Dom Jaeger

I'm much more familiar with Arris, so probably biased, but I think you'd be better off with an Arri S16 SR2 or SR3 (or as others have recommended an XTR). A-Minimas are lovely cameras, but generally not ideal as an A-cam. The archives here have some good info on this, see for example: http://www.cinematography.com/index.php?showtopic=18366 These days even SR3 kits can be had for within your budget. Another option is to buy the glass and rent the camera, which might be a better long term investment. With S16 sized digital sensor cameras becoming available good S16 lenses should hold (or even increase) in value more than film cameras will. Or as Adrian mentioned (depending on the duration of your shoots and how organised you are) it might be worth looking into renting a full production kit including lenses, grip and accessories - the advantage being that you won't need to pay more to have your gear checked and potentially serviced after buying it, and if there are any issues the rental house can simply supply a replacement. When you calculate your budget, it's worth noting that the cost of buying high-end film gear doesn't end with the purchase, you also need to factor in the added expenses of testing, possible repairs and ongoing maintenance (as well as schedule disruptions when something stops working properly). It's the reason most serious productions rent their gear rather than buying. You can negotiate substantial discounts for long term hire, particularly with film-based equipment these days. Try Cameraquip (where I work), Lemac or Panavision. Or (as a final option) you could find a DOP who already owns a S16 camera and hire them and their camera to shoot your project.

I transferred some video footage to Double 8 Ektachrome recently by simply shooting it off an iMac screen. (Thanks to Richard Tuohy of Nanolab for the tip.) Didn't get any flicker, but if the footage is high contrast or changes from say bright exteriors to dim interiors it can be very hard to get correct exposure. Turning down the screen brightness or contrast helps, but at normal speed you then need a very fast lens. I used an f0.9 Kern Switar and it was still a touch underexposed at times. Since the original video was only under a minute I filmed it about 4 times at different exposures and selected the best exposure scene by scene, then spliced it together.

Corrosion and film cameras are a bad combination. Unless it's dirt cheap and you really want the challenge of cleaning it up (with the likelihood of corrosion affecting electronic reliability, fungus in the optics etc) I'd avoid this one. I've seen pretty clean 16BL kits go for as little as $500 lately, far more serviceable SRs for $1000, or even S16 PL mount SR2s and 3s for $2000 - 2500. It's a buyer's market, you don't have to settle for rusty, untested cameras. I'm pretty sure the only place in Australia anymore that is willing and able to service older Arriflex models is the Melbourne branch of Cameraquip (meaning me). Deakin Uni's film dept used to have a few 16 BLs which they occasionally brought in for repair, but they may have sold them by now. Parts for these cameras are scarce, and the blimp can make working on the lenses a pain (especially if they're corroded in). It's the Arriflex model I'd least recommend. You're in Melbourne right? If you're interested in 16mm cameras, you're welcome to come into Cameraquip and get familiar with some working models we have (Sts, BLs, SRs, Bolexes etc). I'd be happy to show you around.

The focal lengths don't change by using the lens on a different format, it's still 20-100 no matter what camera you mount it on. Som Berthiot made 17-85 zooms for 16mm, perhaps this lens is a variation on that with a focal length extender added to increase the image circle, much like (for example) Zeiss 10-100 16mm zooms were converted to 11-110 for Super 16. It may have originally been a conversion for use with 1" video tube cameras which require a larger image circle than Super 16. At any rate it's now 20-100. A grub screw in the focus ring is the most likely way the focus was locked off, look for one that is in a different position than the screws holding the focus ring on to the inner barrel, which will be uniformly spaced around. Otherwise you may need to remove the focus ring (by undoing those uniformly spaced grub screws) and look underneath to see if it's been locked off internally. If you do that, mark the position of the outer focus ring relative to the inner threaded barrel it attaches to so you don't lose the distance settings. You might need to tweak the position later by checking infinity on a reflex camera at the 100mm end.

In terms of current developments this Arri prototype looks interesting: http://www.arri.de/about_arri/press/english/english_single/?tx_ttnews%5Btt_news%5D=1279&tx_ttnews%5BbackPid%5D=7581&cHash=f5fedf50585ddcd337cb839b93cdb94b

Simon, I found some time today to disassemble and measure the early Arriflex 35 we have in our museum. Serial number 700 makes it (I believe) the 200th Arriflex made. I found no imperial dimensions whatsoever (apart from the standard 3/16 mounting thread in the base). Many of the actual cast dimensions are neither precisely metric nor imperial, but then the form is often curved or sloping. Of the unmachined dimensions that were metric: width of door measured from raised light seal: 90.0mm bore diameter for viewfinder tube: 24.0mm outer diameter of turret casting: 112.0mm outer diameter of cast hood base: 120.0mm Some other dimensions: turret shaft dia: 8.0mm main drive shaft dia: 6.0mm claw eccentric cam shaft dia: 4.0mm mirror shutter bearing housing diameter: 20.0mm distance between rear side cover screw centres: 52.0mm flange depth: 52.0mm (obviously!) etc The movement patent by Howell you refer to resembles the later cardiod cam design, not implemented by Arri until the 50s. The first movement was a very simple eccentric with a cam underneath that shifted the long claw arm pivot back and forth. No dwelling at the bottom of the pulldown to increase steadiness, no rigid claw frame: A picture of the simple but effective gear chain, accessed by removing the side cover (held by only 4 screws and the inching knob): And the front turret cavity: I'm no expert in casting technology but the rough unmachined areas look more like they were sand cast than die cast. I have more pictures if you're interested in a specific area. I personally can't see any resemblance to a Bell & Howell 2709, or any other B & H camera.

Fair enough Keith, Arri were the dunderheads. It's more fun to blame Panavision though, especially if you've worked for them. ;) By the time I started in the service department at Sammies in '96 the Arri version seemed the more sensible and already accepted standard. You have to admit, it does keep it simple to make pin 1 negative across connectors.

I'm not really familiar with the high speed models, but we have a high speed medical 2C in our museum (designed to shoot x-ray film) with an 80fps tacho and the pressure plate is the same as yours. Only one other 2C we have has the same pressure plate (but the tacho is standard 50 fps max), all the other standard speed 2Bs and 2Cs we have don't have the 4 nuts on the back of the pressure plate. So I would assume that yours is a high speed plate. Aside from possibly giving the camera a fresh lube, I don't think you need to worry about anything else. The mags should be OK, 435 literature says you can use older magazines up to 130 fps, well past what your 2C can do. Test with some old stock, if there are issues it might be worth having the mag take-up and feed tensions checked. You can build a 32V battery by joining four 8V lead acid batteries in series. Something like this: http://www.allbatteries.co.uk/lead-acid-battery/8v/3-5ah.html joined negative to positive.

Yes, adapters are available for both TS and C-mount. The lens doesn't need to rotate in the mount like some Schneiders and Cookes, so a simple clamping device will work. Very nice lens for its age, the ones I've tested were sharp to the corners with good contrast, minimal breathing and low distortion. Covers about a 14mm image circle, not quite S16.

Panavision batteries are wired with the reverse polarity to Arri (and every other standard I've come across). What that means is that normally for example a 24V battery with a 3 pin XLR or Canon socket is wired with the negative terminal on pin 1 and the +24V positive terminal on pin 2, while Panavision batteries are wired in reverse. Don't ask me why they chose to do that. The upshot is, you need to only use Panavision power cables with their batteries, otherwise a reverse current is sent through the appliance. Some cameras, like Alexas, are very well protected against reverse polarity or short circuits in the input voltage, others aren't and can blow components. With the amount of voltage, some appliances are regulated to accept a broad range of voltages, Alexas for example will work on anything from 10 to 28V. Others may be 12 or 24V only (though generally there is a tolerance of several volts either side). A 12V battery may supply over 14V when fully charged and run down to 10V before the appliance switches off, and often these days the batteries are more likely to be 13.2V or 14.4V, simply because a nominally 12V appliance can usually handle up to around 17V. The danger is applying too much voltage rather than less, so avoid putting 24V into a 12V appliance. If in doubt, check the voltage specs which are usually written on the back.

Should be fine on anything but the Studio, where you need to be careful of the mirror shutter. Angenieux Optimo DPs for example won't fit that one, but will clear the other cameras mentioned. Here's a drawing of maximum allowable lens protrusions for the Alexa Studio:

Great work Lasse and your team, to have gotten so far. For what it's worth (2 cents?) my only suggestions would be to drop the pin registration and add an optical viewfinder. I don't know what sort of pin registration you're using (and never seen a high speed Mekel) but a sprocket-driven loop and pressure plate alone would be a vast improvement over traditional Super 8 steadiness, no need to introduce potential problems with film shrinkage or perf variation by using a registration pin. There's also the issue of added noise, unless you go to the extra expense of adding pitch control to the movement. But perhaps it's not a proper registration pin that fits the perf length, just a sort of stop pin to prevent the film advancing past where the claw positions it (which makes more sense for a high speed camera). Pulling focus off a video split monitor can sometimes be tricky even with state-of-the-art splits for pro film cameras. You need a very good camera in there to judge best focus, and a decent monitor. If you're outside in sunlight you need a deep shade to see the screen. It seems a bit crazy to have a reflex mirror without an optical finder. A simple rotating prism or mirror could switch between viewfinder and monitor out if you wanted to avoid the traditional split prism.

Interesting conspiracy theory Simon, but I'm not convinced by your evidence. Which parts of an Arriflex 35 are Imperial size? Have you dismantled and measured one from that era? There could be other reasons for Imperial sizes other than it was an American invention, imported tooling for example. Why could Arnold, Richter or Kästner not have had a thorough understanding of motion-picture camera design? Arri had been manufacturing cameras since 1924, and their unusual Kinarri design shows they were not afraid to experiment. A nod to the Akeley certainly, but different to other small handheld designs of the day, which generally followed the box-shaped pattern of Ernemann or DeVry. The Arriflex 35 is really quite a simple camera compared to a Debrie Parvo or Bell & Howell Standard, the breakthrough wasn't in the engineering but the concept. According to Raimondo-Souto's other excellent book "Motion Picture Photography" Arnold and Richter patented the mirror reflex idea in 1931, but a prototype wasn't made until 1936. Perhaps it took Kästner's expertise to turn idea into reality. It's true that despite Germany's enormously influential film industry of the 20s professional studio motion-picture camera manufacturing was curiously centred elsewhere. The UFA studio in Berlin used Pathes, Parvos and Eclairs. But Germany had a wealth of compact camera manufacturers as well as arguably the most advanced and innovative optical manufacturing in the world. Just prior to the Arriflex 35 the Berlin firm Askania produced a revolutionary hand-held camera named "Shulter" which for the first time conceived a body shape designed to rest on the shoulder - decades before Eclair or Aaton claimed the innovation. German cinematography had a history of leading the way with free camera movement, from Murnau's "unchained camera" moves in The Last Laugh to Riefenstahl's work in Triumph of the Will, so it seems a natural progression that a German company might invent the ultimate hand-held camera with a reflex viewfinder. I'm not sure what the history of die-casting has to do with it, surely you're not suggesting that a country which by the mid 30s had among the most advanced automotive and aeronautic industries in the world was incapable of casting a relatively simple camera body?

Hi Brian, if this is an important project and you've just picked up a film camera off ebay or something you need to have the camera, lenses and adapters checked over by a technician. Trying to adjust things by yourself is simply not how to do it if you care about the image quality, or want to feel confident about the gear holding up over the course of shooting your project. There are a few factors that can lead to soft images. As you've mentioned, stopping down increases your depth of field, but it also reduces aberrations in the lens itself, which leads to sharper looking pictures. Some lenses perform better than others wide open, but virtually all lenses will sharpen up stopped down a couple of stops. As well as this, with wider focal lengths (combined with larger apertures) the back-focus distance becomes more critical, a tiny error in this distance and everything will be soft. There are 3 factors at play in the focus adjustment of a reflex film camera: The lens itself should be calibrated for correct infinity back-focus, so that you can rely on the distance scale and infinity can be reached. With zooms it's even more important, so that focus is maintained through the zoom range. The camera flange depth needs to be set to factory specifications, matching the infinity back-focus of the lenses (or sometimes a very tiny bit shorter, to allow for the film to float as it runs through the gate, and the light to penetrate into the film emulsion layers). This is usually done with a depth gauge accurate to at least one hundredth of a mm (less than half a thousandth of an inch). It can be checked on an autocollimator with film running through. The ground glass needs to be set at a depth that exactly matches the infinity back-focus distance of the lenses (or the nominal camera flange depth), so that the viewfinder image matches what is recorded on the film. This distance is not physically measureable since it bounces off a mirror (or sometimes a prism), so it needs specialised equipment like an autocollimator which measures it optically to properly check and adjust the setting. You can try and do it by eye, but it's very difficult to get exactly right. Other aspects, like the viewfinder framelines matching the gate are determined by the groundglass/prism assembly being correctly positioned in the 2 axis perpendicular to the depth. I would strongly recommend that you send the camera and lenses to an experienced tech to get at least these basic settings checked. And since adapters can also introduce variations, they should be checked as well.

Like Simon said, the purely mechanical cameras will last longest. A Bell & Howell 2709, stored well, could conceivably still work in 500 years with a little restoration and lubrication. I've restored a Mitchell NC and a Debrie Parvo (both over 80 years old) to working order with very little needing to be done, these things were hand-built to incredible standards. Something like an Arri 35 2C would also last a very, very long time if properly stored and maintained. They're very simple and sturdy, if precise, machines. Run enough film through them and you'll start to get wear, but they'll probably keep working until the motor dies. I work for a rental house that has among its inventory 2Cs, 35-3s and BL4s - Arri cameras between 20 and 50 years old - that are all still perfectly functional (thanks to in-house maintenance). More modern 35mm cameras like Arricams, Moviecams and Milleniums are also very well made, but have more things that can fail, complex electronic controls and such. They need to be quiet as well as steady, which requires fairly regular maintenance if they're being used a lot. In time (or even now) some of the components won't be available any more. But you could probably store an Arricam for a couple of decades, pull it out and give it a light lubrication and it will work just fine. Toughest modern camera is probably an Arri 435, those things are virtually indestructible. Simpler amateur cameras from up to 90 years ago can still work OK today if they've been stored well, though of course they had less exacting performance requirements than professional studio cameras.

I recently supplied a 20 second loop of Standard 8 Ektachrome 100D for an art installation, ran for about 3 or 4 hours no problem on a Bolex 18-5 projector. I didn't notice any major fading, though I didn't study it closely afterwards. But even if it does fade, so what, that's film and this is art right? B) I seriously doubt the film will melt, even for short loops, unless the projector is faulty. Older projectors that used very hot bulbs (like a Bolex M8) might be a different story, but a Super 8 Eumig should be fine I reckon. The bulb might blow though, so I'd keep some replacements handy. The simplest method for running loops I've seen is 2 rubber bands on the take-up spool (at the film edges) and smooth metal rings a bit larger than the film width suspended from the ceiling for the film to loop through. Just avoid too tight a bend through the rings. Works a treat.

I've heard too many people's stories of K3 shoots gone wrong to recommend them. The best thing about them is they're cheap, usually not much more than $100 or $150, so this one is a bit pricey I reckon, even for a S16 conversion. Regarding the conversion - the viewfinder doesn't cover a S16 frame completely, the wide end of the zoom won't cover it either, and even a machined out gate like this one has isn't really ideal for S16, since the left film support rail leads into and out of the expanded picture area (possible scratching) and where it has been machined out there is nothing supporting the film on that side of the gate opening (possible focus softening on one side). But maybe I'm being picky.

It tends to depend on the type of show and how much time production has allowed for gear check and testing. Feature films will spend quite a bit of time going through the gear and building up kits, shooting tests and basically checking everything thoroughly. Usually done by ACs, DITs and Steadicam ops, DPs might come in to shoot specific tests or look at lens projections or familiarise themselves with a new camera system. Some DPs are more involved than others. Shorter rentals tend to spend less time on testing, sometimes these days an AC will barely have time to do a quick gear check.

Pink range is one stop over medium grey, a traditional rule of thumb for correctly exposing Caucasion skin (which can of course vary). Bear in mind that the false colour guide is based on the colour processing set for the particular output signal, so if the MON OUT for example is set to Rec 709, the false colour on that device will be based on a REC 709 image, rather than Log C. Not a big difference for the pink range, but definitely affects the clipping extremes. Lots of this kind of information available on Arri's website: http://www.arri.com/camera/digital_cameras/learn/alexa_faq.html or download a pocket guide user manual for the latest software version: http://www.arri.com/download_search/download_search.html?suchoption1_id=Camera&suchoption2_id=35_Format_Digital_Camera&suchoption3_id=ALEXA

Interesting, according to that site K3s continued to be made up until 1993. Looking at Lossau's book again I notice his dates are for the bayonet mount version while the entry for the M-42 mount version is undated, so I guess those ones kept being made until then.

According to Jurgen Lossau's pretty reliable catalogue in "Filmkameras" the K3 was manufactured from 1971-1976. I doubt there are many notable credits, it's really more of a beginner's camera, the most commonly quoted factoid is that parts of Spike Lee's Get On The Bus were shot with one. This site devoted to the K3 mentions a few music videos as well: http://www.k3camera.com/ Plenty of stuff of varying quality on youtube or vimeo, often camera tests.

Using a follow focus with a higher gear ratio (either one specifically designed for short throw stills lenses or with the ability to change drive gears) will increase the effective throw at the control knob, where you can mark focus distances on the disk. The disks are interchangeable, so you can have ones dedicated to each lens. Obviously the quality of the follow focus, in terms of back-lash and rigidity, will have an effect on how accurately the system works. Unfortunately with some stills lenses, the mechanics of the focus mechanism itself can have back-lash, so even with the best follow focus unit precise focus pulls can be hard to achieve. A discrepancy might creep in when the direction of the focus turn changes. You'd need to test each lens to see how bad it is.

The vast majority of C-mount lenses made for cine use have a mounting thread that only protrudes between about 3 and 6mm behind the lens mount, and always covers the rear element. The exceptions are very old lenses made for early 16mm cameras, but once turret cameras were introduced in 1927 lens manufacturers began limiting the rear protrusion of C-mounts to allow the turret to rotate over the aperture plate that covered the shutter within. C-mounts made for TV or CCTV had no such restrictions however, and can sometimes protrude far enough to cause damage. I recently repaired an ACL that had mirror damage due to the incautious fitting of a TV lens. Super 8 Beaulieu cameras like the 4008 with an oscillating mirror have an aperture plate behind the C-mount that tends to prevent these lenses being screwed in so far as to hit the mirror, which is probably a good idea to consider with this new camera. A C-mount is the logical choice for a reflex mirror S8 camera, plenty of lens options and adapters for other mounts, although maybe a micro 4/3 mount would work also. Anything with a shorter flange depth like CS or D-mount leaves very little room for the mirror. The French firm Ercsam pioneered the reflex guillotine mirror with their reflex Camex standard 8mm cameras in the late 50s but had to devise their own mount with a longer flange depth than the 8mm D-mount, and have custom lenses manufactured by Angenieux and Som Berthiot. Beautifully built cameras by the way, worth examining if you're devising a new version. Nizo and Beaulieu were the only other companies I know of that copied the guillotine reflex mirror system. Nizo used fixed and optically adjusted D-mounts with an extended flange depth while Beaulieu managed to squeeze the mirror within the D-mount standard FFD of 12.29mm for their standard 8 cameras. For Super 8 though they moved up to C-mount. The standard focal length for Super 8 is about 15mm which sort of precludes the use of most larger format lenses for anything other than telephoto, so I'm not sure a relatively expensive intermediate mount system like the P&S Technik one is worth the trouble. There is also the Eclair TS intermediate mount as used on the ACL which does the same thing, but it makes more sense for 16mm.

From the perf shape it looks more like 16mm to me.

Yes I suspect most wide C-mounts won't cover super 16, certainly very few I've tested. At 25mm you might be OK, but wider than 16mm and it gets pretty tight. The 10mm Kern Switar just covers, with slight darkening of the corners, but I haven't found any others that wide. As far as "best" goes, it is a bit subjective, as Philip said. You can generally assume that cine lenses made by the better manufacturers like Kinoptic, Kern, Angenieux or Taylor Hobson are pretty well corrected. Other classic manufacturers like Dallmeyer, Hugo Meyer, Schneider-Kreuznach or Som Berthiot can be good but are a little more hit or miss, and often lower in contrast. American brands were generally derivative and a bit low-end, though Kodak's Cine Ektars were OK, and lenses by other Rochester firms like Bausch and Lomb, Wollensak or Elgeet might be worth a shot. In later years some of these companies also made C-mount lenses for TV or CCTV (video) cameras which might cover a larger image circle, but the specifications would have been lower than for the cine versions. Most Japanese made C-mounts would be in this category, but they can be surprisingly good in terms of correction, maybe not much character though.