Thomas Aschenbach

Yes I am sure. The two pixel formats the scanner outputs is CRAW in bmdFormat12BitRAWJPEG or bmdFormat12BitRAWGRBG. bmdFormat12BitRAWJPEG is a jpeg compressed bmdFormat12BitRAWGRBG.

The Cintel scanner would be great if it had a better sensor and higher resolution. It has an SDK which allows for writing software that can utilize the scanner and its image data output. The Cintel scanner unit itself analyzes the frame to determine the stabilization translation values but doesn't actually apply it. Each frame the scanner puts out has metadata to tell the host computer how to handle the sensor data including stabilization, and colorimetry data determined by BMD. For our application as an input device for a film analyzer it works well because noise and resolution do not matter. Any scanner passes image data from the sensor or sensors and then either the machine or the host computer corrects that data for the known characteristics of the device. Flat field correction, fixed pattern noise, color masking, response linearization, etc. Some of these are characteristics are unique to the individual machine and some to the manufacturers whole fleet. Flat field correction and fixed pattern noise correction are really the only two characteristics that have both spatial and depth components and are unique to the individual machine. They are corrected far upstream so that correction is baked into what could be considered the corrected "raw" sensor output. From there everything else is values for a math function without a spatial component. This is all similar to the way digital cinema cameras operate. The storage space / bandwidth savings are significant. I would hope this technique is used more in the future by other film scanner manufacturers because the raw sensor data and profiled metadata can be utilized by software whose purpose is color science and image manipulation. This is currently not possible because most bayer sensor scanners and some current mono ones do not have objectively measured standards, if they did their default outputs would all look the same for a given piece of film and they don't. This is most apparent in color films.

I have a Cintel scanner we use for one of our film analyzers. We don't use it for film scanning, just live for timing because it has an API. The debayer is not done in the scanner. It has the capability to do so on the hdmi preview output, but the data that comes over thunderbolt is bayer. The stabilization information is stored as x y offsets in the header of the frame data, same with framing, color masking, etc. Field correction / FPN is about the only thing baked in, the image that the scanner transmits is just the 12 bit raw from the sensor.

Colorlab offers 16mm B&W recording onto 3234 duplicate negative stock. For color 16mm we record to 3242 ip/dn. We also record optical soundtracks and make color or bw prints. Here is an example of a music video recorded to BW 16mm here and then scanned. This is a commercial example.

Dupe negatives & IPs do not have a soundtrack. Soundtracks are recorded on a separate B&W soundtrack negative stock. http://motion.kodak.com/motion/Products/Lab_And_Post_Production/Sound_Films/2378.htm This stock is very high contrast. The soundtrack negative roll has the analog optical track and the digital tracks. The film printers have a separate sound exposure head. For a 35mm print the picture negative roll and soundtrack negative roll are threaded on the printer along with the print stock. The picture, analog optical track, and dolby digital track are exposed separately from each other all in one pass. Even though the analog optical and dolby digital are on the same piece of film they require a different exposure and color balance.

The low cost consumer night vision goggles will have an led for illumination you can turn off or on and are sensitive in the spectral range you are shooting. The cheaper ones I speak of are really just high gain near ir imagers. If they have no ir illumination source, they dont work. The ones the cops/military use are sensitive in much longer and wider range of wavelengths. On the consumer ones the LED built in is about where you are shooting 750 -900nm and will vary by goggle manufacturer. You could turn off the LED and use a r72 filter or something on the goggles to see the spectrum the film will see and you cant. We have some here at the lab. They are in the range you are shooting and I will find their name and part number. When in total darkness you can see some red emission from the built in LED as its spectral output has skirts and so do our eyes. We have to use them here because Kodak stopped making 72 internegative. It is used to print from color reversal/pos to color negative. They replaced 72 with 73 which is the 03 50D emulsion on top of estar base with no rem jet. That means no safe lights(total darkness) in the printer room. Color camera neg is not really sensitive above 690nm, lab intermediate stocks around 740nm. So while printing interneg, the operator wears the goggles to monitor the printer.

Roy, Thanks for your comments. Your film looks great, the overcast day really made it have a wide range of details from shadows to highlights. I actually also shot this test on ST8. It also works well and is about the same exposure index maybe a little faster and a lot more contrast. The problems I have with it is in its stability in camera. The stock is so thin that the pressure plate on some cameras does not make contact and it can have a movement/flutter issues. Depending on what type of camera you use this may or may not be a problem. Also everyone should be very careful when using a polyester stock in camera. It is incredibly strong and if jammed can easily damage a camera. I wanted to show the 34 comparison because I miss the clean look of Plus-X(31). 5234 is a duplicate negative. B&W is different than the color photochemical workflow stocks. In color, interpositve and duplicate negative are the same stock. (42 emulsion orange base). IP/DN 42 has a unity contrast. The gamma is ~1. When it goes Camera Neg -> IP -> DN, the dupe neg has the same contrast as the camera original. With B&W IP and DN are different stocks. Fine grain masters (66 emulsion) are the interpositves. They are low contrast clear base positives. While a lower contrast than B&W print stock (02 emulsion), FG masters adds some contrast relative to the camera negative image. So in B&W, camera neg ->FGM->DN, the dupe neg(34) should have the same contrast curve as the original camera neg. Because the FGM(66) adds contrast, the DN(34) reduces it back down. This is why 34 DN's gamma is lower than 1 and it can work in camera and still be printed/scanned with full tone range.

Hello, When shooting IR there are many thing to watch out for. Since you want to shoot "near ir" (720-820nm) lens focus should not be that much of an issue and neither will be heat radiation. The only focus problem will be that the film you are using is sensitive to visible light and so you are going to have to use a filter. The filtering will screw with focus and you will have a very hard time focusing through the lens. Metering will be quite problem, you will have to just shoot tests. If you can, find a spectral output graph of the lights you have. Some have IR/UV filter glass coatings. Fluorescents & visible LEDs will wave no IR output. For the easiest lighting, either buy or create an LED panel. LEDs with that spectral output are very common. If you use these and limit visible light while shooting you can get away with not using a filter. Then you may be able to focus with visible lights then lower them to a minimum and activate the IR. You can diffuse and bounce. You will have to look at the diffusers transmission graph and see which one will be best. The best thing to do is purchase a cheap set of night vision goggles. Most cheap ones have a single LED illumination (which you can turn off) and operate in the spectrum you are going to shoot. That will let you see the diffusion / bounce.

The prices I gave were off our published price list, which are very close to many other labs' price lists, and was using them to give an idea on a per minute price for 16mm. Matt, if you would like a quote I can email you one. I was hesitant in writing the previous reply and decided to do so without specifically quoting this job. I don't think this thread or forum should be turned into a place for bidding wars, but remain a place for the exchange of information and ideas. Robert, I don't think we have ever personally spoken. If you would ever like to discuss anything about the film/lab industry please feel free to contact me. There are not many of us left and it is important that we work to continue the availability of motion picture film lab services.

Because the demise of Kodak B&W Plus-X emulsion (31), shooting B&W film outside is challenging. 5222/7222 is the only Kodak B&W camera stock. Outside it is has an exposure index of 250. In the test described below the daylight was metered at 8,000 foot-candles. At 24fps this would require a f stop setting of around f64, or a very dark ND filter. Most lenses are sharpest at about 2 stops from open, this is impossible for 22 outside without using an ND15 or ND18 to get f5.6 or below. The contrast of direct sunlight & shadows is a struggle for any photographic medium. 5234 B&W duplicate negative is an intermediate film and is quite slow. 34 has a gamma less than one making it suitable to shoot and print. It is also panchromatic and available in acetate 16mm or 35mm. For the test it was rated at EI 6 for D96 processing. Test Information Film Information 5222 Kodak B&W 35mm DOUBLE-X Negative $0.444 per foot Exposure Index: EI 200(Tungsten) 250(Daylight) 5234 Kodak Panchromatic B&W Duplicate Negative $0.391 per foot Exposure Index: Rated EI 6 for test. Lighting Direct mid day sunlight - Metered @ 8000 fc Camera Setup Mitchell Super 35mm High Speed 4 perf 75mm prime lens 5222: f11 ND9 @ 24fps 1/48sec 5234: f5.6 @24fps 1/48sec Processing Kodak D96 B&W Negative Process Transfer Spirit 2K HD 1080P 23.98 1.78 extraction da vinci 2K+ DVNR2K Settings constant for both transfers. The results of the test were the following: The grain structure and response of 5234 is clearly finer. It is a very smooth image and out performs the 5222 outside. It is availible from Kodak in both 35mm and 16mm. We will do a test of 16mm soon, I am sure the diffrence will be even more dramatic. While this stock would be hard to shoot indoors, outside it is beautiful. If processed D97 is would have an exposure index of about 18 allowing for less direct sunlight. It is also cheaper 5222 $0.444 vs 5234 $0.391 per foot. Please right click and save as to see the test video. Youtube upload didnt show a good comparison.

From a lab perspective always edit in the native resolution. Scaling is complex and can be done with many different algorithms depending on the look and content. It will also be easier on your end.

Not trying to advertise, but I want to give some real numbers for S16mm/16mm digital workflow costs. Lab Costs For processing, leader, clean, Best Light HD 4:4:4 scan to hard drive it is $14.40 per minute. For processing, leader, clean, 2K log DPX scan to hard drive it is $20.80 per minute. Distribution Master Costs DCPs are about $10 per minute + dx115 carrier and Hard drive. Digital Output to 16mm (Pix Neg,Neg OT, 1st Answer print) about $150 per minute.

42 is an intermediate stock with unity contrast( gamma of ~1). unless your going from a negative or ip the resulting prints would have a very high contrast. You will also need an orange filter to simulate orange base of color negative. Current Kodak interneg 73 is just the 03 emulsion on top of estar base with no remjet. If you do not require polyester base then use 03 camera film. They do sell 73 in acetate with remjet but it is more expensive than 03 and the only difference is the size of roll you can buy.

For the transfer it depends on whether you intend to shoot super 16 or regular. If super 16 then go with the HD transfer. If regular 16mm go 2K. 2K does not add much to super 16mm even if the file is 1.66 native to match the film, all your getting is 2048 x 1234 rather than 1920 x 1080. Since you are shooting B&W you do not need RGB files as 66% of the data will be redundant. Uncompressed 4:2:2 files will allow you the same 10 bit luminance info with less wasted data and easier data transfer requirements for computer playback. Negative cutters for 16mm A&B rolls are hard to find and pretty expensive these days. You would need keycode to avoid problems and headaches. As said above, no fine grain and dupe neg would need to be made as you can print directly from your A&B rolls. I am all in favor of finishing on film, especially with black & white. DI is a good option. 2K transfer of the negatives allows you to edit and do titling and other effects. You can then have a 16mm film out of the final show. If you do this make sure the negatives and prints are made on B&W stock. We have used this workflow with our clients and the results are great. We have some clients who still have A&Bs cut from 16mm, but most will still require some digital film recording for title composites and credits.

For my previous post when I wrote Kodak color film stocks have basically the same resolving power when looking at it in a digital resolution(pixel count) perspective, I should have stated film of the same size. The emulsion resolving power is the same regardless of format, needed digital resolution is based on the size of the image area used on the film.