Hi John,

Can you mention how it will compare to Workprinter and Cinemate in quality and price? When will you be able to announce it?

Thanks,
David

I have a slide feeder, put in 50 slides and let it do its thing. Insane to even contemplate doing it any other way.
Bob.

I have a slide feeder, put in 50 slides and let it do its thing. Insane to even contemplate doing it any other way.

Me too. Even one box of slides is too many if you have to insert each slide, one at a time.

Can you mention how it will compare to Workprinter and Cinemate in quality and price? When will you be able to announce it?

In answer to the first question, it is identical to the Workprinter quality, since it captures each frame as a discrete entity. There is no difference at all. The Cinemate, of course, captures the film at the same, full-speed 24 fps as my invention, but uses the standard 5-blade shutter technique to reduce the flicker inherent in capturing the film using a 1/60 second shutter. With Cinemate, you of course get some frames of video that contain parts of two frames of film. This increases the blur. Also, even with more blades on the shutter, you still get flicker.

As for announcing, I am going to finish my twenty reels of film, and perfect the software before I describe the actual details. I also want to complete my provisional patent. I will then make the plans available to anyone who wants to build it. This includes Roger Evans, the inventor of Workprinter and Cinemate, whom I have already told about this. He was only mildly interested in our initial email exchange, but I think part of that is because he had considered something similar a long time ago, but had rejected it. The reason for that is that he knows a huge amount about projectors and film, but relatively little about computer software programming. By contrast, I have a relatively small knowledge about movie projectors and movie film, but know quite a bit about software. The keys to my invention are:

1. You absolutely MUST use 1/60 second shutter speed on the video camera, when capturing from a projector that has a shutter. However, once you remove the projector's shutter, using an infinitely fast shutter speed on the video camera means that you can be absolutely 100% guaranteed of getting two complete fields (one complete video frame) of every frame of film. You will also get lots of duplicates and occasional captures of the film while it is moving. These must be removed.

2. The "cadence" of the resulting capture is virtually identical to film that has been telecined using a 3:2 pulldown cadence. When you look at a 29.97 NTSC broadcast of 24 fps film, if you look at it not just frame by frame, but field by field (you can separate fields and look at them as if there were discrete frames, by using the appropriate software), you will see a repeating pattern, where certain fields are duplicates of previous fields. Some very clever people created "inverse telecine" software that is able to recover the original progressive frames by looking at each field of video and detecting where there are duplicates. While this detection can sometimes fail when there is very little movement, it is easy to do when there is movement, and once you have correctly determined the cadence, that exact pattern can be extended both backwards and forwards, and used to eliminate the duplicate fields.

This IVTC software has to be adapted for my purposes because the cadence is not guaranteed to be perfect, since the projector never operates at exactly 24 fps. It drifts over time. Thus, you cannot extrapolate the cadence backwards and forwards in time indefinitely. However, unlike the normal IVTC problem, where it is difficult to detect motion when the subject matter on the film doesn't move, in this case, we get huge, unwanted motion every time the camera captures the film during pulldown. This is dirt-simple to detect, because that field is completely dis-similar to both the previous and subsequent field. In addition, you are absolutely guaranteed that the next two fields will be of the same frame of film.

More than you wanted to know, but the bottom line is that it requires some modifications to existing software. In my prototype, I am just doing the cadence detection by exporting numbers to a text file, putting that into Excel, and then just using IF statements in Excel to find the pattern. I just purchased a compiler and hope to use some of the open source for the IVTC software to modify that so the whole thing can be done without all the cut/paste that is required by the Excel prototype kludge.

Hi fishbelt,

Sorry for the delay in responding - just got back in town this evening.

For the jobs I do, splicing tape makes more sense than film cement.

The combination cleaner/lubricant I use is Filmrenew. It has never damaged any film, even when I soaked an old stiff reel from 1932 for two weeks. As was mentioned in one of the links, Filmrenew can leave some puddling if it does not evaporate before winding onto the reels. An easy way to solve this is to "buff" the film with a soft cloth to be sure it is evenly dry after cleaning.

As was mentioned by Serena (I think) don't over lubricate the film paths. If you use a cleaner/lubricant on the film, that is all that is necessary other than making sure all the film paths are clean and burr free.

A good source for these supplies in the US is www.urbanskifilms.com Larry is willing to sell in small quantities and has always been very responsive on my orders.

RalphM

johnmeyer,
I'm no projector expert but have encountered the film horizontal movement on my machine. Look carefully at the "claw" arm and make sure the two fingers are alined vertically. If the arm is twisted you'll get the side-to-side film jitter you are talking about. A macro photo can be helpful in seeing the twisted arm.