer the camera at lower ISOs. 400 is usable, but it is a little noisy, and as said above, 800 wasn't in the camera the next time I used it. I'm going to do more tests with all of these ISO ranges as well as color, so there will be a lot more testing in the coming days and weeks.

I'm hopefully going to have some RAW footage or single DNGs up to download and play around with, but for now at least you can grab the file straight from Vimeo and mess around with it.

CMOS vs. CCD Gain

The nature of the D16's CCD sensor means that RAW isn't just metadata when you're dealing with ISO. Unlike the CMOS sensors of the Blackmagic (and to a lesser extent the SI-2K), the CCD sensor does not really have a native ISO, as some gain needs to be applied or taken away for every ISO in the camera. This means that choosing an ISO will have an effect on your final output, but with something like the Pocket, you can set it to whatever you'd like and change it in post.

The power of RAW means we can get them all to look very similar without too much effort, but it's something to keep in mind when you're shooting with the D16.

Links:

Your Comment

121 Comments

after watching the first 49 seconds.....BMPC wins all day long :)

December 24, 2013

0
Reply
Alex

I was thinking the same on the test footage...but then the footage of the girl looks great.

December 24, 2013

-2
Reply
Muh

yes it totally wins... at looking like brittle video. I kid, they are both great cameras, but I cannot deal with rolling shutter and the way cmos sensors render motion. This looks much more like film.

The BM4k is a great contender though!

December 24, 2013

1
Reply
elpert

after watching the first 49 seconds…..D16 wins all day long :) absolutely !! unfortunately i don't like the shape of the D16

December 29, 2013

0
Reply
Nick

BTW i love BMPC and price

December 29, 2013

0
Reply
Nick

The tools available to us indie film makers, now days, are exciting.

December 24, 2013

-1
Reply
David

Ya thiiiiink!?!?!?!?!?

December 25, 2013

0
Reply
A.J. Minnick

Yeah !?!?!?!?!?!?

December 26, 2013

0
Reply
David

I love the motion and event he color. But my biggest concern with this camera right now is dynamic range. In every clip I've seen, the highlights clip pretty hard.

December 24, 2013

0
Reply
Kenneth Merrill

What an amazing achievement, creating a camera from zero. Congratulations! (wish I had $3000)

December 24, 2013

-2
Reply

There is a dimensionality with the D16 that the Pocket doesn't have, imho. Also, look at the film credits Evelyn Keyes, the Pocket looks hazy in comparison. Overall, the Pocket's footage seems to have some sort of veil on it compared to the D16.

December 24, 2013

2
Reply
pask

Shouldn't the comparison be between the D16 and BMCC?

December 24, 2013

-1
Reply

Yes, and the 5D3 in RAW if you're going to be in the correct base price bracket, but its a start.
Thought both the Bolex and Pocket looked good and very useable.

December 24, 2013

1
Reply
marklondon

We wanted to compare only cameras that could take Super 16 or older C-mount lenses for this, as that's more why you'd be in the market for a Digital Bolex.

December 24, 2013

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

Understandable

December 24, 2013

0
Reply

I get the impression DB never intended on reaching a larger market.

December 24, 2013

0
Reply
Gene

I don't buy that reasoning in the slightest but hey, the footage looked good given the limited time/test. Like its CCD cousin the Ikonoskope it handles greys so beautifully.
Look forward to seeing it go toe to toe with the other BMs, sensor sizes be damned :-)

December 25, 2013

-1
Reply
marklondon

Haha you can believe whatever you want but we picked the other two cameras for the test because they both shoot RAW and are close to native for Super 16mm lenses. If we had an Ikonoskop handy we would have tested that too.

December 25, 2013

-2
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

Love it but would be cool to get 50/60fps raw at 2k..

December 24, 2013

0
Reply

If it did that I'd probably buy one.

December 25, 2013

-1
Reply
marklondon

"It's great but could be better" ...brilliant comment.

December 25, 2013

2
Reply
Christian Anderson

That Bolex sure makes a pretty picture. The motion has a very nice temporal feel reminiscent of film footage. Must be do to the global shutter.

We live in exciting times. Just a few years ago $3500 dollars maybe got you a nice matte box. Now you can get a camera that produces a picture good enough to shoot a movie.

December 24, 2013

1
Reply
Marc

Even when the camera is free, this will still be a very expensive hobby.

January 29, 2014

0
Reply
Little Mermaido

Joe, if you could test it, I'd be interested in seeing how well the CCD holds the color saturation of different in focus colored light sources versus the CMOS of the other two cameras. For example, red traffic lights and automobile tail/brake lights at dusk or later at night often render white or pink with a pink halo. It's a dead giveaway for digital and its Achilles heel, IMO. I know it's a dirty word for some but film tends to saturate the color way before blowing out to white under bright colored lighting conditions.

It would be nice for digital to overcome that hurdle at some point and I'm wondering if the sensor type makes any difference. It's not a huge deal but it's a pet peeve of mine to see white blobs with colored halos on traffic lights, neon signs and the backs of cars at night where it should be a fairly saturated blob of the appropriate color.

December 24, 2013

1
Reply
Brian

Yeah if I can remember I will try it. I know exactly what you're talking about.

December 24, 2013

0
Reply
avatar
Joe Marine
Editor-at-Large
Shooter/Writer/Director

Cool. Thanks.

December 24, 2013

0
Reply
Brian

I would be interested to see that as well. Here is an example of that problem in some footage shot on the RED Epic: http://www.flickr.com/photos/adrianoc/11552163556/
As shown in the RGB parade, even though the values are well within range, the tail lights of the car look quite blown out and video-like.
I, too, hope that digital will overcome that hurdle at some point.

December 25, 2013

0
Reply

If the values are not clipped in the raw data, then they are recoverable and gradable. Tail lights and stop lights don't have to look blown.

December 26, 2013

1
Reply
David

That's exactly the problem. If you look at the RGB parade you'll see that none of the values are clipped, yet the tail lights of the car are rendered white and with no recoverable details.
For further proof of how digital sensors struggle with brightly colored lights have a look at RED's latest showreel: https://vimeo.com/63060016 - At 4:07 there is a shot of the White House which shows the same problem.
I'd be curious to see how other digital sensors, including the CCD in the Digital Bolex, cope under the same lighting conditions.

December 27, 2013

-1
Reply

It really should be tested with product in the same price range - D800, 5D MK III, BMCC and so on.

December 24, 2013

0
Reply
DLD

I would say it should ALSO be evaluated among similarly priced cameras in terms of a general "one camera to rule them all" approach. However, I think testing it against other cameras with similar sensor sizes and available lens choices, even at significant price differences, is informative. If S16 or thereabouts is the look or format you've settled on (for whatever reasons), it lets you judge performance based on that criteria and helps you draw conclusions about how much cost factors into achieving the look you're after.

December 25, 2013

0
Reply
Brian

When talking $3.5K price range cameras, they all have significant compromises of one kind or another. There is no one to rule them all, just choices based on what set of compromises best fits an individual's needs.

December 26, 2013

0
Reply
David

Many people can only afford to buy one camera and have to make it work for the vast majority of what they shoot. When I said one camera to rule them all, I meant all shooting situations, not all cameras. I know every camera comes with certain compromises, especially at that price point.

December 26, 2013

1
Reply
Brian

Thanks for doing both ungraded and with some grading.

This camera looks able to compete with the Pocket and BB. We'll have to see what it does with the 5Dii/iii, GH2/3. Then against BM2.5K, BM4K, etc. Don't know if it will hold up to higher K's. Will someone eventually compare it to ARRI and Red?

December 24, 2013

1
Reply
Gene

the D16 footage looks absolutely stunning, perfect for an indie film. buying one now.

December 24, 2013

-1
Reply
Hildebrand

The D16 footage reminds me of the Sony F35 and of course the Ikonoskop a-cam DII. All three are CCD and the D16 and A-Cam use variants of the same Truesense sensor. The F35 had a really interesting striped pattern CCD developed by Sony.

One nice thing about the F35 was that its noise often looked like film grain. Hopefully this will be the same for the D16.

December 25, 2013

0
Reply
Marc

Yeah, the F35 was such a great cam.
Pig to work with though. You can get one very cheap!

December 25, 2013

0
Reply
marklondon

Great post guys this is such an interesting camera. I would love to play with this camera. If you guys plan to shoot your short film in Boston I can help out. I know a lot of artists in the area and would love to work with you gentlemen.

December 24, 2013

1
Reply
Caleb

Very impressed with the SI-2K! This surprises the hell out of me as I expected it to perform bottom of the list but dang! Has the almost detail of BMPCC without the horrific moire (before all the fanboys jump down my throat I LOVE that camera, but seriously - that colour pattern is pretty gross).

December 24, 2013

0
Reply
Kraig

I felt the SI-2K had the best dynamic range of the bunch.

December 26, 2013

-2
Reply
Nick

Eh, to me, this looks so much better. [ https://www.youtube.com/watch?v=Ggb9SQJGW3c ] - D5300 into Ninja 2 in 422 ProRes.

December 25, 2013

-1
Reply
DLD

Are you kidding? that looks awful... you need a new set of eyeballs.

December 25, 2013

0
Reply
jason

LOL

December 25, 2013

1
Reply
Caleb

DLD I've always been wondering if you were a troll or not.....aaaand after this comment I can ensure that yes you are...!!!! it's like if you randomly have take a youtube video and post it here.....but the worst part was all that interlacing going on there......XD

December 25, 2013

2
Reply
jesuan

Yeah. Either my sarcasm detector is on the fritz or your sarcasm generator is out of alignment because I can't believe you think this footage looks better than any of these three cameras. Say it ain't so.

December 25, 2013

0
Reply
Brian

Oh, I do. I think the colors are far more vivid and the sharpness is similar and that gives the Nikon footage a far more pleasing overall look. The new D5300 also has a new powerful processor and the time tested 24 MP sensor. The artifacts, wherever they may take place, only bothered me when the camera itself was moving. Otherwise, based on the stationary shots, it looked better. So, take CCD or high FPS cameras when you're doing whip pans. When you're shooting an interview, whip pans won't be on your shot list.

December 25, 2013

-2
Reply
DLD

Translation: "The Nikon is better when nothing is moving"

Nobody cares about that.

December 25, 2013

0
Reply
Christian Anderson

Nikon does make good video:

Nikon D800: [ https://www.youtube.com/watch?v=B_zZuQzPMe8 ]

Two 1080p video samples from the $307.00 Nikon 1 V1: [ http://www.youtube.com/watch?v=1OvV00-lIIk ]
[ http://www.youtube.com/watch?v=xqebWbwvIW8 ]

some 4K bursts from the Nikon 1 V1 http://vimeo.com/72247280

I'm not saying these samples are better/worse than the 3 cameras in this post. Just wanted to show Nikon does make good video, their strongest point probably being the appealing color palette.

December 25, 2013

-1
Reply
Gene

yeah, no.

December 25, 2013

0
Reply
alan

LMAO. DLD/Gene crack me up.

December 26, 2013

0
Reply
Mitch

Pages