F35: F35 vs Digital Bolex

Hey thanks for posting your impressions of the D16 Keith! I've been looking into purchasing one to use as a walk around. Like many of you guys here I too was brought to the F35 by my love affair with the A-CAM DII....I actually tracked one down in the states that has never been used, brand new in the box, But I find it very difficult to pull the trigger because the owner wants almost what it cost when it was first released (about $12k)....If I had my choice I probably would've chose the A-Cam DII over the F35 ..but now I'm here and I also couldn't see myself without the F35 either. Which brings me to the more affordable A-Cam ..the D16. Keith, please do share your first test whenever you have time buddy. Email me grabs of your setup when you time. Super excited for you pal.
 
Great info here on this thread! I've never shot with either of these cameras, but would love to do so. As a Bolex owner, I have just purchased 500ft of film for it. I got a client who wants to have the Bolex as an extra camera at their wedding. It will be my first time cutting film with digitally sourced video. I know it will have its challenges but I can't wait!!

in talking about motion cadence relative to film, I'm wondering if anyone has tried to insert black frames into their timeline to mimic the way that there is black space between frames when we see film projected. Just how the shutter mechanisms worked. For example, in a film theater, we sit in the dark for almost a third of the time. I was always taught in film classes that this was integral to differentiating between film and video. The theory about this was that because the image was not visible the whole time, that our brains would interpolate it, and thus seeming more natural/real. Or to say, that because we see the images the entire time in video, that the artifice is raised.

Has anyone also been thinking about this and tried it in a digital way?
 
Great info here on this thread! I've never shot with either of these cameras, but would love to do so. As a Bolex owner, I have just purchased 500ft of film for it. I got a client who wants to have the Bolex as an extra camera at their wedding. It will be my first time cutting film with digitally sourced video. I know it will have its challenges but I can't wait!!

in talking about motion cadence relative to film, I'm wondering if anyone has tried to insert black frames into their timeline to mimic the way that there is black space between frames when we see film projected. Just how the shutter mechanisms worked. For example, in a film theater, we sit in the dark for almost a third of the time. I was always taught in film classes that this was integral to differentiating between film and video. The theory about this was that because the image was not visible the whole time, that our brains would interpolate it, and thus seeming more natural/real. Or to say, that because we see the images the entire time in video, that the artifice is raised.

Has anyone also been thinking about this and tried it in a digital way?

interesting I am going to try this
 
interesting I am going to try this


Awe, snap!! I can't wait!

Im sure there is a lot of math to consider, from the frame rate to shutter speed to potentially the hertz settings of your monitor (amongst other things) but I'm super interested in your potential findings!
 
It isn't that easy. A projector normally has two "phantom" blades in it (from memory) - thus for each single 24p image it is refreshed three times. That fools our brain into thinking it sees more than it does. The dark portion is a fraction of a second - it is not a whole frame, really far from it. If you start including black frames in a video you're just going to make it look like it is strobing.
 
Interesting point, JM.

Well, I don't mean to hijack the thread, just wondering. I figured it wouldn't be like inserting whole frames, but thought that whatever rates of darkness that manifest in film could be duplicated digitally. It may not be possible, but my curiosity led me to ask :)
 
It could be duplicated digitally, but not easily. And in any case it is only done at the projector stage.
 
i am playing with a flicker plug in - luca light kit - that may do some simulation of this - no it's kind of cheesy - there is film grain with flicker in it - that's about all I got
 
Last edited:
Dunno if it's appropriate, but I figured I'd share some D16 grabs with various lenses and grading workflows. I'll present them in the order that I shot them. The usual caveats regarding the operator's skill and learning process apply....

This one was taken with a Sun Cosmicar 22-90mm TV zoom lens. A cheap $100 hunk of glass. Super heavy, very big and not wide enough for everyday use. I sold it the next day. Looking back on this I'm not a huge fan of this shot. But I also had no idea what I was doing with the camera at the time, so....
Impulz LUT and some other mish mash of stuff.

PugetSunset_CosmicarZoom.jpg

These next three were taken with a different cheap-o TV zoom lens, the Computar f1.2 12.5-75mm. Cost about $100. It's a bit of a Coke bottle, no doubt, but it worked until I could get my C-mount adapter for the Angenieux. Now that I have that I'll probably eBay this one since for grab & go shooting I'll use primes.

These are terribly under exposed and they're all pulled up a stop or so in post. One thing I forgot to mention in my other post was the difficulty in finally finding a decent way to measure exposure in camera. The HMDI out image luma mimics the brightness settings on the internal screen on the D16. The user can set the brightness of that screen to LOW, MID or HIGH for easier viewing in different light. That's how I botched the exposure on this shoot- I had the screen brightness on HIGH, so my HDMI feed was giving me feedback that the image was too bright, but the actual true exposure on the sensor was fine. So I closed down the iris to get the false colors on the DP4 in line and thus I pushed the exposure way down in camera. Thankfully I figured out that setting the internal screen brightness to LOW gives the most accurate sense of the sensor's true exposure using an HDMI monitor's exposure tools (which the D16 doesn't have yet- no zebras, false color, histogram, etc.) It does have a RAW view mode which shows you the actual luminance of each photosite on the sensor in B&W- great for seeing when highlights clip, terrible for anything else since it's all shoved down into the darks (inverse square, etc.)

Grading is an Impulz LUT, but I forget which.

nature_1_computarZoom.jpg

nature_2_computarZoom.jpg

nature_3_computarZoom.jpg


These next two were Nikon AIS 35mm and 50mm primes as taking lenses with a Sankor 16c anamorphic in front. Grading is a mix of Impulz LUTS and some Filmconvert.

mountains_1_Nikon35Sankor16c.jpg

mountains_3_50mmNikonSankor16c.jpg

This next one is a Nikkor 75-150mm AIS push/pull zoom, probably full 150mm here. A secondary on the sky to pull it back a bit, Impulz/Filmconvert combo.

mountains_2_Nikon85.jpg


And finally the Angenieux f1.9 9.5-57mm zoom. No LUTS here, I just processed the RAW images with BMD Film color space, BMD Film 4k gamma, and a simple luma curve to push contrast and dialed in the saturation to taste. These best show the camera's noise natively. The interior shots were night time using only a practical lamp with CFL bulbs and the lens was WFO- hence the softness.

cat_Angenieux.jpg

JM_Angenieux.jpg

Penny_Angenieux.jpg

Plant_Angenieux.jpg

SkyTrees_Angenieux.jpg

Like I said, there's something organic going on here with the D16 that I like. I chalk that up to the Super 16 CCD sensor and their very good color science. But you'd NEVER mistake the images out of a D16 for the F35. It's just got its own flavor of nice going on.
 
Last edited:
I think those look gorgeous - there is a richness to the color and the softness and everything is really lovely. I do think there is something in this that is so completely organic and great - okay I am curious again about a side by side shoot out

but then you get videos like this
And one is just like - okay well it's the grade that makes it good, sensor be damned it's shots, lighting, and color just as importantly and then I need to go back and do my own work instead of drooling over the camera I don't have.

I mean I've shot on Alexa, Film, Red Dragon - and have shot stuff that looks horrible - so I need to settle down and work on myself.
I am putting my tongue back in my mouth.
 
Last edited:
Seriously? That donut thing looks awful much of the time. Take out all the (somewhat subjective) color arguments and try getting over the micro-jitters. It looks like someone is hand-holding a GH3. Amateur work at best and there's no excuse for that kind of thing other than simple, newbie laziness and ignorance. I'll stick to real cameras and craft.
 
I don't have a DBX but i have spoken to the engineers and i used to mess around with similar Kodak sensors.

The CCD sensor is fundamentally different to CMOS, colour rendition is much better but it's more expensive to make, lower resolutions, doesn't handle flare as well as CMOS and it's tricky to integrate into a camera so that's why CMOS took off. One of the biggest differences is the colour red, the CCDs are just so much better at it. (of course CMOS is so well developed that it's not quite the same gap as it used to be and some CMOS are probably better now).

The DNGs that the DBX produce are simple and contain all the data but they are 12 bit linear, so you'll never have more than 12 stops in there. The magenta cast sounds like some firmware thing and that will pass i'm sure.

Some other cameras actually don't contain the full information in the DNGs, it appears that BMC pad the lowest bit, so 12 bits is actually 11. There are some that are even worse. This means that banding can appear and it's all a bit of a spin really. DBX doesn't.

For what it is, it's a lovely camera.

And i wonder if the motion cadence that Kholi is famous for :) is to do with rolling shutter? I have to say after messing with an A7s i find myself disturbed by the cadence of that whereas i'm a lot less sensitive to others...

cheers
Paul
 
Thanks, Ed. Nice of you to say. I agree with you- one can get so lost chasing the thing they don't have that they fall prey to not working on their own craft with what they do have. But at the same time I think it's kinda normal for artists to be a little bit peculiar about the tools of their craft and to keep striving until they find the one thing that fits them perfectly. Personally the Internet Camera Fan-Bro Spec War doesn't drive that peculiar pursuit in my life, so I'm happy with whatever quirks come with getting something that fits what I like to do and what I like to make.

And yeah, Vince, that donut thing has all the earmarks of a person not knowing what the footage needs regarding grading. It is soaked through with that cyan and magenta poison. The micro jitter, that is on the shooter not the camera. Sadly micro-jitter is another plague on a lot of D16 footage online.

If you want to see something tasty in a more modern & clean style, this was shot on a D16.

Frederik consistently gets excellent results from the camera.

This is a bit more stylized and quirky- a VERY D16 kind of production, but it's got quality.

I'll be shooting a spec promo piece here in the next week or two and I'll show some stuff from that when I get the chance. That is if people are still interested. I don't want to overrun the F35 board with too much D16 stuff.
 
I personally think we should talk D16 bolex in here - this thread could become a CCD board.

I think CMOS was all about cost savings. The Sony F35 I think was around 200k when it came out then the Arri Alexa came to the scene for around 60k and so Sony needed to compete so the F65 came out I guess. Also CMOS is cleaner and can get to high ISO and it seems to be able to get to higher frame rates - but I think the Sony F35 sensor could go up to 120 FPS too if they wanted to at the time.
In terms of sharpness, I still think the Sony F35 has a ton of photosites and resolving power and if someone wanted to, they could easily make a 4k CCD camera but most people don't really see the value in CCD I think because of the noise of the sensor.

However using programs like Neatvideo, I am able to make the Sony F35 footage as clean as alexa footage - so I'm sure someone could implant this into the camera itself.

Anyway I'm mostly just throwing words and ideas out, not science - but I do wish someone would come in and recover the Sony F35 sensor and all CCD sensors and make a modern camera. Who knows though what it takes to make a camera - as seen by how long the DB16 took to make - it seems completely daunting.

Who knows what it costs to make a sensor and why people make them? I guess for the medicial industry and military and photographic industry? And what people want out of a sensor? And in the world of still photography, they say sales of DSLR are way down - it's all about cell phone technology - small small sensors.

And those pieces you posted Keith - unexpected farewell - nothing in that I think that felt too organic - the other one is quick lovely - but yea it's the artist not the tools really.

Anyway now I want to interview a sensor manufacturer.
 
I'm not a big fan of "multi-quoting" Ed. But you touched on a few different points that are better addressed this way.


I think CMOS was all about cost savings. The Sony F35 I think was around 200k when it came out then the Arri Alexa came to the scene for around 60k and so Sony needed to compete so the F65 came out I guess.

Sony has pissed itself twice. First by contradicting everything they said together with Pananvision and killing all the F35 promotional material when they released the F65. Then doing the same thing when they released the F55 which goes against what they preached with the F65 and putting a lid on the F65. And they wonder why ARRI has a better rep.


Also CMOS is cleaner and can get to high ISO and it seems to be able to get to higher frame rates - but I think the Sony F35 sensor could go up to 120 FPS too if they wanted to at the time.
In terms of sharpness, I still think the Sony F35 has a ton of photosites and resolving power and if someone wanted to, they could easily make a 4k CCD camera but most people don't really see the value in CCD I think because of the noise of the sensor.

In a "perfect world", movie cameras would have ISO 800 CCDs and video cameras would have CMOS with these ridiculous night vision level ISOs of late. Although for production I'm fine with ISO 500 for movie/production cameras.


However using programs like Neatvideo, I am able to make the Sony F35 footage as clean as alexa footage
If you believe the rumors the Alexa has "Neativideo" built-in. So it's fair game.
 
I'm not really into multiquoting either.
Alexa has a much cleaner sensor than the F35 - the DP on that wicked witch movie with Angelina Jolie who usually shots F35 shot it on the Alexa probably because it was cleaer.
Using an algorhythm like neatvideo can clean up the F35's CCD noise and I have done that on a bunch of projects myself.

There is no perfect world of ASA - until the 90's it was max of 200 ASA - now we are living in a 800 ASA world - things have changed a lot - but have movies looked better because of it? Me personally I don't think so. I still think Fargo and No Country for Old Men look better than anything Deakins has done on the Alexa.

But hey, it's more than that - it's budgets - it's production schedules - and as I find when I used to shoot at 50 ASA on the Letus 35mm DOF adapters - I brought lights in because I had to - now I am usually lazier - and don't bring in the lights as much - and sometimes I wish I did pretend I was still shooting that way.

Sony didn't screw themselves by going with the Sony F65 over the F35 - they hurt themselves by not making the F65 easier to work with in post - a la the Alexa that could crank out a Pro Res copy right there on the camera - this was their fault.
And the F55 wasn't probably the smartest move either - but I'm not sony nor do I know much about how they run as a company.
I also think Arri should have made a newer sensor by now over the Alexa sensor - but that's just me - a guy who is using a camera from 2008.
And finally Red screwed themselves by making the smaller epic body - the red one mx worked fine - no fan noises, more connections - it was just smarter.
Now they are really paying for it with the Red Weapon which is basically admitting they really screwed up with the design of the Epic body.

Anyway, I don't think we should criticize company's and their cameras - we should learn how to work with what we have and make it better.
If you don't like something, don't use it.
 
The problem with a lot of DB footage is that there is a ton of color information in the native footage and there didn't seem to be an established workflow when it was released (that I could see) that properly explained how to fit those colors into a normal gamut. Here's a bit more info: http://www.digitalbolex.com/normalizing-d16-color-space/.

When done more or less correctly, the colors coming out of camera are really beautiful:

https://vimeo.com/94800395

https://vimeo.com/74887302

thanks for this - this is useful for anyone using raw with any camera - now I am begining to see how there are steps to ensure that the RAW output of the DB16 is read properily. The camera definitely can create a really beautiful image that has a pop to it that looks better than a lot of stuff I have seen shot on the Alexa.
 
Sony went with CMOS because the technology matured, you need a whole production facility for a CCD (CMOS is easily made on a standard production line), and a CCD is nowhere near as flexible as a CMOS. The F35/Genesis were never going to do 100fps - they didn't even do 50 when the Genesis came out. Also it became unacceptable for a camera to cost $200,000 post-RED.

It is easy to pick faults with the F35,F65,Alexa or the RED but the truth is we have never had it better! Ten years ago no-one would have believed these kinds of cameras would be available at this level.
 
Back
Top