Jump to content

Image quality and sharpness comparison


Julian Fletcher

Recommended Posts

Hi,

 

Please can you help me answer a bit of a conundrum in my head. I have been using my Bolex H16 super 16mm camera more and more confidently and getting reasonable results. I have been mostly using the Switar RX lenses. As you know, shooting on film and scanning at 2k is not a cheap hobby. I know that the lens is the biggest contributor to image quality (assuming good scans), and that the camera is only ever a box that transmits the light onto a film. However, I have always wondered if the likes of Arri and Aaton super 16 cameras will inherently produce a sharper (not more stable) image. This has just come from what I have seen on Vimeo etc.

 

Please imagine this scenario.

 

* 2 cameras - one Arri SR3, one Bolex H16

* Both have the same lens - say a Zeiss Masterprime 75mm and shooting is at f8.

* Its PL mount, and is attached to the Bolex via a good adapter.

* Kodak Vision 3 50D is used and the same subject is shot

* Tip top 2k Scan onto DPX

* Assume that the prism issue with the Bolex is no longer an issue at 75mm and f8.

 

I know that being pin registered the Arri's image will be more stable, but would there be a difference in image quality and sharpness between the two? Would the Bolex produce a softer image, or be just as sharp? I know that the RX lenses correct for it's prism, but I have read that Bolex themselves suggest its effect is absolutely minimal when a) the lens has a focal length of 50mm+, and the lens is stopped down to f5.6 and beyond.

 

What do you think guys?

 

Cheers

Julian

Edited by Julian Fletcher
Link to comment
Share on other sites

  • Premium Member

Pin registered or not, all things being equal, an image shot with the same lens, on the same film, camera body will matter little.

 

I have a non pin registered camera, it produces stable images.

  • Upvote 1
Link to comment
Share on other sites

  • Premium Member

I think you're correct that the lens plays a very important part. Not just the quality of the optics and coatings but the mechanical build and how well you can fine tune the focus. But sharp focus is also dependent on how well the operator can judge it. In this regard a modern Arri or Aaton viewfinder will far surpass a Bolex one.

 

Anything placed in the optical path affects image quality, so even with the lens stopped down the Bolex prism has an effect, however small. Any dirt, scratches or smudges on the glass will obviously be detrimental.

 

There is actually a lot more to a camera than just being a box that holds the film. Things like mechanical vibration, accuracy and flatness of the flange depth, how accurately the ground glass depth matches the film plane, the gate design and whether it holds the film perfectly flat, all contribute to the quality of each individual frame. Variations in this quality together with registration accuracy determine how sharp the image looks in motion. Things like light tightness, and the quality of the transport mechanics determine whether the film gets flashed, scratched or bruised, all of which affect image quality.

 

I'm not saying Bolexes or any other older cameras necessarily have any of these defects, but the more modern professional cameras can be very accurately set up to achieve the best image quality possible from the lens/film stock combination.

Link to comment
Share on other sites

  • Premium Member

The Bolex pulldown system is very unique because it can run forwards and backwards. It's kind of a compromised system in a lot of ways, even though it's pretty ingenious and mechanically works well.

 

Dom hit on the key points... but the critical one that separates the Bolex from the modern Aaton/Arri cameras really comes down to that prism and viewfinder system. It's really hard to focus a Bolex to the level you can of the other, newer cameras.

 

Having spent a great deal of time shooting with both wind up and electronic Bolex cameras, I've not been able to get the stability of the Aaton or Arri cameras. There is always a little bit of jitter, which can be corrected digitally. You just don't get that with the newer cameras.

 

But to answer the question... the difference is really in the stability. The Aaton and Arri gates, pressure plates and pulldown systems are a lot better then the Bolex. That superiority is really what makes the rock stable images we're accustom to seeing from modern S16mm. So gate wobble and focus issues, would be the two deciding factors. Now that modern S16 cameras are so inexpensive, the benefits of the Bolex are becoming less and less. Which is too bad because they're still great little cameras.

  • Upvote 1
Link to comment
Share on other sites

  • Premium Member

I have a somewhat different view. An Arriflex doesn’t have better registration than a simple claw movement, both mechanisms adjusted. The Paillard-Bolex H claw with serial number 100,401 and higher together with proper side guidance actually positions the film with no play.

 

post-35633-0-13416600-1469262257_thumb.jpg

 

Here is an H-16 film gate that I have altered from the wrong way around to conforming DIN ISO 69. I have ground the aperture plate to a width of 15,92 mm, a tad under the lower deviation of the film width tolerance, and weakened the leaf springs on the left side so that the film is not warped like mad. In its original state the H-16 pushes the film to the left by the long leaf spring at a point below the claw travel. The aperture plate has 15,85 mm width there but over 16 mm in general. Please note that the scratched film rails have not yet been buffed when I made the photograph.

 

Register pins must fit the perforation holes (plural!) also with no play to compete and there is the rub. We have two hole edges and one pin size that must be brought to fit. Tapered pins is the solution but we still have play with the pin guidance and/or bearings. The standard Arriflex 16 has the better mechanical arrangement than the later models in this respect. It is even better than what the Mitchell 16 offers (at lower speeds). Cancellation due to identical vertical register is also part of the game. The SR3 positions in the +3 hole like the Paillard-Bolex H-16. The Arriflex 16 positions +1.

 

Optically, the Paillard-Bolex H reflex prism system is doubtless inferior to no glass between lens and film. We have the very simple and therefore rugged arrangement of the matte surface etched directly into the upper of the two block prisms but the light path through glass is longer between lens and focusing screen than between lens and film. In fact, the light towards the film leaves the prism block while the light towards the matte surface doesn’t. This asymmetric situation is acceptable for long object distances. With macro and micro setups, however, we can be fooled by the inverted geometry. Mirror-shutter reflex cameras on the other hand need careful selection of shutter bearings plus a minimized run-out. Less than 4 tenths is tolerable, 0.0003" max total run-out preferred.

 

Standard Paillard-Bolex H cameras on rackover device offer the old reliable focusing system which can be checked and aligned with relatively little effort by a specialist. By the way, at f/8 you can shoot with simpler lenses than a Zeiss-Arri Master Prime. A Kinoptik C-mount lens or a Switar will do as well, perhaps a Tessar variant such as the Wollensak Cine Raptar 1"-1.9 can keep up.

 

The lens mount is another point. You’d better compare an H-16 M to an other single-port camera. I have encountered non-flat discs and burrs on lens seats with H-16 turrets. Bolex cameras have aluminum parts where professional cameras are made of steel. The Zeiss Master Prime 75 weighs 2,8 kg or 6.2 lbs. No Paillard-Bolex H-16 is made for such weights.

 

Last, only big-base Paillard-Bolex H cameras must be taken into consideration. If you look at the Ciné-Kodak Special or the Arriflex 16 the body can rest on a support with a large area. The separation of professional cameras from amateur gear often has to do with inconspicuous things.

  • Upvote 2
Link to comment
Share on other sites

  • Premium Member

Hey Simon, what do you consider an "Arriflex 16" ? S/M, BL?

 

The original SR I/II's have some stability issues for sure. The SR III's don't. I was always told the reasoning was the gate and how the edges moved in the later generations. I understand how the registration pin can be an issue, depending on the interior size of the perf varying. A more standard pulldown system, will only use the bottom edge of the perf for registration alignment, which is a far superior method in a lot of ways.

 

What is your take on the "modern" SRIII's vs the original one's?

 

Also in the case of the bolex, wear and tear is the largest factor for the cameras registration issues.

Link to comment
Share on other sites

  • Premium Member

I have a somewhat different view. An Arriflex doesn’t have better registration than a simple claw movement, both mechanisms adjusted. The Paillard-Bolex H claw with serial number 100,401 and higher together with proper side guidance actually positions the film with no play.

 

We might have to agree to disagree on this. I'm actually not sure you can make this sort of statement without at least having viewed a number of steady tests of both cameras.

 

I viewed dozens and dozens of SR2 and 3 steady tests when they were regularly being rented for TV series, docos, TVCs, shorts and features from the rental house I work for. Steady tests were a standard part of the prep for anything with a decent budget. I regularly used to test up to 75fps on each mag, and if there was the slightest unsteadiness the camera and/or mag would be serviced. We viewed the tests projected onto minimum 20ft screens, the cameras were almost always rock-steady at every speed, certainly only steady ones went out on jobs.

 

A couple of times I saw steady tests shot on a Bolex, and while they were remarkably steady considering the simplicity of the Bolex movement, they were not as stable as an Arriflex, particularly at higher speeds. In my experience people generally don't shoot steady tests on a Bolex, they're not expected to be at that level of perfection.

 

Even with the careful modifications to the gate that Simon is talking about, there are other factors at play. Exactly controlling the gate width and side spring tension might help to control lateral unsteadiness or gate weave. But it has no impact on vertical unsteadiness. The Bolex claw is made of bendable aluminium and barely enters the film perf, pushing the film forward with a small edge that can and does wear. The claw then drags back over the film (during exposure) before moving forward again to engage the next perf. The pressure plate and maybe a bit of side rail pressure are the only things keeping the film stable during this time. The tension pushing the claw edge towards the gate as it shuttles back and forth is controlled by spring washers, grease, and the spinning of the shaft that the claw pivots on. A single, small ball bearing controls any play in the the claw pivot shaft as it rotates on a gear. It's a marvel of simplicity that works incredibly well, but to say it creates the same stability as the steel, hard chrome, multiple ball bearings and precision engineering of an Arriflex movement is a bit ludicrous. I mean, why go to all that trouble if it makes no difference?

 

It's partly a question of durability as well as precision, the materials used by Arri were much sturdier, the lubricants more varied and carefully chosen for application and longevity. And of course the later movements also needed to be quiet as well as precise.

 

There's also the question of what sort of viewing conditions the footage will be shown under. Stuff shot on an Arriflex might end up in a cinema whereas Bolexes were more often used by students and experimental filmmakers. They were made for different applications, with different expectations.

 

I've found the Bolex to be pretty stable, although the Arri SR has been seen to have some weave issues during telecine transfers in the past.

 

If Bolexes had been subjected to the same professional scrutiny as Arriflexes, they would surely have fared worse.

 

I was told by retired industry pros that the SR gate weave issue was due to the fact that telecine scanners often used different registration methods than the cameras, so while SRs guide the film edges above and below the aperture, many scanners used the film edge next to the frame for lateral registration. Any weave in the film stock itself would thus show up as the image weaving. The camera stability itself was very rarely the cause. There was a thread on this:

http://www.cinematography.com/index.php?showtopic=63117&page=2

 

Like any camera, wear will eventually cause issues, and SRs were probably the most widely used 16mm cameras during the 80s and 90s for professional use. Many would have seen millions of feet of film pass through them, in all sorts of conditions. Of course some will have ended up noisy, unsteady and unreliable, but a properly maintained one is not by its nature any of these things.

Link to comment
Share on other sites

  • Premium Member

The original SR I/II's have some stability issues for sure. The SR III's don't. I was always told the reasoning was the gate and how the edges moved in the later generations.

 

The "stability issues" of SRI/IIs was due to wear to the side rails after many, many thousands of feet passing through, which was easily solved by replacing the side rails on the gate. Rental houses and professionals who had their gear regularly serviced never had these stability issues. It's a bit like saying BMWs have "tyre issues" because after thousands of miles the tyres don't grip like they used to.

 

SR3s used saphire guides for lateral stability that don't wear, but the movement is just as susceptible to wear over time (or more correctly workload) if not maintained.

 

I understand how the registration pin can be an issue, depending on the interior size of the perf varying. A more standard pulldown system, will only use the bottom edge of the perf for registration alignment, which is a far superior method in a lot of ways.

The interior size of a film perf is cut with a die so only shrinkage due to poor storage or very old stock should cause any variations. At any rate a slightly undersized perf might cause noise problems but not instability, since the reg pin still holds the film still while engaged during the exposure phase.

 

 

Also in the case of the bolex, wear and tear is the largest factor for the cameras registration issues.

 

Wear and tear is the largest factor for any camera's registration issues, but amateur cameras tend to start with the bar set lower.

Link to comment
Share on other sites

  • Premium Member

In the interests of a comparison to Simon's Bolex gate, here's a pic of the gate and movement block from Arri's first 16mm camera, the 16S:

 

post-46614-0-88310100-1469350293_thumb.jpg

 

One point of interest is how the gate aperture has support rails on every side, so that there is no possibility that the film can bow in, causing focus discrepancies. It can take a shift in depth of only one or two hundredths of a mm (a piece of sticky tape is about five hundredths thick) for a sharp image to soften a little. Similarly, if the gate is not perfectly parallel to the lens mount, there can be a slight focus drift across the frame. But depending on the shot, the application or the final viewing resolution this sort of error might not even be noticeable.

Link to comment
Share on other sites

  • Premium Member

I can't speak to a Bolex, but the difference between multiple Scoopics I've owned and multiple SRs is noticeable. I love the Scoopic fixed lens and get excellent results with it, especially close up on people and things. I've had many complements from colorists on images from it.

 

Where I notice a difference is in large gradient areas in the image like a big blue sky. There's a pulsing that comes from the film transport in most Scoopics that can be very subtle, but there. I still love the Scoopic for its portability and small form factor.

 

With the three SRs I've owned (and are well maintained) never have that issue. My SRs just feel rock solid to me and are incredibly quiet. I have very nice versions of the "stock" 10-100 Zeiss/Arri zoom lenses (T3.1 and the T2) for my SRs which are amazing too.

 

As has been said, theoretically if these cameras are tuned perfectly only lenses should make a difference but a camera like an SR or LTR is designed to be tweaked and adjusted whereas a Scoopic has very little adjustment possible in the actual film transport so when something is off it's all about moving a little here and there which then knocks another part out a little. Not tech friendly from what I understand.

Link to comment
Share on other sites

  • 1 month later...

The solution to all this might be to own a Bolex, Scoopic or similar camera for personal use, to get into 16mm film making and learn how to use a camera well, and if you can eventually get to the point of making a feature film for cinema release, rent some Arriflex 35 or S16 gear for the shoot. Or buy and then sell the camera afterwards. By the way, although a lot of people say 16mm (and film in general) is an ancient dinosaur that is fading away, what happens if, contrary to popular belief, film starts to make a big resurgence? Will Arriflex, or anyone else, still have the machining skills and technical knowledge to get back into making great, professional 16mm cameras again? Or will it become like Stradivari and Guarneri del Gesu violins, endlessly maintained and re-used by the top pros? This is said partly in jest. I'm not seriously comparing Arri with Strad!

Edited by Jon O'Brien
Link to comment
Share on other sites

...."Now that modern S16 cameras are so inexpensive, the benefits of the Bolex are becoming less and less"

 

The so called "modern S16 cameras" and the Bolex present, presented, quite different technical and aesthetic opportunities. When the indies and artists got hold of the humble Bolex, all sorts of cool stuff was possible. People walked around single framing pixilated stuff. They undercranked stuff in cool ways, manipulating the shutter angle. The made DIY optical printers by mounting a Bolex in front of an old projector movement etc, expanding their undercranked frames into real time. Making fades and disolves.

 

I suppose you could try some of these cool things with the SR 1, 2, the Aaton LTR 7, 54. You could get or make an external device that might let you single frame. You could put up with the weight of the Arri or the bulk of the Aaton, feeling stylish, pushing to the corner of your mind the fact that they did not have a variable shutter.

 

If the benefits of Bolex are perceived by some as becoming less, it's possible they were never correctly perceived in the first place.

Link to comment
Share on other sites

  • Premium Member

Comparison among cameras such as Arriflex 16, Paillard-Bolex H-16, Beaulieu Reflex 16, Canon Scoopic, and more is like comparing eggplant to pineapple to tomato. Paillard had departed from the initial H camera design over the decades, partly for better and partly for worse. It’s an amateur camera, always was and never became anything professional. It is probably the best amateur camera. As Dom and I said, many things are aluminum with a Paillard-Bolex H (Bolex was the Geneva equipment of Bogopolsky 1923 to 1930). One of the incredible Paillard improvements is the built-in frame counter. One must not assume that it is reliable. It’s driven by friction and there’s friction a second time between the 1-to-50 wheel and the hundreds one. While a little error isn’t catastrophic with the hundreds only a quarter of one permille slippage results in a frame off over 100 feet (H-16). The thing is that one can’t see slippage happen.

 

I admit an H-16 compares unfavourably to a register-pin camera at higher speeds, no question. There we also have the cause for its typical noise, the travel is longer than full perforation pitch so that the claw beats the hole edges. The flung-on claw wouldn’t function reliably were the travel near perf. pitch.

 

The CP-16 movement has a very effective means for vertical positioning of the film. The film actually sits down on a hardened steel ball hole by hole, a centering method comparable to a tapered pin. The Ciné-Kodak Special, by the way, bears a unique lateral centering gate which was important for the use of Kodacolor. Described it over here (German).

 

The Arriflex 16 family offer ten times magnification of the GG. 13 and 14 times magnification comes in an H-16 EBM or EL finder. The Bolex 16 Pro affords 20 times magnification but a rather dark GG. Beaulieu also had a big view but some play in the mirror shuttle guide slots. The optics of a professional reflex viewfinder, ARRI, Eclair, Aaton, are an entirely different breed from the amateur finders. Not only the technical optical register is important, I think how well I can focus by the aid of the finder is as well. An image splitting device (Berthiot lenses with built-in finder), a microprism array or a fine enough GG makes the difference plus the eyepiece.

Link to comment
Share on other sites

  • 2 years later...

 

The CP-16 movement has a very effective means for vertical positioning of the film. The film actually sits down on a hardened steel ball hole by hole, a centering method comparable to a tapered pin. The Ciné-Kodak Special, by the way, bears a unique lateral centering gate which was important for the use of Kodacolor. Described it over here (German).

 

Simon, the CP-16 movement is the Auricon movement, correct? Actually manufactured by Auricon and I think unchanged throughout its history except for model CM-74 which had Sapphire inserts in the gate?

Link to comment
Share on other sites

  • Premium Member

It is but you obviously know the CP better than I do.

 

I’d like to bring in the subject of printing. We can discuss cameras and steadiness for ages, negatives are not projected. In the professional workflow theatre positives are made at some point and these, continuously exposed, will never contain the precision of which the original was located by the camera. You can have dimensionally stable polyester-base originals and polyester-base prints, the bringing together of original and raw stock on a printer makes the difference. I mean a difference for the spectators.

 

Now that film technics are almost forgotten about, thought of to belong in museums, intermittent contact printers in use have become still scarcer than they were thirty years ago. We’re in the field of 16mm film on this thread but I confess here that it makes me angry that 35mm productions have been worsened, strictly technically, in this respect. The film world didn’t fight the video menace by mechanically perfect prints and precision projectors. The latter are still the club foot of film cinema. To say nothing of second-class dupes

Link to comment
Share on other sites

  • 7 months later...

Simon, I don't know the CP well at all. Jeff Kreines set me straight on this: the CP movement was a copy of the Auricon movement (the patents were expired) with a few modifications. It was the precursor to the CP-16, the Frezzolini, that actually used bodies that were purchased for Auricon and (very) heavily modified.

While on the subject of 16mm movements and pin registration; I recently picked up a Milliken high speed camera and have been examining its pin-register movement. Attached is a photo with the pressure plate removed. Have you taken a look at these before? If so, what is your opinion of the quality? The register pin coming in thru the back is similar to the Arri S, as I recall.

 

milliken_gate_small.jpg

  • Like 1
Link to comment
Share on other sites

  • Premium Member

Yeah, similar to the Arriflex 16 with the difference of the swivel axis closer to the film and further to the rear. The rounded protruding knobs along the gate are a cheaper version of the CP ground steel balls which on their turn go back to the round flat pins with the 1933 Ciné-Kodak Special. A truly perpendicularly acting register pin is found in the Doiflex-16 which is not a copy of the Arriflex 16 but a design on its own. It was given a similar outer appearance because the Arriflex 16 had become the most successful professional small-gauge camera up to the early 1960s. I’m preparing an article on the more probable development of it, I mean more probable than what ARRI and its surroundings tell us.

Link to comment
Share on other sites

  • Premium Member

Not being an engineer but having owned many 16mm cameras, my take on different camera bodies with the same lens as stated in the original premise would be that an individual frame would be quite similar for multiple cameras with the same lens, it's the movement from frame to frame that is different. 

While I don't have a Bolex, I can say that my Canon Scoopic has beautiful individual frames, but I can see flutter in skies quite often whereas my SR2 is steady and continuous as a rock. I've been told the movement on a Scoopic is extremely finicky and difficult to adjust as changing one element leads to issues on another element and basically it wasn't designed to be worked on and adjusted as precisely as the Arri. Bolex's are fine cameras and I'm sure a well-maintained one will be great; it just may be harder to find qualified Bolex techs vs. Arri techs at least in the U.S. Not sure in London.

Link to comment
Share on other sites

The only way to settle this is to shoot some sample film with both setups. If I had the excess $$ or equipment I would get it done for you.

This is the type of thing that intrigues me. While I'm not a pixel peeper, I like to know where I stand. I do lots of available light shooting.  I don't have the luxury of adding light many times. I shoot all my lenses wide open, then test stopped down some.

I know which can produce decent results wide open and which can't. It is good to know if your setup will produce less than highest level results from the start. 

 

Edited by Daniel D. Teoli Jr.
Link to comment
Share on other sites

  • Premium Member
On 9/6/2019 at 2:37 PM, Will Montgomery said:

While I don't have a Bolex, I can say that my Canon Scoopic has beautiful individual frames, but I can see flutter in skies quite often whereas my SR2 is steady and continuous as a rock. I've been told the movement on a Scoopic is extremely finicky and difficult to adjust as changing one element leads to issues on another element and basically it wasn't designed to be worked on and adjusted as precisely as the Arri. Bolex's are fine cameras and I'm sure a well-maintained one will be great; it just may be harder to find qualified Bolex techs vs. Arri techs at least in the U.S. Not sure in London.

Usually the flutter is a shutter being slightly out of calibration. 

Link to comment
Share on other sites

Guys, I had a Beaulieu R16 (later donated to a museum here), which I did a steadiness (registration) test on it in the 90s with fine results as far as I remember.

It may be unfair to compare the semi-pro cameras with the pro ones mentioned above I think.

Doiflex 16 made by DOI Seisakusho was aimed at a reasonable-priced camera, there was no need to put a complex intermittent mechanism like in the Arriflex 16S/St , M and BL. It was well enough to utilize a simplified mechanism having two cams opposite each other and inside of frames as the followers. The one gives a sinusoidal pull-down movement while the other lateral for the registration pin. But the rest is similarly designed as 16S/St. At first, it was accepting 100ft da-lite spools and a 400ft. magazine provision was planned by modifying the camera top at the factory. However never became a reality. Also, it has a guillotine reflex mirror not a shutter, its shutter is focal. Very helpful for quick pans handheld to eliminate the stroboscopic effect.

 

 

Edited by Erkan Umut
Link to comment
Share on other sites

Stable image definition may be notional. To what extend? 

If you need rock steady images to be screened on a big screen (assuming the print and projection quality within the given standards) or conventional SFX work, then it really matters.

That's why you gonna see image steadiness data is given generally, and engineers were designed complex and quite expensive mechanisms, as well as near impeccable machining and workmanship were required, to achieve that.

Otherwise, I am free to say: I had a Nikon R10 camera with an excellent lens mounted and the resulting images were stable.

 

Edited by Erkan Umut
Link to comment
Share on other sites

  • 2 years later...
On 7/23/2016 at 6:59 PM, Simon Wyss said:

 

post-35633-0-13416600-1469262257_thumb.jpg

Here is an H-16 film gate that I have altered from the wrong way around to conforming DIN ISO 69. I have ground the aperture plate to a width of 15,92 mm, a tad under the lower deviation of the film width tolerance, and weakened the leaf springs on the left side so that the film is not warped like mad. In its original state the H-16 pushes the film to the left by the long leaf spring at a point below the claw travel. The aperture plate has 15,85 mm width there but over 16 mm in general. Please note that the scratched film rails have not yet been buffed when I made the photograph.

 


In an oscillation between the left side spring guide and the right side spring guide (were it not otherwise moved out of the way), the right side has more give, so the film would align to the left side guide, and between the left side guide and the right side hard points, the film would end up aligned to the right side hard points (because the left side has more give than the right side hard points)

This makes the right side spring guide appear somewhat redundant. But perhaps it's for film that isn't as wide as it should be. At least it will aligned to the left edge guide.

Carl
 

Edited by Carl Looper
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...