Jump to content

Exposing 7266


Steve Milligan

Recommended Posts

I need a sanity check please; I have my answer but I'd like to hear yours before I burn film.

 

You have a Bolex Rex4 (with Rx lenses) a roll of 7266 (process as reversal), and an 85N6 in the gel holder. If your meter is set to 180 shutter and 24fps, what would you set your meter ISO to so that you could transfer the F-stop directly to your lens, including an underexposure strategy if that's your bag?

 

I get 32, for what it's worth.

 

And thanks!

Link to comment
Share on other sites

  • Premium Member

Are you shooting a day exterior? 7266 is 200ASA in daylight and 160ASA under tungsten. If I remember correctly you lose 1/2 stop to the prism on the Reflex Bolex. Also, I'm not sure the normal shutter angle is 180 degrees, I seem to recall it's more like 135-150. So that's another 1/2 stop.

 

So under daylight before filtration you are down to 100ASA. 85N6 is -2 2/3 stop, so that's 16ASA. I'd definitely check the manual for the shutter thing.

Link to comment
Share on other sites

The Bolex shutter is 130°, combined with the semi-transparant mirror gives you 2/3 of a stop loss for the camera. The 85N6 will give you tungsten light filtration and 2 stops density, total of 2 of ND plus 2/3 for 85. This makes a grand total of 3 1/3 stop loss or indeed 16 ISO as Satsuki mentions.

 

I would suggest to do an exposure test beforehand because the reversal stock has very little latitude. Nothing looks worse than an underexposed or overexposed reversal film;

Link to comment
Share on other sites

Confirming Satsuki: The Bolex RX4 shutter angle is around 135 deg. while the reflex prism light loss closer to 1/3 stop than 1/2 stop. According to Bolex, the combination yields 1/80 second effective exposure at 24 fps. So the factor versus a straight 180 degree shutter is 48/80 = 0.6. Kodak publication B-3 indicates that the 85N6 gel will cut the light to 18%. So the ASA to set would be 200×0.6×0.18 = 21.6. Call it 20.

 

fiwh8b1261b7afl6g.jpg

My old Luna Pro meter with 1/80 sec marked for the Bolex RX.

  • Like 1
Link to comment
Share on other sites

Since he is using an 85N6 he is using 'tungsten light' (as seen by the film) so the basic starting speed would be 160 ISO.

With a setup like this I would really recommend a test beforehand. Unknown light meter + unknown camera + unknown lens + reversal stock = LOTS of room for error.

Link to comment
Share on other sites

Thanks for the math, everyone, that's exactly what I was looking for. It was the tungsten filter/tungsten rating that worried me...i.e. whether the 85N6 would cut the 200D ISO by 2-2/3 or 3. Looks like 3 is the consensus, which is logical.

 

Satsuki, yes, this is for the extreme case, bright exterior, 12fps, opened up--should it come to that. I have red, yellow, ND .3 and ND.6, but not ND.9...I thought 85N6 might be useful if it got that far over.

 

A further question: I derive 32 from consistent advice of more experienced shooters: expose reversal for the opacities, close it down as you might open up for a fat negative. I think for myself, I'd spot meter and expose dead on, but I'm handing this off to students. Is 7266 just too contrasty for this? Any other thoughts on exposure or metering technique? I'll be doing the scan myself on a JK.

Link to comment
Share on other sites

I use a Sekonic L398 and I always set it for 2/3 of a stop under my film stock's ASA. 1/3 compensation for the prism and 1/3 for the 135° shutter. After that, I don't have to think about my readings until I start playing with filters or change film speeds.

Link to comment
Share on other sites

I'm a firm believer in mathematical calculations for photography. Experiments, such as exposure testing, are costly in time and money and are best avoided IF there's a trustworthy calculational alternative. That's a big IF.

1. The calculation must be logical -- the photographic situation must be fully understood.

2. The data must be firm

3. The uncertainties that propagate through the calculation must stay tolerable.

 

On review, what Satsuki, Dirk, and I did was borderline illogical in how we found the filter factor for the 85N6 filter. This filter is seldom used with B&W films. The 7266 data sheet does not include a filter factor for it. Satsuki and Dirk used the 2/3 + 2 stop factor based on use of the 83N6 with daylight color films. I found a 0.18 factor, which is 2.47 stops, based on how dark the 85N6 looks in daylight to the human eye. B&W 7266 film sensitivity matches neither daylight color films' nor the eye's.

 

So what is the filter factor for the 85N6 with 7266 film shooting in daylight? The data for calculating that are available. Kodak publishes the spectral transmission of the 85N6. Kodak also publishes the spectral sensitivity of 7266 film. The CIE publishes the spectral power distribution of D65 daylight. Here they are.tkpf30saoqsf9k76g.jpg

 

You multiply the D65 curve by the 7266 curve and add it up over all wavelengths to find the unfiltered exposure.

You multiply all three curves and add it up over all wavelengths to find the filtered exposure.

The ratio is the filter factor. As calculated here from 380 nm to 640 nm the factor is 0.116 or 3.1 stops. Higher than Satsuki and Dirk figured, much higher than I figured.

 

The film is practically dead above 640 nm. But the film and daylight are both alive below 380 nm. This implies a filter factor more than 3.1 stops. Unfortunately Kodak doesn't publish the film's sensitivity below 380 nm, and we also don't know how transmissive the lens is below 380 nm. On the other hand, if you used a minus-UV filter over the lens, the 85N6 filter factor would become less than 3.1 stops.

 

So 3.1 stops is the filter factor, and again using Bolex's 1/80 second figure the new answer is ASA = 200 × 0.6 × 0.116 = 13.9.

 

I hope this exercise helps more than it hurts.

Edited by Dennis Couzin
Link to comment
Share on other sites

  • Premium Member

Dennis,

 

That's very interesting, thanks for that explanation and also thanks to Steve for asking the questions.

 

Can I ask a basic question:

 

 

You multiply the D65 curve by the 7266 curve and add it up over all wavelengths to find the unfiltered exposure.

You multiply all three curves and add it up over all wavelengths to find the filtered exposure.

The ratio is the filter factor. As calculated here from 380 nm to 640 nm the factor is 0.116 or 3.1 stops.

How do you do that/those calculation(s), forgive me if its a really basic thing?

Link to comment
Share on other sites

How do you do that/those calculation(s), forgive me if its a really basic thing?

 

First you need to make the three curves into numbers. D65 daylight and the 85N6 filter are published numerically, wavelength-by-wavelength. The film's spectral sensitivity is only available in graphs. So digitize the graph using the nifty "Digit" program available free from W. Theiss. The digitized graph will have funny wavelength values, so you must interpolate to get all three curves on the same wavelength scale. Also since the film's spectral sensitivity is published on a log scale, you must delog it -- replace it with 10 to that power.

If you work in Excel, you will now have 4 columns of numbers: wavelength, D65, 85N6, 7266. It's unnecessary to normalize the D65 and 7266 numbers, although that makes a nicer graph. The 85N6 numbers must stay as is; except if they're percent change them to decimals. Now you can make a 5th column from the D65 number × the 7266 number. Also make a 6th column from the 5th column number × the 85N6 number. Add the 5th column of numbers. Add the 6th column of numbers. Divide the second sum by the first sum. I got 0.116. That filter knocks the daylight exposure of that film down to 11.6%. You don't need to convert this to stops to solve Steve Milligan's problem, but the filter factor is of interest in itself. The log of 0.116 divided by the log of 2 gives you -3.1 for the stops. This means the factor for an 85 filter is 1.1 stops, significantly different from the 2/3 stop we use for it when shooting tungsten balanced color films in daylight.

 

 

Or you could just use a spot meter, metering through the filter. That's what I would do, but I'm terrible at math.

If the spot meter measures nits, footlamberts, or any photometric quantity accurately, then it will give you the lousy 2.47 stops figure I used at first. 7266 responds to the spectrum its own way, shown by the black curve which is different from your meter's.

Edited by Dennis Couzin
Link to comment
Share on other sites

  • Premium Member

Dennis,

Thanks for taking the trouble to do all that. I'll bash my head against it for the next while (couple of years given the way my brain works) and I may come back to you with some more questions if you have time or else just point me to a text etc.

 

Much appreciated.

 

Jeremy

Link to comment
Share on other sites

I have never processed 7266, but plenty of reversal 7242,7241,7256,7252, 7250,7251,7239 (in order of appearance). A couple of years ago I had an interesting technical conversation with the tech guys at Kodak Chalons about the true speed of 7222. After exchanging many sensitograms (they expose, they process, we expose, we process and then all change), we came to the conclusion that yes, the 7222 does not have its rated speed and secondly and most importantly, every decent cinematographer should test before production and establish his own look and speed with his own equipment.

 

So, unless you have a blind faith in Kodak, lab processes, lens stops, exposure meters, filter factors etc, etc, please do some real-life testing before starting.

  • Like 1
  • Upvote 1
Link to comment
Share on other sites

  • Premium Member

It's probably faster to just shoot a test roll and bracket in 1/3 stops at this point. I've shot and projected a lot of 7266 and it always comes out looking great except in low light, low contrast situations. I usually use a Red #25 filter for exteriors. Surprisingly for a reversal stock, I've found 7266 to be very forgiving of slight exposure variations.

Link to comment
Share on other sites

I have never processed 7266, but plenty of reversal 7242,7241,7256,7252, 7250,7251,7239 (in order of appearance). A couple of years ago I had an interesting technical conversation with the tech guys at Kodak Chalons about the true speed of 7222. After exchanging many sensitograms (they expose, they process, we expose, we process and then all change), we came to the conclusion that yes, the 7222 does not have it's rated speed and secondly and most importantly, every decent cinematographer should test before production and establish his own look and speed with his own equipment.

 

So, unless you have a blind faith in Kodak, lab processes, lens stops, exposure meters, filter factors etc, etc, please do some real-life testing before starting.

 

Dirk, your whole list of reversal stocks (which are not in order of appearance) are color Ektachromes. 7266 enjoys a simpler sensitimetry and simpler processing than those. But 7222 has yet simpler processing so it's sad to hear that 7222, which has been around since 1959, no longer matches its published characteristic curve. Do you think it is the film -- Kodak's manufacturing gone downhill -- or your and the French Kodak's sensitometry? How are you simulating "daylight" for the sensitometry? How close to photometric is your lux metering? Has the status M blue filter been spectrophotometered? Etc. Etc.

 

When the film manufacturer can no longer produce film with consistent speed, you indeed must test each emulsion batch. That is if you can trust the laboratory to process consistently. When you can't trust your exposure meter either, then it's really time to switch to video where you can see and scope the image as you shoot it.

 

Is the question in this strand still suited to calculation? I understand the question as: If I know how to expose 7266 with a standard camera and no filter, what should I do differently with a Bolex RX camera and 85N6 filter. The Bolex RX means you use 1/80 sec instead of 1/48 sec. That part of the calculation is solid. For the 85N6 filter factor, post #8 showed how it depended on the spectral distribution of daylight, the spectral sensitivity of the film, and the spectral transmission of the filter. How solid are these?

  1. D65 daylight is in CIE-ISO standards, but will the daylight when you shoot resemble D65? If it doesn't, film testing some days before won't help you. If you're worried about the daylight quality's effect on the filter factor and thereby on the exposure, then you'd better buy a spectrometer and measure the daylight when you shoot.
  2. The spectral sensitivity of the film is less likely to be off the published spec than the film speed that you found to be off because it depends on the emulsion formulation but not the coating. Yet, if a manufacturer is slipping that could slip.
  3. The 85N6 is a gel. Wratten gelatin filters, later sold as Kodak filters, used to be well controlled as they had many scientific uses outside photography. Though many have been discontinued, they're easy to manufacture. The last time I spectrophotometered some Kodak gels was in around 1995 (for a non-photographic use) and they were good. I hope they still are but don't know.

Post #8 stated that the calculational method requires both that the photographic situation is fully understood and that the data is firm. Hopefully the longer one works with film the better his understanding. I actually quit film in 1993 but continued to think about it, and came to be ashamed at how much film testing I did for lack of understanding and good modeling. While our understanding may be increasing, our data is simultaneously decaying because of the sad state of the industry. So the calculational method does struggle.

Link to comment
Share on other sites

Dennis,

 

the reason people still shoot film is because it looks different from digital (in a good way). You could argue why some artists still use oil or watercolors for painting while photography has been around for 150 years or more and digital pictures can be taken with most phones now.

As a lab owner/colorist I have to balance between science and art. There are a few customers that want 'the full story' but most of them are interested in getting the images they want in the easiest possible way.

I just got a fresh batch of 7222 and it comes to a gamma of 0.72 with our standard process, it will probably drop with aging. I have a feeling that Kodak may now bring it to market sooner after manufacturing, letting it 'age' less than before. If someone would come in with a major production, we would process this emulsion differently to get a gamma of 0.65 (unless agreed otherwise).

Link to comment
Share on other sites

Dirk, we here want "the full story". I've never heard of films changing drastically as they wait some months to be shot. A standard processing producing gamma 0.65 vs. gamma 0.72 is a drastic change.

 

The aging you describe has nothing to do with the emulsion ripening and after-ripening, which are completed during manufacture. It's some kind of decay which must be reducing sensitivity of the smaller grains more than the larger. Has there ever been a study of shelf-life of B&W motion picture films? (Some spectrographic films have NO shelf life.)

 

Mees & James (1966) says: "...some emulsions show a rapid drop in sensitivity during the early period of storage, without a conconcurrent increase in fog. This behavior has been termed anomalous aging by some authors." While 7222 seems a simple emulsion, performance-wise, we don't know what black magic goes into its formulation and manufacture. The danger comes when Kodak doesn't know.

Edited by Dennis Couzin
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...