Venice Beach on a Digital Bolex

[Robert Ruffo]

2001 called, they want their harsh video-ish highlight roll-off back.

 

[combatentropy]

DR is nowhere near film/Alexa/Red quality, not even a ballpark. Where you see it is not in the shadows, but in washout and overblown highlights (which are horribly and harshly clipped). Maybe if they used polarizer it would come out better. Skin tones isn't everything, the clouds are just just white blobs, without any detail.

Then when you shoot with the Digital Bolex, expose for the clouds. Then, in post, raise the mids and shadows (but not the highlights) until the skin tones are at the level of brightness that they should be.

I'm totally armchair quarterbacking, because I don't own a camera and haven't shot with anything but my phone for the past few years. But from 1990 till about 2004 I shot over a hundred videos --- mostly weddings, but also a few promos and a lot of for-fun short movies. I also went to film school. I quit working in film and video about 10 years ago, having suffered terrible burnout. But anyway, maybe someone in the field can verify.

There is already at least one test already done in the field. In 2004, Roger Clark used something called "the photoshop shadow tool" to bring detail out of the shadows of a Canon 1D Mark II. He measured its dynamic range at nearly 12 stops. He also said it beat Fujichrome Velvia slide film and Kodak Royal Gold print film. See his article, Dynamic Range and Transfer Functions of Digital Images and Comparison to Film.

The pictures that we're getting from video are generally of lower dynamic range than film. However, at the same time, they have drastically less grain. I mean, they're totally clean, to my eye. It seems to me that these are two sides of the same coin. I believe that digital has a lot of detail in the shadows. I believe that most camera makers are hiding it with their default presets, because maybe they think the average consumer will object to any grain at all, but if dynamic range is limited, the average consumer will just blame themselves or the lighting. I'm not sure. But there are some people who say that a little film grain actually looks pleasant. So much so that some of us are artificially adding grain to video. At the same time we're wishing that video had more dynamic range. I believe that if you just raise the shadows, as Roger N. Clark did, that you will get both. You will get more dynamic range and you will get grain. I don't think you have to make it look like a grain silo explosion to get comparable dynamic range to film. But I really would like to see more people who have RAW cameras and who have the post tools to try it. Try just raising the shadows enough to bring the grain up to the same as 35mm or even 16mm. Then see how much dynamic range you get. It may not be the 14 or 15 stops often quoted for film, but you might just find another stop or two.

 

[Razz16mm]

My reference to the HPX was two things:

One being the fact that the depth of field look here is something you could have got with any 1/3" camera. That's not a knock, I'm just saying that a prosumer CCD camera would have been similar in look and DOF. If it was mostly shot at f16 (which I had guessed) well then that explains why the DOF is the way it is.

Secondly, the D16 realistically is only really 1-2 stops over some cameras and to me this video does not show a lot in terms of dynamic range that couldn't be achieved with a handful of cameras which offer 10-12 stops of dynamic range, and even use a quality-limiting codec for encoding. Some of the shots in this video (like the interviews) are controlled with a reflector in order to reduce the overall scene contrast ratio and help out the balance. Its not like it was shot without a reflector and then pushed in post drastically.

I think in general however the video was over-exposed by a couple stops. There appears to be some blooming apparent in the wide shots where you see a lot of sky or sand which might have happened in post (not sure), and in the interviews the white tent in the background is pretty hot and the highlights on the skin look clippy. This is not the fault of the camera, probably just the result of a slight over-exposure in-camera.

Knowing this is a RAW camera and shooting outdoors under the sun if it were me I would have under-exposed everything by at least 2 stops to help allocate some of that dynamic range towards the bright tones and highlights, then brought the skin tones back up in post to where they should be.

I would love to test one of these cameras to see how it compares to some of the other cameras I've worked with on-location and in post.




I didn't read the background info on the video by Kurt, so it is good to know how it was shot.

Rich colors and non-blocky shadows are definitely an attribute to shooting RAW but again depending on how much you're really really pushing in post most grades can survive 422 8/10-bit. I'm not saying it is the same as RAW, just saying you don't always need the vast color space and bit-depth of RAW.

As for Kurt's statment about skin falling apart with 8-bit codecs, judging from his interview shots he exposed for the skin tone and not the background. So had he exposed for skin-tone using any other camera like the HPX, AF100, F3, 7D, 5D (all 8-bit 422 or 420 codecs) would he really say none of those cameras could produce pleasing skin tones? Interesting.

I agree that pro grade video cameras shooting 4:2:2 high bit rate broadcast grade i-frame codecs can produce comparable footage. The closest general aesthetic comparison would be a decent 2/3" format broadcast camera or S16mm film camera. That is precisely what is missing in the under $10k market. There is a gaping hole between low end small chip and large chip cameras that is currently only served by the original BMCC and the pocket cam, with a host of compromises necessary for price targeted mass market retail.

The BMCC can produce slightly better overall DR, but that comes with aliasing artifacts and rolling shutter issues on fast movement that the D16 does not have. It is also significantly noisier in its lower range. The D16 is to me a much handier camera that can be used with little or no external rigging or work arounds to make a practical small crew/solo operator camera for the kind of independent creative films that 16mm shooters have done for decades.

 

[Kholi]

I agree that pro grade video cameras shooting 4:2:2 high bit rate broadcast grade i-frame codecs can produce comparable footage. The closest general aesthetic comparison would be a decent 2/3" format broadcast camera or S16mm film camera. That is precisely what is missing in the under $10k market. There is a gaping hole between low end small chip and large chip cameras that is currently only served by the original BMCC and the pocket cam, with a host of compromises necessary for price targeted mass market retail.

The BMCC can produce slightly better overall DR, but that comes with aliasing artifacts and rolling shutter issues on fast movement that the D16 does not have. It is also significantly noisier in its lower range. The D16 is to me a much handier camera that can be used with little or no external rigging or work arounds to make a practical small crew/solo operator camera for the kind of independent creative films that 16mm shooters have done for decades.

Slightly is a bit understated. The DR gap is pretty obvious to me, although it still looks fine. 2.5K resolves more detail (at the expense of aliasing/moire until mosaic releases), more recording options on-board, on-board HD-SDI. Wouldn't be surprised if the Pocket Camera out-resolves it, and I'm pretty confident in greater DR. May or may not be as clean, but it certainly is cleaner than the 2.5K counterpart.

I don't see this camera as a "no-compromises" alternative, it has its own set of will-be issues as well.

This footage does look better, but leaves much to be desired. Well, to me, anyway.

 

[Razz16mm]

Slightly is a bit understated. The DR gap is pretty obvious to me, although it still looks fine. 2.5K resolves more detail (at the expense of aliasing/moire until mosaic releases), more recording options on-board, on-board HD-SDI. Wouldn't be surprised if the Pocket Camera out-resolves it, and I'm pretty confident in greater DR. May or may not be as clean, but it certainly is cleaner than the 2.5K counterpart.

I don't see this camera as a "no-compromises" alternative, it has its own set of will-be issues as well.

This footage does look better, but leaves much to be desired. Well, to me, anyway.

There are inherent compromises in any camera in this general price range.
The engineering team completed initial ISO and color calibration last week so it will be interesting to see what the next round of footage looks like.
98% accurate base ratings for middle gray reference to a Sekonic L758 at 100,200,400, 800 ISO. This is one camera that you can expose by light meter out of the box.

 

[Razz16mm]

Here is a BMC pocket cam sample shot under similar outdoor conditions. No filters or lighting control. Small aperture.
I haven't seen anything near the overall IQ of the D16 coming out of this camera so far. It has serious issues with clipped highlights.
Maybe when it gets upgraded to raw there will be some improvement.

 

[Kholi]

Here is a BMC pocket cam sample shot under similar outdoor conditions. No filters or lighting control. Small aperture.
I haven't seen anything near the overall IQ of the D16 coming out of this camera so far. It has serious issues with clipped highlights.
Maybe when it gets upgraded to raw there will be some improvement.

Hmmm doesn't seem similar to me: there's far more contrasty situations in this video than Venice Beach, which is a wide open space. Also, why is "no ND" an important detail?

I have Pocket Camera footage that I shot m'self at Venice Beach, though with Anamorphics and NDs, so I guess it doesn't qualify.

I'm sure Hot Rod could organize a comparison, but I imagine that wouldn't happen until the camera was actually shipping.

 

[Razz16mm]

Hmmm doesn't seem similar to me: there's far more contrasty situations in this video than Venice Beach, which is a wide open space. Also, why is "no ND" an important detail?

I have Pocket Camera footage that I shot m'self at Venice Beach, though with Anamorphics and NDs, so I guess it doesn't qualify.

I'm sure Hot Rod could organize a comparison, but I imagine that wouldn't happen until the camera was actually shipping.

The footage was shot at a small aperture so one would expect to see similar softening from diffraction.

 

[Kholi]

The footage was shot at a small aperture so one would expect to see similar softening from diffraction.

The latest DB footage looks nice (Portrait Day). Very different feel/texture, CCD depth, and rich color.

Not sure you could get that kinda depth from a CMOS sensor.

 

[Jade Lane]

The latest DB footage looks nice (Portrait Day). Very different feel/texture, CCD depth, and rich color.

Not sure you could get that kinda depth from a CMOS sensor.

Just posted it up, interested to hear what people say.

 

[snowblindfilm]

So let me explain my disappointment is. I want DB to be a successful alternative, I really do. But I honestly think these guys are shooting themselves in a foot. For one it would be far better to release 20 seconds, with a note- "see we are working on it, and this is how it looks. If they decide to hand the camera to someone, please hand it to somebody, who will take is serious (remember Mobius release for C300?). I don't care what credentials guy shooting Venice Beach had- he made the whole venture like a bunch of amateur wankers. Is this image DB wants to portrait? Yes, like every camera it will have the set of it's own issues. But for crying out loudDB- please put your best foot forward.

 

[Razz16mm]

So let me explain my disappointment is. I want DB to be a successful alternative, I really do. But I honestly think these guys are shooting themselves in a foot. For one it would be far better to release 20 seconds, with a note- "see we are working on it, and this is how it looks. If they decide to hand the camera to someone, please hand it to somebody, who will take is serious (remember Mobius release for C300?). I don't care what credentials guy shooting Venice Beach had- he made the whole venture like a bunch of amateur wankers. Is this image DB wants to portrait? Yes, like every camera it will have the set of it's own issues. But for crying out loudDB- please put your best foot forward.

Check out the new portraits test footage. This is with the newly color and ISO calibrated firmware. Ungraded.

http://www.dvxuser.com/V6/showthread.php?314196-Digital-Bolex-Test-Footage-Portraits

 

[David W. Jones]

So let me explain my disappointment is. I want DB to be a successful alternative, I really do. But I honestly think these guys are shooting themselves in a foot. For one it would be far better to release 20 seconds, with a note- "see we are working on it, and this is how it looks. If they decide to hand the camera to someone, please hand it to somebody, who will take is serious (remember Mobius release for C300?). I don't care what credentials guy shooting Venice Beach had- he made the whole venture like a bunch of amateur wankers. Is this image DB wants to portrait? Yes, like every camera it will have the set of it's own issues. But for crying out loudDB- please put your best foot forward.

I wish no bad will toward these guys, but I have to agree 100%

And the new portraits video... WTF?

Are you guys marketing to 16 year olds?

It's a darn shame.

 

[Samuel H]

It's a fun video, and the music was great, so I liked watching it. It didn't blow me over with gorgeous images, though. In particular, I'm not sure I liked those skin tones, but it's all in the grading with these cameras anyway.

 

[Razz16mm]

It's a fun video, and the music was great, so I liked watching it. It didn't blow me over with gorgeous images, though. In particular, I'm not sure I liked those skin tones, but it's all in the grading with these cameras anyway.

If you want to get a feel for what this camera is about, download the Portrait DNG's and play with grading them.

What a refreshing change to work with a sensor that can produce such rich saturated colors and fine discrimination of hue differences. Instead of struggling to get nice well modulated saturated color out of pale flat grey images, you actually have to dial this one back a bit. This is the Kodachrome stock of digital cameras. It even picked up the color cast differences between the HMI primary key the Kino flow fill, and the Divalight used for the back light, all nominally daylight balanced. The D16 is a lush canvas for painting with light.

This camera is really really good at seeing color. In fact, it’s so good that it amplifies some of the color variance in daylight balanced lights that look similar to our eyes. At the beginning of the test, we set up three lights: a Fotodiox LED ring light (front), a 2 foot 4 bank Kino flow (backlight), and the hair light was a Kino Diva 400. When looking at the subjects we noted that the LED was slightly greenish, but when we looked at the footage in the computer it was even more evident.
The Diva also had a strong magenta hue when compared with the other two lights. The 2 foot 4 bank was actually the closest to 5600K, but it was a little blue. We changed the setup halfway through the day, this time with an ARRI 1200w HMI, a bounce board, and keeping the 2 foot 4 bank Kino as the hair light. We found the HMI to be the most accurate light, so we’ve now bought a few more HMIs to make sure we have an evenly balanced setup for future shoots. Joe and Elle

 

[philiplipetz]

It's a fun video, and the music was great, so I liked watching it. It didn't blow me over with gorgeous images, though. In particular, I'm not sure I liked those skin tones, but it's all in the grading with these cameras anyway.

these shots were ungraded except for slight contrast and saturation tweaks. There are DNGs for download.

Here is a user color corrected frame, not bad.


https://dl.dropboxusercontent.com/u/13696496/BLX-0079-20130926-0086402.jpg


color me impressed. However, it picks up subtle differences in color temperature of sources. It appears that the KinoFlo bulbs on their set were dimmed, which produces a shift to the magenta on our team's Kino Flos so strong we never dim, but diffuse.

 

[Kholi]

Sorry, not to rain on the parade, but if you shoot enough with Divas and Kinos you know that they have massive color shifts. RED ONE picked them up, DSLRs will pick it up, that's just how the units are if they aren't maintained thoroughly. Kinos may skew green, Divas love to go magenta.

Heck HVX and Letus Adapter picks it up.

I think there's a lot of merit in the deeper color here, the color shift thing is kind of a well known... at least if you're working with this equipment often.

 

[combatentropy]

I don't care what credentials guy shooting Venice Beach had- he made the whole venture like a bunch of amateur wankers

The Venice Beach footage is what got me excited about the Digital Bolex again. Its colors and tones are film-like (which reminds me, when comparing it to the other, videoy Digital Bolex samples, how much it depends on the grade) and its handheld camerawork and whip pans remind me of its global shutter.