Jump to content

Buying film for Bolex, and other questions


Jon O'Brien

Recommended Posts

Hi, I'm about to order a roll of 16mm. Is there anything that I need to specify to the supplier about roll or perf type? It's for a S-16 Bolex reflex and the only thing I know is that it's going to be 100' Kodak Vision 3 50D.

 

A video on youtube says that for a reflex Bolex you need to open the diaphragm about 1/3 of a stop over what the light meter says, due to the viewfinder prism diverting light away from the film. Is this correct? It makes sense:

 

 

Another question if anyone knows the answer: is a Bogen 3066 fluid head designed to have the horizontal pan travel locked down? I've located a tripod and head but the seller says he's not sure if the travel can be locked. Thanks for any help.

 

 

Link to comment
Share on other sites

  • Premium Member

You do need to overexpose by about 1/3 stop to compensate for the prism diverting light to the viewfinder.

 

The manual has exposure times listed, as well as 'adjusted' exposure times that compensate for the light loss. So for a reflex Bolex with a 133 degree shutter, at 24fps your exposure time is 1/65 sec, but the adjusted exposure time that compensates for the prism is 1/80 sec.

 

Some people prefer to alter the ASA setting on their meter instead, some set it to 24 fps cine mode (which assumes a 180 degree shutter so 1/48 sec) and open 2/3 stop to approximate 1/80 sec, whatever you find easiest.

 

There is occasional misinformation out there that states that RX lenses have their f stop marks adjusted to compensate for the prism light loss but this is not true, they are marked like any other lens.

Link to comment
Share on other sites

  • 3 weeks later...

Thanks Dom, very helpful. I've read Nanolab's very informative page about using handheld metering too. I will have to do what he suggests, and do an exposure test.

 

As a very approximate guide, though, it looks like the Bolex reflex 5 model needs to have the lens opened up about 1/3 of a stop to compensate for the viewfinder light loss, and opened about 1/3 of a stop more to compensate for the fact that the light meter (for instance a Sekonic L398A) is calibrated for a 180 degree shutter, whereas the Bolex has a 133 degree shutter and is thus not letting in the same amount of light.

 

Have I got that right?

Link to comment
Share on other sites

I recently shot some 100 ASA 16mm film without any light meter at all, for the simple reason that on the day of the shoot, my 25 year old digital light meter finally kicked the bucket. I had to guess the f/stop. Surprisingly the results were quite okay. So along with stuff I've done without a viewfinder, I can add "without a light meter" to my list of achievements.

It is generally accepted that 2/3rds a stop (0.66 stops) compensation is required for the Rex Bolex, to compensate for the odd shutter angle and the prism. There are a number of reasons for accepting this, and the principle one is that filmmakers using this rule of thumb get good results.
Using this information, along with knowledge of the shutter angle, we can (out of pure curiosity) calculate the percentage of light lost to the viewfinder (by the prism).
The literature suggests a fully open shutter angle on the Rex Bolex is 133 degrees or 135 degrees. We'll use both values in our calculation.
The light admitted by the shutter (compared to a 180 degree shutter) can be calculated like this:

133 / 180 = 0.7388' (73.88 %)
135 / 180 = 0.75 (75.0 %)
A nice round result of 75% may very well indicate that such a value is a "rule of thumb" value rather than an actual value, but in any case we'll work with both values and use the difference in stops to estimate whether the difference matters.
Now if the Bolex did not have a prism, the required compensation in stops (for the shutter only) can be calculated like this:
133 degree shutter: log2(1) - log2(0.7388') = 0.4365 stops
135 degree shutter: log2(1) - log2(0.75) = 0.4150 stops
We can see the difference in stops between the two shutter angles is negligible: 0.0215 stops.
In any case, assuming the full compensation for a Rex bolex is 0.66' stops (2/3rds of a stop), then the light lost to the viewfinder, in stops, would have to be:
total compensation - shutter compensation = prism compensation
133 degree shutter: 0.66' stops - 0.4365 stops = 0.23 stops (prism loss)
135 degree shutter: 0.66' stops - 0.4150 stops = 0.252 stops (prism loss)
Converting such back into fractions/percentages, we get (if I'm not mistaken):
1 - (2 ^ -0.23) = 100% - 85.3% = 14.7 % lost
1 - (2 ^ -0.252) = 100% - 84% = 16 % lost
But as a check for errors lets perform the forward calculation where we should get 2/3rds of stop as a result (for both shutter angles). We compute the amount of light that reaches the film with:
shutter admission x prism admission
for 133 degree shutter: 73.88% x 85.3% = 63%
for 135 degree shutter: 75% x 84% = 63%
So far so good. Converting such to stops we should then obtain the original rule of thumb (2/3rds of a stop).
log2(100) - log2(63)
= 6.6438 - 5.9773
= 0.6665 stops (2/3rds of a stop)
Assuming there is no error in this calculation (based on the check), and that 2/3rds of a stop is correct (based on experience), and that the shutter angle is 133 degrees or 135 degrees (based on how could this be wrong), then the light lost to the viewfinder must be in a ballpark of:
14.7%
16%
Or to put it another way (on the same assumptions), were the prism to divert more than this, then the required compensation in stops would need to be more than 2/3rds of a stop.
C
Link to comment
Share on other sites

Here's an interesting post I just came across, giving an account of different shutter angles on different models of Bolex.

 

http://bolexh16user.net/ExposureAdvice.htm

 

Of interest to me is a paragraph on "133 degree" shutter models. According to the article, the 133 degree shutter angle is an incorrect shutter angle. It was inferred by a particular author from information in a Bolex manual where an exposure time of 1/65th sec, for 24 fps, was recommended.

 

In reverse engineering the shutter angle from the exposure time, the author of the 133 value had used the following formula to obtain that shutter angle.

 

Shutter Angle = (1/65) x (360) / (1/24) = 132.9

 

In reality (to follow the article) the shutter angle was actually 130 degrees (ie. physically so), and the exposure time of 1/65th sec (given in the manual) was simply a good enough approximation. If we otherwise need the exact exposure time for a 130 degree shutter angle, it can be obtained like this:

 

​ T = 130 / 360 x (1/24)

​ = 0.15046296 sec

 

which is between 1/66th second and 1/67th of second.

 

​To account for a prism at 24 fps, the exposure time given in a Bolex manual is 1/80th sec (which is another way of "tricking" your light meter into giving you a good aperture setting). Like the 1/65th second value, the 1/80th sec value is probably a "good enough" value as distinct from an exact value.

 

​C

Edited by Carl Looper
Link to comment
Share on other sites

The stop compensation can be reverse engineered from the exposure times given in a bolex manual. If we otherwise assume that 1/80th sec (at 24 fps) is a good enough value, then stop compensation can be reverse engineered from such:

 

log2(1/48) - log2(1/80) = 0.737 stops

 

We see that the rule of thumb value of 2/3rds of a stop (0.66 stops) is very close to this computed value.

 

If we look at all the values provided in the bolex manual for a 130 degree shutter compensation with prism compensation, we see that the compensation value is either 0.781 stops or 0.737 stops, suggesting 0.759 stops might be a good number to use as a more accurate number. If one ever needs such a number.

 

64 fps: log2(1/128) - log2(1/220) = 0.781 stops

48 fps: log2(1/96) - log2(1/160) = 0.737 stops

32 fps: log2(1/64) - log2(1/110) = 0.781 stops

24 fps: log2(1/48) - log2(1/80) = 0.737 stops

18 fps: log2(1/36) - log2(1/60) = 0.737 stops

16 fps: log2(1/32) - log2(1/55) = 0.781 stops

12 fps: log2(1/24) - log2(1/40) = 0.737 stops

 

 

 

C

Link to comment
Share on other sites

  • Premium Member

Before calculating too far the basics need to be firm. With which Paillard-Bolex camera do we have 130 degrees opening angle? The subject has been treated before, http://www.cinematography.com/index.php?showtopic=11912&do=findComment&comment=365326, but reverse engineering is against a technician’s approach, especially mine.

 

I understand every questioner very well, it’s natural to be curious. Yet we have to live with mean values and uncertainties, for example when considering the fact that a frame is not evenly exposed across (or aslant) its entire width. The reason for that is the sectional obscuring and unveiling of the aperture by the shutter edges. Attempts have been made to compensate for it, Paillard-Bolex changed the shutter blade edges from stricly radial to diagonal. A higher-order curve is actually necessary for equal distribution of exposure time of every point of the image. To comfort everybody: there were cameras that expose perfectly evenly, namely the 1897 Smith, the 1915 Akeley, and the 1924 Kinarri.

 

Let’s reckon with 135 degrees open angle for a PB H-16 Reflex VS. Deduct 25 percent of light as reflex finder loss. Accept two percent of difference in exposure time between left and right image sides. Not much compared to about 8 percent of exposure time difference between upper and lower image halves with Guillotine-type shutters.

Edited by Simon Wyss
Link to comment
Share on other sites

According to the article, the 130 degree shutter belongs to those cameras which have the published shutter time values of 1/65th (non-reflex) and 1/80th (reflex) at 24fps.

 

These shutter times are compensation values, ie. to compensate for the non-180 degree shutter angle and (for reflex cameras) the prism. The other way of compensating is to use 180 degrees as the shutter angle (1/48th of a second as the shutter time for 24 fps) and then compensate using aperture or ASA/ISO. The latter is convenient as it means one can read aperture values directly off the meter.

 

Now lets suppose we did "reckon with" the shutter as 135 degrees, and the prism loss as 25%.

 

That would mean the shutter admits 135/180 = 75%

And the prism admits the same: 75%

 

The total amount of light reaching the film would then be: 75% x 75% = 56.25%

 

The aperture compensation would then be reckoned as:

 

log2(100%) - log2(56.25%) = 0.83 stops

 

Now normally (if not invariably) this would hardly matter, as the difference between 0.66 stops and 0.83 stops is only 0.17 of a stop.

 

But there are situations where it might matter.

 

C

Link to comment
Share on other sites

Here's an example where it could matter what values we're using for shutter and prism loss, and relates to the original post.

 

Suppose we want to compensate using ASA/ISO. Armed with the values of 135 degrees for the shutter angle, and 25% for the prism loss, we would conclude the total light reaching the film was 56.25% (as calculated above)

 

With 50 ASA film, we'd then compute the compensation as:

 

50 ASA x 56.25% = 28.125 ASA

 

Looking at the available light meter settings (on a digital meter), we'd see 25 ASA or 32 ASA are the available options. And based on what we calculated we might very well choose 25, when we should have chosen 32.

 

C

Edited by Carl Looper
Link to comment
Share on other sites

Thanks Carl, and Simon, have read everything you wrote. Am shooting the first reel in just a few days time. I've been very slow getting around to it - but soon the film will be whirring through the gate - I promise. Then ring up Werner and hope all is well for getting it processed. As far as I can see, 2/3 of a stop will do for the first reel. Then see what the exposure generally come out like, and adjust from there if necessary.

Link to comment
Share on other sites

a frame is not evenly exposed across (or aslant) its entire width. The reason for that is the sectional obscuring and unveiling of the aperture by the shutter edges. Attempts have been made to compensate for it, Paillard-Bolex changed the shutter blade edges from stricly radial to diagonal. A higher-order curve is actually necessary for equal distribution of exposure time of every point of the image. To comfort everybody: there were cameras that expose perfectly evenly, namely the 1897 Smith, the 1915 Akeley, and the 1924 Kinarri.

 

Thanks for this info Simon.I found this quite fascinating.

 

On face value a rotational radial design (amongst others) seems as if it should work. For any location under the shutter, one might think that such a location should experience the same exposure interval as any other location. While locations closer to the centre of rotation have a shorter arc between the open edge and closed edge (of the blade) those blade locations are moving along that arc at a correspondingly slower speed. The net effect (one might think) is that all locations on the film should experience the same exposure time, if at different times (phase).

 

In an effort to understand why there should be a difference in exposure between one side of the frame and the other, we can expand the model of light used in the above. For example, I can see the above is based on a model in which a particle of light travels along a line to the film, passing through the shutter along such a line. But if we otherwise model light as a wave, it spreads out from it's source as a sphere, a subset of that wavefront being captured by the lens, before coming to focus on the film (or not as the case may be). Either way the light can be understood, at the time it passes through the shutter, as an "out of focus" point (more often than not). Or to put it another way, for each point on the film there is a correspondingly out of focus "point" (a circle) on the shutter blade. And it is between a circle and the shape of the shutter blade that one should compute how much light reaches the film at any location. Deliberations on such can then result in a better shape for the shutter.

 

 

Carl

Link to comment
Share on other sites

Thanks Carl, and Simon, have read everything you wrote. Am shooting the first reel in just a few days time. I've been very slow getting around to it - but soon the film will be whirring through the gate - I promise. Then ring up Werner and hope all is well for getting it processed. As far as I can see, 2/3 of a stop will do for the first reel. Then see what the exposure generally come out like, and adjust from there if necessary.

 

With negative film such as Vision3 50 D, being out by one stop from an ideal exposure, isn't going to matter very much at all. The stock has such wide latitude I imagine you could be out by two stops and it would still look brilliant. It's with reversal film stock that a tradition of fussiness emerges (aiming to be within one third of a stop to the ideal exposure) because reversal stock has limited latitude and so is much more difficult to correct in post. Indeed, traditionally, there is no "correct in post" as the camera original is shown, as is, in the projector!

 

Of more importance (as a rule) is consistency (whether exposing negative or reversal). Between shots that are meant to look the same (same location, same light) it's of more importance that the shots use the same exposure, ie. regardless of whether they are "correct" exposures. There are, of course, exceptions to this rule.

 

C

Edited by Carl Looper
Link to comment
Share on other sites

  • Premium Member

Phew, a professor got lost to/with/at you, Carl. Which preposition is the right one?

 

There is a similar problem with a photographic chamber shutter. If we take the leaf or central shutter, we have little to complain about. But focal plane shutters can cause distortion in moving objects. An image by Jacques-Henri Lartigue, 1912, © Estate of Jacques Henri Lartigue:

 

lartigue1.jpg?w=620&h=417

Link to comment
Share on other sites

Phew, a professor got lost to/with/at you, Carl. Which preposition is the right one?

 

There is a similar problem with a photographic chamber shutter. If we take the leaf or central shutter, we have little to complain about. But focal plane shutters can cause distortion in moving objects. An image by Jacques-Henri Lartigue, 1912, © Estate of Jacques Henri Lartigue:

 

lartigue1.jpg?w=620&h=417

 

I understand the distortions that can occur with shutters exposing a dynamic scene, since each location of the film is being exposed at a different time. But I've yet to correctly understand why a radial shutter design (exposing an otherwise static scene) would experience a difference in exposure time across the frame. When I try to model it using a circle pattern at the shutter plane (based on a wave model) I'm still getting an even exposure across the frame. I take it as correct that there is a difference (motivating high order curve designs), but am having trouble modelling it.

 

C

Edited by Carl Looper
Link to comment
Share on other sites

Hi Simon,

 

I'm going to suggest that the reason there is a difference in exposure between one side of the frame and the other, on film exposed with a variable shutter Bolex camera, (as small as this difference might be) is due to the variable blade (for whatever reason) being other than along a radial line. In other words, I'd suggest the diagonal design isn't solving any problem to do with exposure time, but creating one in the first place.

 

At least that's the current conclusion I've reached trying to model (understand) the situation.

 

I am, of course, open to correction on this.

 

C

Link to comment
Share on other sites

Ok, I've worked out the reason for the diagonal design.

 

It's for when the shutter angle is being made smaller (towards closed). For a fully opened shutter (130 degrees) a radial design would be appropriate (and optimum), but for a shutter approaching closed (smaller angle) it will reach a certain angle where both blade edges are shadowing the film at the same time. If a radial design were maintained, areas of the film further from the centre of rotation would receive more light than areas closer to the centre of rotation (for the duration of the exposure). The diagonal design (or a high order curve design) would act to compensate for such.

 

C

Edited by Carl Looper
Link to comment
Share on other sites

  • Premium Member

Now that’s technical thinking, along what one has. Regardless of the opening angle radially straight edges cause uneven exposure times across the image line intersected by those edges. With the ARRIFLEX, CINEFLEX, and the early Arriflex 16 designs the intersection goes over the image height, with a Camerette/Caméflex Eclair and most other younger designs it goes over the image width. I do not know of a movie camera that has a simple disc shutter whose axis lies somewhere within the film’s width. The bigger the opening angle is, the smaller the difference in exposure time becomes. As you say, only towards almost closed do we have a noticeable error. One should study fades. What a cool job . . .

Link to comment
Share on other sites

If the angle of the shutter blades is such that only one blade can intersect the film frame at any given moment, then a radial design seems to me optimum. For any given moment in which one blade splits the frame in two (between shadow and exposed), one can find a reciprocal moment in which the other blade has divided the frame in the exact opposite way. It would only be when both blades are intersecting the film frame at the same time, that they can no longer act as the inverse mask of each other.

 

C

Edited by Carl Looper
Link to comment
Share on other sites

So here's a graphic visualisation of the logic I'm using.

For any shutter orientation (left column) can be found (at some other time) a shutter orientation (right column) which exactly counter balances the exposure given by the orientation in the left column.

But if both edges of the shutter occupy the frame at the same time (third row), there is no such counter balance to be found (at any other time). This motivates an alternative to a radial design.

 

post-48441-0-51394600-1509912789_thumb.png

Edited by Carl Looper
Link to comment
Share on other sites

Just a quick question, probably a dumb one but I've looked around the internet and can't find the answer. Haven't read the preceding posts yet but will, with interest, later on. So here's my question.

 

You've just filmed a full 100' roll of film in your Bolex. Lo and behold, the film is now all on the bottom take-up reel. Before sending the film off for processing, do you have to re-wind the film back on to the top reel, in-camera (with shutter closed and lens caps on etc. and motor disengaged - I have a re-winding lever btw). Or do you simply pop out the full bottom reel, in semi-darkness of course, put it back in light-proof box, and post off for processing? I last filmed in 16mm in 1983 and can't remember what to do! Nearly everything I did was on Super 8. Thanks for your help.

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...