new super 16 camera

[Robin Phillips]

someone took a swing at it. should be interesting https://www.cineluxcinematools.com/sixteen

[Brandon Paterno]

this would be awesome without a digital sensor. the digital sensor takes the place of an optical viewfinder.

the digital stuff seems to be the main focus and the magazine looks a bit like an unsightly growth. like holy shit! that gate is so deep inside the camera body. bet that’ll be fun 

wonder what sensor they’re using/who they’re contracting. i love that they’re throwing in some little film gimmicks like hand crank emulation, even if 99% of everything shot with it will be going straight to a hard drive. at rental house prices how will they get people to trust its reliability? 

[Tyler Purcell]

It's interesting, I always wanted to develop something similar, but when doing the math it makes no sense.

To make something truly work, the quantities sold and price of them would need to be high to offset the development cost. I'm not sure if their idea is to only place at rental houses, but even that is technically challenging. After talking to several companies about the concept, it seemed not worth the effort. The idea that the high end market (which this is targeting) will suddenly rely on a one-off custom camera, has been proven to not be the case in modern times.

My concern has always been that making everything from scratch, leads to the same problem we have now with parts and support down the road. It's almost better to base the camera on something with large quantities of parts still out there, where you're using a reliable pre-existing design with some alterations which don't affect the original design much. 

When reading the specs, there are a lot of things which are hard to agree with such as a 0 - 360 degree shutter angle? Maybe when not shooting film? The prototype is also MOS, which makes me feel the movement may not be where they want it today. 

I'm interested in seeing it at some point since they are local. 

 

Edited June 5 by Tyler Purcell

[Geffen Avraham]

I feel like not using the Eclair ACL flange distance system is a tremendous mistake in today's era of mirrorless cameras! We live the post-mirror age - most cool new lenses coming out nowadays are for mirrorless and offer incredible capabilities.

You want a t1 cine lens set? You want a small anamorphic autofocus? You want an f0.8 prime? All of these exist now - for mirrorless. An interchangeable active PL/EF/E/L/MFT mount would have been table stakes imo.

 

Edited June 5 by Geffen Avraham

[Brandon Paterno]

devils advocate but I wouldn't want to know what an f0.8 prime wide open looks like with the crop factor of 16... 

[Simon Wyss]

Contradictory is a video chip behind a shutter, so you won’t have 24p over the whole time.

[Robin Phillips]

20 hours ago, Geffen Avraham said:

I feel like not using the Eclair ACL flange distance system is a tremendous mistake in today's era of mirrorless cameras! We live the post-mirror age - most cool new lenses coming out nowadays are for mirrorless and offer incredible capabilities.

You want a t1 cine lens set? You want a small anamorphic autofocus? You want an f0.8 prime? All of these exist now - for mirrorless. An interchangeable active PL/EF/E/L/MFT mount would have been table stakes imo.

 

the digital sensor likely is receiving reflected light from the shutter, so it may just not be doable. at least not without shoving a bolex style beam splitter in front of the gate

[Ross Yeandle]

Hey I just love the fact that someone is actually making a film camera. It's certainly on in the eye for ARRI. 

And let's not forget RED and many new lens manufacturers are all recent additions to the movie industry. Anyway these guys are probably just testing the market response, the mk 2 will probably be sound synced. 😉

[Brandon Paterno]

it should be sound synced first go. it’s not easy but there are a couple of good options. my beaulieu r16 firmware/hardware has sound sync  with two decimal points of precision at the shutter using simpleFOC and a dc motor loop.

and this camera should have sound sync — not necessarily for sound recording, but for the purpose of digital video and film sync. cause that camera does not look to be very quiet! 

supposedly word on the street is this is a 35 grand camera. as a crash cam… put a pellicle in a filmo instead. 

Edited June 7 by Brandon Paterno

[Robin Phillips]

it looks like it will. but if people conform the audio off the video tap to the film scans, they'll all slowly go mad from the half frame off sync you'll get.

but I do think people might underestimate the demand for something like this at the medium to high end. I suspect anyone 30 or under would not be willing to use anything but an arri grey camera for a real production. and even in the crash cam scenario, having an integrated solution transmitting the image stream out could be more ideal. 

[Brandon Paterno]

re. sound sync — some claim that the kodak s8 cam lacks sync, fluctuating between +- 1fps. if a modern mopic camera that is seemingly aiming for a similar sort of use case (price notwithstanding) lacks this i have my doubts about this one implementing it. but that might be kodak cutting corners, or the people claiming these numbers just making them up. 

and for the crash cam scenario — i suppose. what will the after sales support look like? — it will be interesting to see how they handle maintenance and parts distribution. 

and the bigger question looms about what exactly the movement is doing that makes all the risks worth their tradeoffs! it seems that for the past one or two years they’ve been exhibiting this thing at at least one trade show so has no one asked?

[Robin Phillips]

the real question is did anyone get to check it out at cinegear?

at this pricing crystal sync shouldnt be an issue. Aapo has been doing miracle work in this area, especially given his pricing. it can obviously be done, and no one will touch it if it didnt hold sync. my point was just that the image on the negative and the image captured on the tap will be slightly off sync from eachother

[Dom Jaeger]

The technical challenge of sync sound is not crystal speed regulation - which is very straightforward - but making the camera quiet. 

[Brandon Paterno]

would the cost cover the R+D? aapos sync design uses discrete components — and he sets his pricing as his own in house R+D and manufacturing source that’s been operating for a while (years? more?).

as cinelux is a fairly new company, seemingly crossing over from a different part of the industry, i’d assume they’d have to come up with a whole new backend (both mechanical and electronic) that hasn’t necessarily been done before, with the components and manufacturing scale available today. it’s a pretty steep hill between the half frame gap, the structural integrity, and the mystery 3k (!!) digital sensor 

[Gautam Valluri]

21 hours ago, Robin Phillips said:

he real question is did anyone get to check it out at cinegear?

Here's an interview with one of the founders of cinelux, I didn't quite understand what he said about the double capture. Is he saying that he replaced the groundglass with a digital sensor? So the film part of the camera is filming it off of a screen of some sort? Does that mean you could slap a LUT onto it and then film it on 16?

 

[Robin Phillips]

interesting. at 1min 24 sec you can see there is a sensor reflected in the shutter. so no ground glass

[Jon O'Brien]

So it looks like a conventional film camera, with a spinning mirror deflecting light onto a sensor, where the groundglass would normally be. Surely a sync sound timing difference between the film and digital capture would be an easy fix in post. It's a clever idea. Regarding making a camera quiet, I've loaded and run a Panavision Millennium XL-2 and it was so quiet while running at 24 fps that you had to lean your ear right into the body, just inches away, to hear a thing. I was amazed at how whisper quiet it is.

[Aapo Lettinen]

21 hours ago, Brandon Paterno said:

would the cost cover the R+D? aapos sync design uses discrete components — and he sets his pricing as his own in house R+D and manufacturing source that’s been operating for a while (years? more?).

as cinelux is a fairly new company, seemingly crossing over from a different part of the industry, i’d assume they’d have to come up with a whole new backend (both mechanical and electronic) that hasn’t necessarily been done before, with the components and manufacturing scale available today. it’s a pretty steep hill between the half frame gap, the structural integrity, and the mystery 3k (!!) digital sensor 

All my crystal sync designs have software based speed stabilization, I never got discrete component analog crystal sync working well enough. And making software based systems is much cheaper after getting them working well.

The main reason I can drive the prices relatively low for a hand assembled device is that I am trying to keep parts count to a minimum and recycling software components to next designs to save developing time (you may have noticed I often have the next product partially based on the previous one). And I have this weird approach of separating the user interface and speed reference to one microcontroller  and crystal sync motor control to another so that I can treat them as "separate building blocks" which saves lots of developing time even if making the device a little more complex.

This allows running everything on tiny affordable low pin count 8bit microcontrollers which are easy to manually solder and reflow at home if needed. People who want to do everything on one microcontroller need a really powerful one and tons more developing time to work out the bugs as the software becomes too complex and software components interfering each other and competing for resources.

Making stuff cheaper necessitates making it in large quantities and minimum parts count to save work costs on assembly and direct the money to developing and testing instead. For example my 4speed/6speed board is already cheap-ish but has the leds and switch and backside manually soldered which raises price. Most small parts on frontside I reflow solder at home to save work.  When I designed my new 5-speed board I lowered parts count and changed voltage regulator and led type to be able to reflow solder everything on the frontside and price per board dropped to half compared to 6speed board. The 5speed has very different software and some drawbacks but it is about as affordable as manually assembled board can be if not sold at loss. These low cost systems use separate Chinese bldc board to run the motor though, if that needs to be made from scratch it is much more expensive. So they have the custom crystal sync and user interface board I made and that runs the factory made motor control board using 0-5V control line or other standard means. 

If the motor is brushed like cp16r then I generate the pwm signal controlling it directly and everything is custom made in the system.

Digital sensor image processing is much harder than making crystal sync system for motor control though crystal sync stuff is already hard. Making a good video processing from bare image sensor to video feed and recording internally at the same time can be dozens of times harder and need very advanced fpga programming, you may need to order it from somewhere as custom work because learning that stuff is simply too difficult even for most programming people and designing the boards is hard too because of the high frequencies, dozen layers or more, it needs to be factory assembled for impossible to solder manually, etc.

So on this type of hybrid camera it can be 70% of the work on the digital image processing, 25% on mechanics and 5% on crystal sync approximately if doing all parts well 

[Jon O'Brien]

It would be relatively easy I'd think these days to make a blimp/barney for this camera. It's quite compact.

[Brandon Paterno]

watched the interview, the film side just doesn’t seem to exist yet. they’re aiming for a launch late next year. ergonomically the thing looks questionable, but with proper sound deadening/mechanical isolation i’d assume the whole things gonna balloon up a bit, maybe for the better. 

it seems like they’re selling to those that won’t stand for an image pickup traveling through ground glass optics? the sensor is 3k because it is quite literally cropped only to pick up a super 16 area over a 4:3 sensor. it will be interesting to see if they leave the option of opening the digital gate to the entire sensor width.

regardless, hopefully they continue their campaign and continue to build traction/buzz. 

Edited June 9 by Brandon Paterno

[Aapo Lettinen]

I assume the main idea of this new camera is to provide good enough digital backup version of the image so that producers are comfortable using film origination despite the added financial risks associated with film nowadays. Much easier to sell the 16mm film idea to them when can convince them that reshoots are not needed if the negative has issue, they can just use the digital backup to finish the movie. This is pretty signifiant advantage and allows shooting film more regularly. People hate risks if they can create extra costs and taking the perceived risk away may double or triple film originated shooting as it is almost always about the risk management and logistics, not about costs

[Simon Wyss]

Kodak has the same arrangement with its latest Super-8 camera, a mirror shutter reflex system in combination with a digital sensor.

[Tyler Purcell]

17 hours ago, Simon Wyss said:

Kodak has the same arrangement with its latest Super-8 camera, a mirror shutter reflex system in combination with a digital sensor.

The Kodak camera works like a standard motion picture camera, with the video assist shooting the ground glass. 

The new camera does not do this at all, it's entirely different. 

The video camera is at the same focal plane as the 16mm imager, no ground glass, no optics outside of the mirror shutter. 

Thus, the film gets 100% the same image as the digital imager and visa versa. 

This has never been done before on ANY motion picture camera. 

[Brandon Paterno]

here's a photo (not mine) of the super 8 camera front.

[Steve Switaj]

16 hours ago, Tyler Purcell said:

The video camera is at the same focal plane as the 16mm imager, no ground glass, no optics outside of the mirror shutter. 

Thus, the film gets 100% the same image as the digital imager and visa versa. 

I wonder why they went with a spinning mirror?

Yes, it's traditional , but it's going to make things more complicated for a camera that wants to be equally good at delivering a digital or film image.

First, it's going to affect your exposure timing, you're going to get 180 degrees max minus whatever the shutter "clearing time" is over the gate. It looks like a vertically running shutter and a 16mm frame is only about 7.6mm tall, so that may not be a big issue, but it will prevent you from going over, say, 170 degrees while running film.

Second, you're trying to make a high-quality digital image off the moving mirror parh, which, in my experience, is usually not the best optical surface in the world.

I feel like every film camera I've ever handled eventually develops some sort of damage to the mirror. It's usually slight, but it's there.

Dust gets into the port when you change the lenses, bits of stuff get stirred up by the air currents in the box because of the spinning mirror, some of this dirt clings tenaciously to the mirror, maybe because of static. Sometimes the AC's get a little to enthusiastic about cleaning in there. The front-surface mirror is insanely easy to damage and almost impossible to clean well without something exotic like CO2 dusting. 

You get the idea. Even in normal service with conscientious operators the mirror eventually shows some defects, and you can  often see it through the finder as you inch the camera slowly. 

This does not matter in normal operation, since the finder image is just for the operators, but if you're trying to make a for-real picture with it... it's probably going to show.

Tyler, I suspect this might be the right place for your pellicle camera to shine. 

Beamsplitters have their issues, and we've talked about them here, but in this application the trade off might be worth it.

A beamsplitter would have the same problems with dust and dirt and scratches as a mirror, but a well-designed system could make a pellicle swap-out for a fresh one quick and straightforward, something you could do at the camera truck. The Fries R35 conversions did this successfully back in the 90's.

A beamsplitter would also decouple the timing requirements between the film and video sides, not to mention how much more straightforward the mechanical design would get.