Dirk DeJonghe

I think most customers would be happy with the prepaid stock + processing + workprint deal we have: any Kodak Vision 16mm negative stock (50D up to 500T) in 100 ft or 400 ft reels. You buy the prepaid stock from us, it comes with a sticker and you get a workprint or HD Quicktime. Send a PM for prices, they are very close to the Reversal stock plus processing only.

I had a S16 production that used 50D stock to shoot sunlit exteriors in Greece. The cameraman had to do some reshoots and used his own 35mm camera. It was an older Cameflex with vintage lenses. After seeing the results he regretted his decision.

Keep in mind the difference between camera aperture and projector aperture. During shooting he negative will be carefully framed to the projector aperture, if you show more than this, you may run into dollytracks, microphones, etc. The director and camera operator are not going to like it if you show unintended extra image. Unless it is your own film, it is best to stick to the official formats.

Here is a picture of the 35mm bi-directional panel printer, image head on the left, sound head on the right.

I will get a picture on Monday;

David, that is correct. I would like to add that in case of a direct blow-up/reduction on an optical printer, the sound negative is exposed after the image on a separate contact printer, using only the soundhead. Agfa ST8D was a very good stock, but no longer available and no longer manufactured. Kodak and Orwo have comparable stocks.

The printing machine has two printing heads, one for picture and one for sound. Two negatives are threaded to make one positive print.

It used to be that the soundtrack was copied to the duplicate negative, but only for B&W and certainly not in the last fifty years or so. I see this configuration only on very old archive material in B&W from the 1950s or before. Currently all soundtracks are printed from a separate sound negative that can be used with the original negative (if suitable for printing) or from the duplicate negative. Friends in large labs told me they could strike about 2500 prints from one set of duplicate negatives. A separate sound printer head is used because the exposure of the soundtrack is quite different from the picture. Also, light changes in the picture would do no good in the soundtrack.

With PFClean, you take a sample of previous and next frame, just large enough to cover the defect. The program will motion interpolate if possible, in case of very rapid movement between frames, it can be set to only take local pixels from current frame. You can also make it interpolate for flicker if present; we always dustbust before colorgrading. There also is an automatic mode that works well if the shot is fairly static. We prefer the manual mode. I think you can get a 30 days trial version. It is a very deep program and will require many years of practice to master most of it. On the other hand, the dustbusting is fairly straightforward and you can see the result immediately. You can render to a different directory so you don't overwrite your original files if you are not sure of your skills. I think photoshop also has a similar but less powerful tool, if only for a few dozen frames, it might work.

Your best bet is to do digital dustbusting. We use PFClean very effectively but there are other means. Wetgate will not remove embedded dirt as Richard indicated.

Yes we still do AB-roll printing, both contact and optical reduction/blowup or recentering from S16 to Std16 with optical soundtrack which we make in-house now. There is still a reasonable amount of Kodak black leader available in good hands, we can make lightstruck black leader if that runs out.

If there is a requirement, we could make a 1.78 mask on 4Perf to be used when optical printing 3Perf full frame to 4 Perf. There is little demand for this kind of masking, we mostly use 1.66 and 1.85.

If you do direct blow-ups or reductions, it is essential to print a black mask because the unexposed area outside of the negative frame would be transparent. We would take our standard 1.66 or 1.85 mask and then take the frame leader for this particular production and adjust the optical printer until the frame leader arrows fit exactly in the mask. If going via IP/DN this would be done at the IP stage.

We can time the film during optical reduction/blow up and make fades as well; our optical printer has the B&H light valves and fader. We did many hundreds of S16 productions where only a few direct blow-ups were made purely for festivals. First a trial print was made on S16 contact positive, then viewed and color correction fine-tuned, then in many cases only one blow-up print was made. We did the same for 3Perf, but no contact print here, only reduction to 4Perf direct. A properly made direct blow-up from S16, with good lenses and good photography still is mind-blowing quality even today.

You would also have to provide optical sound recorders for 3 perf. I don't see that happening.

Very few 3Perf projectors around.

You need both an optical reduction plus a 3Perf to 4 perf change. This is done by having a 3Perf movement in the projector and a 4Perf movement in the camera of the optical printer. Once this is done, you still need to put a mask by exposing a mask film on a contact printer. It has been many years since we did 3 perf work on the optical printer, but all the parts are still there. What is more, when doing direct optical reductions from B&W negative to B&W positive, you may run into halation problems, depending on image content (black halation around backlit hair for example). This is because of lack of anti-halation layer in the B&W positive, this problem is much reduced when going via Interpositive/Duplicate negative.

We only use PFClean's automatic mode on shots that have almost no motion, on everything else manual dustbusting is done; if you are very good, you can do about 10 minutes in a day if the film is not too dirty. So, even just counting labor, we are talking about several hundreds per day.

We use PFClean with great success. See http://www.thepixelfarm.co.uk/. You can also send us the footage to process. Grain reduction is quite separate from manual or automatic dustbusting. It is a great program but with a steep learning curve.

We charge processing and cleaning per meter and telecine and scanning per minute. Some labs have a price for processing and scanning per foot, this means you get more minutes for the same footage when using 3Perf instead of 4Perf. Maybe this leads to the confusion?

The middle one is Ektachrome reversal, it has edgenumbers. I don't think Kodachrome ever had edgenumbers, maybe the Kodachrome Commercial but that was before my time.

It would be better for everybody if Apple would make the ProRes codec 'open' like Avid did with DNxHD.

On my scanner I can do 8, 10, 12 and 16 bit per channel DPX, but most applications I have won't recognize 12bit version. It is indeed open format, SMPTE approved, which means you are not vendor-locked into one particular workflow. Try combining Windows or Linux with a full Quicktime workflow.

There is nothing to prevent you from working in DPX 16 bit. If you need to change one single shot in the middle of a 2 hour program, DPX suddenly becomes much more easy to work with than any Quicktme codec. Just rerender the shot and you are done. That said, you need a fairly heavy RAID to playback DPX in real time without proxies.

After many years of remaining dormant, flashing is now used in our lab quite often these days when making proper internegatives from positives or reversal originals. We mostly use 200T stock preflashed and Pull processed to a particular gamma to get really good internegatives. This method is also used with the current crop of internegative stock from Kodak which is really 50D in longer lengths and on polyester base.