Everything You Wanted To Know About Film Recorders by Andy Cross.

            What’s a film recorder ? Oops maybe I should back up a bit. Film recorders were a stop gap in the transition from film photography to digital capture. In the early days of digital capture the resolution of these devices was pretty lousy. They could produce files suitable for output to a printing press to about A3 in size. Some not even that. A large amount of film was still being shot, much of it colour transparency film, which was in turn scanned on higher resolution drum scanners.

            These files were then given a photoshop treatment and made press ready. The art directors were often shown a PDF before it went to press but colour management was also in its infancy and the result from screen to paper would often show up issues. A better alternative was to proof the files on a colour transparency film that had all the corrections. Then the art director and client would sign off on that. The transition from film to printing plate was something that everyone in the daisy chain was familiar with.

            Data projectors of the quality we have today didn’t exist and if someone wanted to present an audio visual you could present graphics on OTF film. Images could still be displayed by projecting film. The labs were still using optical systems like enlargers to print for the wedding portrait industry because digital capture and high end inkjet printing was still being developed. Often corrections would be made by photographers by scanning C41 colour negatives then having a new negative written. This was then printed on RA4 paper the conventional way.

            Although the motion picture industry was making use of digital most cinemas were still projecting film. As a result the finished film still needed to be transferred to film. 

            All these variables in output required the use of a film recorder where the file is output to a device that exposes a conventional photographic film to points of light or circles of confusion. This film is then processed in the darkroom as if it were exposed in a camera. The end result is a film image that can be printed using an enlarger onto conventional B&W or colour materials.

            These days with the resolution capabilities of digital capture it’s possible to integrate digital capture with film based photography. You can begin with a film based image, scan it high res, make any corrections required and output the file to the film recorder and you need not dodge, burn in or perform any other manipulations again.

            Alternatively begin with an RGB digitally captured file, make a B&W conversion, output to a B&W film and print conventionally in your darkroom. Even making B&W transparencies is possible by inverting the file before output. Creating surreal compositions which although possible in the analogue darkroom became much easier to create for people who don’t have those skills. Composites for commercial uses became much more affordable.

            So with film recorders explained I should elaborate on some of the tech bits regarding these devices and the differences between them. Just as oils ain’t oils film recorders ain’t film recorders. They can be broken down into several categories. The first are the CRT types. 

These machines were essentially a flat field corrected Cathode Ray Tube that had a camera pointed at it with a special flat field corrected apo lens attached to it. The resolution capabilities greatly depended on the resolution limits of the CRT. They were also sometimes called flying spot film recorders. Most were limited to 35mm format although some would accommodate medium format. The films from them could be enlarged a moderate amount but were really intended for producing images for audio visuals. Cost reductions were possible because they could use 35mm sprocketed film. 

Another type were classified as digital enlargers. They were intended to write digital files to either B&W gelatin silver or RA-4 colour paper. Units like the DeVere D504, Durst Lamda and Light Jet printers would fall into this category. They were never intended to write to film and the results I have seen where they were fed some film had artefacts through them particularly in areas where smooth continuous tones existed. The results when printing on paper are very good but are still not capable of resolving very high frequency information the same as printing from a negative can.

Similar machines like the ZBE Chromira printer and Durst Rhino printer that expose the paper to RGB LED arrays that move over the paper on an air bearing produce similar results. 

The next category are the LVT film recorders that use what’s commonly known as light valves. Light valves have a long history that I won’t elaborate on but the first film recorder built using this technology was made for Kodak. These machines can write to virtually any photo sensitive material all you need is the appropriate profile. The resolution capabilities are almost as good a first order result from a camera. 

Provided the file size is sufficiently large enough, these machines are capable of writing to roll material. Images a meter wide and as long as you like are possible. Writing to backlit film for display purposes needs to be written on this type of recorder.

Finally we get to the laser film recorders. Machines like the Symbolic Sciences and Fire 1000 fall into this category. They in turn can be broken down into two categories. The dye pulsed and gas laser types. The later can also be separated into two categories. The dye pulsed units make use of dye pulsed lasers. The exciter material inside the laser tube are dyes. There are three such tubes in these film recorders, one for each of the three primary colours being red green and blue.

These tubes can be manufactured to work in any colour space but wear out a lot more quickly than the gas lasers. The recorders that use gas lasers usually have a helium tube for the red, neon for the green and krypton for the blue wavelengths. The purchase price is higher than the other types but can write the files to film much faster. Another type of laser recorder uses a single white light laser using xenon gas.

These units rotate a filter wheel in front of the laser as it writes each of the colours. Just white light is used to write B&W film. This is my preferred type as it allows the user to change the colour space in which the recorder operates by changing to a different filter set. Although this isn’t recommended by the manufacturer it does work. With the other units B&W film has to be written in either RGB mode or one of the lasers has to be selected. The red laser is most often chosen as it is the cheapest of the three to replace and red light scatters through an emulsion the least.

The resolution capabilities of film recorders is not expressed in the usual way. We are familiar with inkjet printers output expressed in DPI or dots per inch. Scanners, contrary to popular belief, do not scan in DPI or PPI pixels per inch. There are no dots or pixels in a scanned image. Scanners scan in SPI or samples per inch. They also generally over sample in that there are many more samples per inch than there are pixels being displayed on the monitor. The only place where pixels can be seen is on a monitor.

Just to complicate things a little more a film recorders resolution capabilities are expressed as an addressibility factor in K. The magic number here being 1,024 because there are 1,024 circles of confusion in 1K. So a film recorder with an 8K capability can write 8,192 circles of confusion per inch to a film along the longest axis of the film. A 10K recorder 10,240 and so on. The highest resolution capability I am aware of is some CGI equipment with a 12K addressibility factor. My film recorders have an 8K addressibility and I have rarely needed more to enlarge 6cmx7cm or 10cmx12cm formats to make 50cmx60cm prints. 

As you can imagine the file sizes you need to feed these machines are by most printing standards quite large. The software driving many film recorders are raster imaging processors or RIP’s. I don’t know of any RIP’s for film recorders that will read interpolated information. So starting with a small file and just interpolating to increase the file size doesn’t work. The software will often put a splash on the screen telling the operator the file size is insufficient for the resolution selected.

 These large file sizes were once an issue when HDD space was small compared with the capacity of the drives we have today. But more importantly the cost of RAM was another limiting factor. Most bitmapped image editing applications are RAM hungry even today. Building a computer system that can deal with all this no longer requires a significant cash outlay.

I have written many files to film for my clients and sometimes I have had to decline the job because of the way the file has been manipulated. The most common fault is the manner in which the selections have been made before the alterations were carried out. You simply can’t use the same approach for output to a high resolution film recorder that you use for output to an inkjet printer or data projector.

Selections usually have to be made via channel operations in order to maintain a smooth continuous transition between the selected areas and those which have not. Alterations in exposure, colour balance, tonal distribution and contrast have to be carried out proportionally rather than globally. Carrying out the work without using these techniques almost always results in what is known as artefacts in the final film. 

With only one company still making film recorders these days ( Laser Graphics ) if a darkroom worker wants to purchase one there are some pitfalls they should be aware of. The first is the type of plug the film recorder connects to the computer with. The earlier types used SCSI – 1 and SCSI – 2. I would avoid this type as getting custom SCSI boards made that will plug into modern motherboards is very expensive. The firmware and software that drive them also must be compatible with the OS. Although you can get SCSI to USB adapters I have heard that process is fraught with problems.

The recorders that have a USB plug on them are a far more viable option. The machine needs to come with all supporting software including the firmware. There aren’t any more updates for the drivers or firmware anymore. Many of them won’t run on current operating systems. So the computer will often need to be dual boot so the latest OS that will run the film recorder software can be chosen. Maintaining old computer systems that did run them is another option although not a practical one for obvious reasons.

The film look up tables are another issue. Sometimes called CLUT’s short for colour look up tables. These tables are selected by pointing the film recorder at them before writing a file. It tells the recorder what type of film you have loaded into it. Many look up tables won’t include the film type being used. 

An application that can write custom look up tables in the file extensions suitable for that recorder can be used to work around this issue. These applications will allow the user to write a table for virtually any emulsion now or into the future but they do add to the cost of setting everything up.

Another addition that I would consider necessary is a good transmission densitometer. These machines pass a known quantity of light through a film sample and measure the amount that emerges on the other side. They are essentially vacuum tube volt meters that readout in log units of opacity instead of voltage. Determining colour shifts when calibrating the system for colour films will require a tri-colour densitometer.

In order to use these machines some knowledge of sensitometry is required but once a person has that fine tuning a film recorder becomes far easier to obtain the results required. Bearing in mind that the film is being processed in a conventional darkroom. Sometimes you need to bring the darkroom into line with the film recorder. Other times the film profile needs a tweak.  Those are some of the pitfalls to look out for when using these older types of technology. 

It is difficult to illustrate an article like this with comparisons between images printed from in camera first order results, scanned originals and those printed and scanned from a film recorder especially when they are all being displayed on unknown monitors many of which probably won’t be calibrated. 

So I have just limited the images to those that were either captured on a medium format capture back or scanned from films. Then those files were output on the film recorder. Like all technologies they have their limitations. Can the results from film recorders match those from an in camera original ? No and never will. But when a viewer has to start guessing which results were  produced in the recorder compared to those in camera it means the operator has worked within the limitations of the technology. 

Some photographers wouldn’t ever consider integrating this type of technology into their workflow. Some will embrace it and there are those like myself who will use it when it makes sense to do so. When it comes to making prints using colour assembly processes like  dye transfer ( See articles So What is Dye Transfer and Back to the Future ) it saves me a full day’s work making the separations this way and reduces the likelihood I will make an error therefore saving materials as well. 

I hope this article has enlightened people to another type of imaging technology not widely known about and answered some of the questions others may have had about it.