A question on experimental formats

[wado1942]

With all this talk about trying to get more out of your film, has anybody tried non-perforated film? Think about this, using a capstan/pinch roller movement, you don't need perforations and the movement is more stable. You could use maybe 7mm of non-perf'ed 8mm film and program it to move in 4mm increments to get a 16:9 aspect ratio. Likewise, you could do the same with 16mm and get maybe a 14mm x 8mm frame. Better quality, less waste.

[Will2]

I think the issue would be stability. It's those perfs that allow cameras to maintain good registration. For instance there is a really big difference between an Arri 2C and an Arri III that has the registration pin.

[aj]

If it could move fast enough and have enough play between frames it could work. Nowadays all new images on film are converted to digital so if you expose a marker in the frame border the digitizing software can automatically recognize this and line up all frames...
Works on 70mm too and there is still 70mm non-perf production I believe/hope

[Nicholas Kovats]

This exists on the electronic projection side of things, e.g. Muller HM73 scanner which utilizes a laser aimed through the film perforations as a registration method.

I have sample UP8 scans arriving from Germany and I am curious if Frank provided overscans that the perforations are visible. If so I will create side by side comparisons of 3 different scanning systems for analysis.

I have sent a query out to a contact regarding an all electronic motion picture camera transport/drive. There may be a patent lurking somewhere.

Happy New Year everyone! Keep film alive. Shoot!

[Nicholas Kovats]

You may wish to check out this 1998 Nikon patent regarding using a photosensor to track perf alignment, i.e.

"SUMMARY OF THE INVENTION:

It is, therefore, an object of the present invention to assemble efficiently a device having a photosensor for detecting perforations in film."

http://www.google.com/patents?id=6aggAA ... &q&f=false

[Nicholas Kovats]

Here is the real meat from a Canadian engineering angle by the IMAX corporation as per their 2012 patent. It appears this has been perfected on the projection side, i.e.

http://www.google.com/patents?id=AcwLAg ... &q&f=false

[wado1942]

This is great. I know film tracking exists on the projection side of things, like in the 1999 system Maxivision, where a grid is printed just below the frame and a CCD detects this to adjust framing before the shutter opens. You'd just need a camera that uses a capstan/pinch roller system to transport, say non-perfed Super-8 carts and only leave a small gap between images to burn in some sort of registration point. Imagine a 7mm x 3.8mm image on Super-8 film with near-perfect registration. The actual transport would be the difficult part by a long shot, since it would need an intermittent movement.

[Nicholas Kovats]

I found a United States Patent 3697017 called FILM TRANSPORT DECK UNIT 1972 which details a capstan and pinch roller system for non-perforated film. Check out the diagrams. It does not appear to utilize a registration system.

http://books.google.co.uk/patents/US3697017.pdf

[Nicholas Kovats]

I found another more detailed patent from 1980 called "Cine Camera", i.e.

http://www.google.com/patents/US4198133

Realistically I suspect it would be impossible to convince other's to invest in the tooling and/or cnc machining. You have not presented a valid technical argument regarding it's potential superiority over "traditional" pin or non pin registered pull down intermittent movements. I will put this down as a thought experiment. Perhaps when 3D printing reaches a critical mass we can once again revisit expired patents and potentially model alternative film transports.

[wado1942]

Thanks for the PDFs, I'll take a look at those when I get time. The argument for a perfless film is more efficient use of said film. As is, Super-8 is about 1/2 the cost of 16mm but 1/3 the area and that's hardly a fair trade. This would simply do away with the wasted area and make it 1/2 the area of 16mm, so it improves clarity, registration quality & due to the adaptive positioning after capture) and proportional, cost wise.

[Will2]

As is, Super-8 is about 1/2 the cost of 16mm but 1/3 the area and that's hardly a fair trade.

Look at Regular 8mm! Those perfs are giant compared to Super 8.

If you are talking about 8mm a more realistic approach might be a variant on the Max 8 or Super Duper 8 and the wider gate idea. Obviously if you get into 8mm without perfs the tolerances would have to be extremely precise; more than it would be for 16mm and you'd of course have to design and manufacture an entirely new camera and possibly magazine/cartridge system.

Also keep in mind processing for the sprocketless film. Not sure if that's an issue or not.

Find a good machinist and give it a try. If you can get a prototype working there's always Kickstarter!

[wado1942]

I didn't even think about processing. So you'd be stuck with home processing. For the record, I'm not really suggesting anybody actually build this, I'm just exploring the possibilities. My first thought would be to offset the pull-down of a Super-8 camera by 1/2 perf and shoot between them, but there's the distinct possibility of getting scratches in the perf area and you would be throwing away more frame area with a 7mm x 3mm frame and a 2.35:1 aspect ratio.

[Will2]

Since we're doing the thought experiment, perhaps going "VistaVision" on Super 8 might be fun.

35mm VistaVision:


In other words, rotate the image 90 degrees and make it as wide as you like...say 16:9 or even CinemaScope. That would mean giving two to three frames away for each frame of this format so running time would be more like 1:30 but it would be fun and the quality would be closer to standard 16mm cropped.

[wado1942]

I thought about that, actually and even wrote an article about it years ago, though you can't access it through my index anymore http://www.gcmstudio.com/ultra8/ultra8.html
I removed the link because Pan-8, where you take a 16mm camera and mask off one side and run it sideways already does the same thing but without any major mods.