if you can have your footage transferred frame by frame and dumped to hard drvies/removable media(dvd, cd, whatever), then that would provide you with rgb images, that can then be converted to quicktime movie format of your choice for nle editing. but the point being, you have highest res possible rgb images as your archive, meaning if you wanted to ... you could print back to film using those images as they are(just edited with cuts),or with added compositing/sFX.
dpx/cineon files work well ... as does jpeg2000 ... both mac and pc computers have utilities to read these file formats. if at all possible, AVOID video, espeically today, as one can now transfer their frames of film at hd, 2k, 3k, 4k, and even 8k if need be(4k & 8k for 35mm, very expensive, and slooow).
if price is a consideration, find someone with a WORKPRINTER(moviestuff.tv i think) or a jk optical printer, either of them having a dSLR camera attached, digitizing frame by frame. this should not cost too much.
i myself have prototyped my own telecine device, and am using firewire and cameralink solutions, with various camera heads ...10bit and 12bit, up to 4k. although the process is slow(1 frame every 3 seconds), no telecine in the world can compare in terms of quality.
my co-developer and i have been working on osx software that manages the large numbers of sequencial files, allows color correction, printing out to applicable format, be it video or data, for distribution or archive, and more. currently, there is no software on the market that allows these processes to occur in a simplied manner. we are creating special filters for negative, posative, and reversal filmstocks covering kodak, fuji and ilford product lines(current stocks for now ... discontinued stocks will follow when we have completed more important aspects of the software).
also, our software will work/identify almost 100 differant types of cameras from usb to firewire(both versions of each), cameralink, ethernet, and i/o cards like decklink and aja, with appliable cameras attached to those cards. this includes dSLR cameras, and does so in the sense that when a dSLR camera is plugged into the computer, it will be recognized as a video device as well as a still camera device(those of you who have played around with dSLR cameras doing stop motion animation will understand why this is benefical).
the software is currently called F2HD(film to high def), and has the following modules:
LOCALIZE
COLORIZE
ORGANIZE
STABILIZE
ARCHIVE
each module having numerous functions and processes, both automatted and manual for full control. so lets say you are digitizing Kodak 50D super 16mm stock using a kodak DCS Pro/C DSLR camera, the software will take the physical measurements of both the film and the sensor, and provide the ideal capture resolution based on benchmarking tests we have performed. lets say if you are using an i300 firewire camera at 1024x768 with 35mm film loaded in your transport device/telecine/workprinter/jk printer/our printer/whatever printer, our software will FIRST tell you/show you what type of image you'll end up with(in this case, the sensor/res is too small, and therefore that digitized 35mm film will only be suitable for 720p video output, for a very simplified example). additionally, various camera heads/sensors(ccd/cmos/foveon) havew differant pixel sizes, ranging from 2-3 up to 10-12. the larger the sensor pixal size, the BETTER the image quality. in many consumer cameras/sensors, the sensor is very small, but also has very small pixel areas, thus making a higher res image from a smaller sensor ... but this does not mean a better image. so knowing the thresholds will help select the right sensor for capturing your film stock, something our software will provide.
because cameras/camera heads/sensors are readily available from many suppliers/manufacturers, it makes it easy for someone to purchase a camera that is right for them and their use. but film transport and illumination devices are RARE, so unless you are using a contact closure mouse type device like the workprinter, or a modified optical printer, the only other options are high end, very expensive telecine machines like rank, spirit, arri, etc., either being a telecine machine or a film scanning machine. this is why we are also building a transport/illumination device capable of gating all formats of film(with applicable gate, pin registered), simply swap the gate, and only the gate, and voila ... you can digitize whatever format you so desire. right now, the device is connected to the host computer via USB, and has on-board memory for programming additional functions and controls for the on-board componants, which include steppers motors and servos and LEDs. steppers/servos are used for film transport, opal/diffusion linear adjustment, glass/resin filter linear adjustment, and LED lighting linear adjustment. on the other side of the gate, the taking lens can be adjusting up, down, left, and right, as can the camera move up, down, left, and right(camera and lens are connected using a bellows system, and in the case of 8mm and 35mm, extension tubes are employed between the bellows and the taking/printing lens). the lens we use is custom made, and provides exceptional resolving power, far more then is needed for taking illuminated film to digital sensors in the 4k range.
but i've gone off topic ... sorry, i could not help myself :)
regarding taking your film to the desktop ... always remember that digital/data files in rgb(10 or 12bit) is going to provide you with future proofing abilities and archive ability, something VIDEO cannot provide, as video formats come and go, and are YUV, a lessor quality when dealign with them on the desktop, which is what we are all doing now. having RGB data files means one need only convert them, or downsize them to an applicable video format/codec, somethign all computers are capable of doing, using one software application or another.
eric
