16mm pc film scanner used as a telecine

[Jonathan Bryant]

Is there any such computer film scanner that can feed in the 16mm film and capture each frame independently? Virtualy a cheap telecine.

[David Mullen ASC]

For 24 fps material, that's 14,400 individual frames you'd have to scan just for ten minutes of footage.

 

Even if you only took ten seconds to scan each frame individually by hand, it would take 40 hours -- basically two days straight -- to scan that ten minutes of footage.

[Filip Plesha]

Is there any such computer film scanner that can feed in the 16mm film and capture each frame independently? Virtualy a cheap telecine.

<{POST_SNAPBACK}>

 

If you use a flatbed you can get about 15 frames in one scan, and cut them in photoshop. Also a very long job.

 

But a flatbed will give you troubles with sharpness because 16mm needs so much enlargement, even for SD.

[Phil Rhodes]

Hi,

 

> basically two days straight -- to scan that ten minutes of footage.

 

Sounds OK to me!

 

Even if it was a supervised process, which I'd hope like hell it wouldn't be, that's still not a bad rate of pay.

 

Phil

[Jonathan Bryant]

For 24 fps material, that's 14,400 individual frames you'd have to scan just for ten minutes of footage. 

 

Even if you only took ten seconds to scan each frame individually by hand, it would take 40 hours -- basically two days straight -- to scan that ten minutes of footage.

<{POST_SNAPBACK}>

I thought there were scanners that feed the film in by themselves. I have seen scanners that scan 16mm film so you would think someone would make some software to turn it into a slow yet cheap telecine.

 

Here are the only inexpensive options I could find.

http://www.moviestuff.tv/sniper16_pro.html

http://truetex.com/telecine.htm

 

Time doesn't matter so much to me as the quality of the transfer and the low cost of ownership. I would use this for mostly 30 sec ads with anywhere from a 10:1 to 30:1 shooting ratio. I could set up a machine for this and run it over night. This would be great because I could shoot projects on film that would have only been able to budget video. Though I am sure a telecine on a Rank or Phillips would be better not to mention the experience of the colorist.

[David Mullen ASC]

I thought there were scanners that feed the film in by themselves.

<{POST_SNAPBACK}>

 

That's what a film scanner is for motion picture work, scanners like the Arriscanner, Northlight, Imagica-XE, etc.

 

The problem with adapting a consumer still film scanner is speed (of scanning and writing the data) and registration. I'm sure one could be built though but I'm not sure of the quality. But intermittent drives to pull a frame, stop it & register it, then release it and advance to the next frame, are not necessarily easy to toss together. I'm sure someone could manage it though -- it's almost like you'd want to take a Filmo or Mitchell set-up for stop-motion work and somehow sandwich a CCD on one side of the frame and a light source on the other in the gate...

 

Do still film scanners pull the film through to each frame on their own? Then the other question becomes speed again. How many seconds does it take per scan?

 

Something like what you want may show up on the market but it won't be as cheap as you want because people who want to do their own 16mm scanning is a very small niche market compared to people who want to scan 35mm slides and negatives at home, so no device is going to come down in price due to volume.

 

So this falls under the category of: if it were easy, everyone would be doing it already.

[Nate Downes]

The Moviestuff sniper is a good home-telecine system, I've seen the results and it is definately worth the $$$. Almost as much as a good Nikon film scanner setup, and much faster. But not as much resolution, granted.

[Phil Rhodes]

Hi,

 

> it's almost like you'd want to take a Filmo or Mitchell set-up for stop-motion work and

> somehow sandwich a CCD on one side of the frame and a light source on the other in

> the gate...

 

S'what I'd do. You could do it with even a pretty cheap digital stills camera.

 

> Do still film scanners pull the film through to each frame on their own?

 

Some do. Canon and Minolta make scanners targeted at the high street photo labs which have automatic advancement and take long roll attachments; this could be something of an off the shelf solution costing a few thousand rather than a few million. I doubt the registration is anything like good enough for motion picture work and you would be reliant on post-scan stabilisation (as if Spirit isn't!).

 

> Then the other question becomes speed again. How many seconds does it take per

> scan?

 

A few, which is one reason I wouldn't do it that way - the other reason is registration. Far better to marry a decent DSLR (Even something as fairly-basic as an EOS-350 would be OK, although a Nikon D70 or Canon EOS-1DS would probably have better dynamic range) to a camera movement. It could probably be done with off the shelf macro lenses and a small amount of machining or just some X/Y adjustable tables on a solid benchtop. Most of these cameras will operate tethered and it'd be an afternoon's work to knock up some computer scripting to automate the process.

 

I'd also be tempted to add an RGB lightsource, which could also be easily computer controlled, to allow initial setup for the negative mask and possibly intensity control to allow two-pass, high-dynamic-range scanning.

 

An EOS-350D will shoot 3K frames at about 1fps sustained, or even slightly faster, in a situation like this.

 

Phil

[Stephen Williams]

I doubt the registration is anything like good enough for motion picture work and you would be reliant on post-scan stabilisation (as if Spirit isn't!).

 

Phil

<{POST_SNAPBACK}>

 

Hi,

 

You are right about registration on a Spirit! Sometimes its almost perfect and other times its not. I often do a double exposure grid over a colour chart on the first roll! On a few occasions the camera registration (a Mitchell) have been slagged off by TK operators!

 

Stephen

[Phil Rhodes]

Hi,

 

I didn't mean to impugn the Spirit's stability; what I meant was that it uses post-scan stabilisation in the vertical axis - which it does.

 

Phil

[Michael Morlan]

I would add that just about every consumer level film scanner is really soft - no matter what resolution specs they are touting. I discovered this when comparing my CanoScan 2710 against drum scans of still 35mm frames. Canon declares a native resolution of 2720 pixels per inch. But the resulting image is much softer.

 

You simply aren't going to get the detail that a high-end scanner can muster.

 

That said, I have a CanoScan 2710 I'm selling real cheap... :D

Edited July 25, 2005 by Michael Morlan

[Richard Mills]

First post so please be gentle :)

 

I would love to remove tape from of the equation.

 

I can't understand why a desktop scanner has not been developed for 16/s16. Even super 8.

With technologies available today, that some scanner company hasn't come up with a scanner that can do this is unbelievable to me. Or even a desent adaptor for that matter so I don't have to use strips.

 

Something that is the size of a Movieola, with RS422 control, to only scan selects, that will give you a TIFF/JPEG file that can be converted with QuickTime for editing in a NLE. Granted it will be slow, but it couldn't be any slower than waiting for your transfer to come back from the transfer house. All that would be needed is a ton of drive space, which is relatively cheap. Then output to what ever you want.

 

With all the software that is available, ie After Effects, Photoshop, Final Touch etc. You would have more control of the image and achieve the look that you are after.

 

Now I'm not saying that you would use this for a 35mm feature, but something for the middle range for someone that doing spots, music videos, or industrials that want to shoot 16mm and not have to deal with tape. Tape, to me seems to be the weak link in the whole transfer chain. Even HD although close but no cigar as far as image quality goes.

 

Rant over.

 

Thanks for letting me blow off steam

Richard Mills

Cinematographer, Photographer, Editor

Phoenix, AZ

[Jonathan Bryant]

My point exactly. This wouldn't be like reinventing the wheel, it would only be adapting a existing technology. The transfer is the most expensive part of the film process if you intend to goto video or hd. Kodak or Fuji could really compete against 24p and HD solutions more on a cost basis if they would come to market with something like this. I don't think it would hurt the telecine houses too much but would allow people who couldn't originally afford film to use film. Wouldn't that be good for people who make the film though? Supply and demand.

[David Mullen ASC]

Let's say someone put one out and it cost $20,000 -- how many of you would buy it?

 

And if they only sold it for $1000, how would they make a profit considering the low number of people who would buy one compared to people who buy one for still photography?

 

If one could be made cheaply and sold at a profit and give halfway decent results... then why hasn't it been done yet? Could it be that the market is simply too small for a company like Minolta or Nikon, etc.?

[Bon Sawyer]

Richard,

 

Are you familiar with MovieStuff's telecine machines? I think they're about as close to what you're describing as we're likely to see.

 

-Bon

[Nate Downes]

A few, which is one reason I wouldn't do it that way - the other reason is registration. Far better to marry a decent DSLR (Even something as fairly-basic as an EOS-350 would be OK, although a Nikon D70 or Canon EOS-1DS would probably have better dynamic range) to a camera movement. It could probably be done with off the shelf macro lenses and a small amount of machining or just some X/Y adjustable tables on a solid benchtop. Most of these cameras will operate tethered and it'd be an afternoon's work to knock up some computer scripting to automate the process.

 

I'd also be tempted to add an RGB lightsource, which could also be easily computer controlled, to allow initial setup for the negative mask and possibly intensity control to allow two-pass, high-dynamic-range scanning.

 

An EOS-350D will shoot 3K frames at about 1fps sustained, or even slightly faster, in a situation like this.

 

Phil

<{POST_SNAPBACK}>

Actually, this has been discussed before. THe main issue is that DSLR shutters are not engineered for the longevity you would need. The EOS-1DS has a shutter that is considered the best in the business, and it will fail after 150,000 shots, barely enough to record a single short. Replacing an $8000 camera per 30 minutes of final footage is hardly economical.

 

You would have to design a whole new camera, with a more ruggid shutter design, for this.

[Stephen Williams]

Actually, this has been discussed before.  THe main issue is that DSLR shutters are not engineered for the longevity you would need.  The EOS-1DS has a shutter that is considered the best in the business, and it will fail after 150,000 shots, barely enough to record a single short.  Replacing an $8000 camera per 30 minutes of final footage is hardly economical.

 

You would have to design a whole new camera, with a more ruggid shutter design, for this.

<{POST_SNAPBACK}>

 

Hi,

 

Well get an extended waranty and you could have a new camera every month for a couple of years!

 

Stephen

[Richard Mills]

Richard,

 

Are you familiar with MovieStuff's telecine machines? I think they're about as close to what you're describing as we're likely to see.

 

-Bon

<{POST_SNAPBACK}>

 

 

Yeah I have looked into their line of machines. Interesting concept but, they are telecines and you can't do S16 with them. I'm speaking of a mid-range scanner where you scan the whole frame, wheather 16 or S16 and not lose information by going into a camera. Where you are not dumbing down the frame by the camera's limits. Your only limit would be drive space.

 

Richard Mills

Cinematographer, Photographer, Editor

Phoenix, AZ

[Richard Mills]

Let's say someone put one out and it cost $20,000 -- how many of you would buy it?

 

And if they only sold it for $1000, how would they make a profit considering the low number of people who would buy one compared to people who buy one for still photography?

 

If one could be made cheaply and sold at a profit and give halfway decent results... then why hasn't it been done yet?  Could it be that the market is simply too small for a company like Minolta or Nikon, etc.?

<{POST_SNAPBACK}>

 

 

David

 

You are correct, no one would spend $20K for a scanner. But, they will spend $20K for a midrange HDV camera :)

 

For example, I think the Coolscan has an adapter for rolled film, why not make a larger transport with a coaxial mag to hold 200ft of 16. Doesn't seem that hard to do to me.

Yeah it is a nitch market, but a market none the less.

 

With the world going digital, the scanner market will shrink anyway so what are the going to do with all those scanners? Maybe we as potographers didn't scream loud enough in the beginning. It seems to me that we could be extending the product cycle of another dying product :)

 

Richard Mills

Phoenix, AZ

[Phil Rhodes]

Hi,

 

The camera shutter issue isn't a problem; you rig it open, and flash the backlight.

 

Phil

[Mike Lary]

For example, I think the Coolscan has an adapter for rolled film, why not make a larger transport with a coaxial mag to hold 200ft of 16.

<{POST_SNAPBACK}>

Nikon has an adapter for 35mm roll film, but not 16. They have a 16mm strip film carriage that does two strips at a time, for the ridiculous price of $350 :blink: It's just two pieces of plastic and a plastic hinge!

[Nate Downes]

Yeah I have looked into their line of machines.  Interesting concept but, they are telecines and you can't do S16 with them.  I'm speaking of a mid-range scanner where you scan the whole frame, wheather 16 or S16 and not lose information by going into a camera. Where you are not dumbing down the frame by the camera's limits.  Your only limit would be drive space. 

 

Richard Mills

Cinematographer, Photographer, Editor

Phoenix, AZ

<{POST_SNAPBACK}>

 

Why can't you do S16 with them? The whole frame, including the sprocket holes, are available to access.

[David Mullen ASC]

The issue isn't technology, it's economics of scale. Look how much that guy is charging to just point a 16mm projector at a CCD camera (and I'm not saying that it's a rip-off either.) There isn't the market demand necessary to make such devices cheaply.

 

This would have to be a do-it-yourself engineering project.

[Richard Mills]

Nikon has an adapter for 35mm roll film, but not 16. They have a 16mm strip film carriage that does two strips at a time, for the ridiculous price of $350 :blink: It's just two pieces of plastic and a plastic hinge!

<{POST_SNAPBACK}>

 

 

Mike

Sorry, my bad.

I couldn't remember which scanner company had the roll film adapter for 35mm roll :(

And the two strip thing for 16mm is a joke.

 

Richard Mills

Phoenix, AZ

[Richard Mills]

Why can't you do S16 with them?  The whole frame, including the sprocket holes, are available to access.

<{POST_SNAPBACK}>

 

 

Their site has no mention of S16, just 16mm.

There is the single perf factor with S16.

 

Plus there is the telecine factor. I don't want to be at the mercy of a camera. I want to scan right to a file without having to deal with a camera's chip and compression.

 

I'm not saying that this is a bad product, it is just not what I'm looking to do. Though it maybe the only option out there.

 

Has anybody seen the results of this product or used it?

 

Richard Mills

Phoenix, AZ