Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Fastest and cheapest 35mm film scanner on the planet

Peter Majtan

Peter Majtan

Well-known member
Joined
Feb 29, 2008
Messages
3,986
Reaction score
0
Points
36
Location
Tokyo, Japan
Website
www.petermajtan.com
Did You guys realize that the EPIC could be modified to become the fastest (100 FPS) and cheapest (bellow $100K total setup with fast raid to capture the RAW sensor data uncompressed) 35mm 4K film scanner (after de-bayering from 5K) on the planet?!?!?!

Dibs on the patent...
 
I'm dying to see how it will stack up against the 6K ARRI-LASER (my favorite for now...). The speed alone is amazing. You could even slow the speed down for multi-exposure for HDRI exceeding the latitude of the film stock itself. Even if I did that at 10 FPS - it would be awesome... (and quite ironical...)
 
Yeah, there was such a discussion here and also - some options were already in the works.

Maybe this would work with RED1 ans Scarlet too?

Scarlet - for 8mm and 16mm?
 
I hope so :).

But what about the intermediate device? Maybe it will be $$$? :(

And I guess that won't be a scan of the negative? Yes?
 
What intermediate device? Do You mean the film transport? Any old movie projector could be modified for this purpose - just don't use the same light...

You could scan negative just as well. All You will need to adopt is the light-source and then invert the image with some curve control in the post...
 
when you shoot a 3 perf film, you still need to process the film in the lab, am i correct? is there something called 'positive' film where you shoot and directly view it in projector?
 
I thought about this with a red1 a while ago. I started a thread which ended badly. Lots of people arent happy with 'truncating' films color resolution with a bayer pattern scanner, (although i'm pretty sure several major film scanners use bayer pattern sensors).

Difference in quality aside, i think you might find a market with a Red based (epic, one, or scarlet) scanner, simply because the cost would be easily 1/5th the price. If i shot my own film again thats what i'd do.
 
A projector as the film transport for a film scanner? Have you seen how bad the registration is on a film projector?

Also, the light source for a film scanner has to be very tightly controlled and perfectly even. The most costly part of a scanner is the film transport.

You could probably use a modified Mitchell NC as the film transport but it would be difficult to integrate the light source. And there would be no way to integrate a wet gate.

I do agree that RED has some technology that would be adaptable film scanner application (especially REDCode). It remains to be seen if they would be interested in entering that market.
 
As far as the light goes - high powered LED grid with balanced diffusor could do the trick. It will require calibration to match the sweet-spot 320 ISO (if that is going to stay with Mysterium), but it is doable.

I am aware how bad the registration can get, especially on a retired film projector. The thing is that film itself is not exposed 100% in the exact position by the original camera, but there is a solution. First You overscan the area by lets say 5% and then a simple SW utility could re-align the footage on a frame-by-frame basis. And since I would not compress the RAW signal, I would not have to worry about compression artifacts. Sure it will be challenge - but one that I am confident to take up...

I know in my right mind that this will not replace the the films scanners. But do you remember those who would transfer film to video by projecting it on a screen and filming it with video camera? I am sure there would be a market for such a installation. I could see for example Scarlet to make a fine and affordable 2K scanner for 16 mm film...
 
DIY scaner and film recorder

DIY scaner and film recorder

If you are serious about making a DIY film scanner and or film recorder I already have software made just for that purpose for "freeish" at my Web site.

DANCINES.EXE (tm) runs a scanner projector and digital camera in single frame mode, can make multi-exposures for HDR and can operate a IR+RGB light house. It can read a cue sheet to skip over frames in the film roll you do not need scanned to save time.

DANCINEL.EXE (tm) runs a DIY film recorder, currently the highest resolution video board I know of that works is 2048x1536x32bit, but if you make color seperation exposures that comes out to true RGB pixels.

My Program DANCAD87.EXE (tm) is used for color correction of both color negative and positive and is being upgraded to handle up to 15K+ resolution DI (Digital Intermediate) and have intergrated 16 channel sound mixing and true 24/25/30 "proxy" editing with SMPTE master/slave for the sound playback on a dual monitor two computer sync setup.

DANCINES.EXE (tm) was developed from a 35mm printer control program and can be used for both film printing and scanning.

If you have any questions feel free to contact me at the email at the top of each page in my Web site.

I have been able to import RED ONE (tm) frame by use of the freeish program XnView (tm) so you could mix frames from Epic and Film scans probably, in the edit list.
 
I'm looking forward for your tests, Peter.

If you manage to make an affordable [not cheap! :)] solution... I'm your potential client.

These things are very exciting.
 
Cheep...

Cheep...

karapetkov said:
If you manage to make an affordable
Click to expand...

If you can hook up your RED ONE (tm) to single frame you can use a old Moviola editor, movie projector, or optical printer and it to scan for probably less than $500 to $2000.

I am using a Canon XTi (tm) in our test scanner now made out of Oxberry printer parts. A monochrome industral camera may be better since the Bayer filter adds artifacts.

The advantage of the Epic (tm) and RED ONE (tm) etc. is the lack of a mechanical shutter to wear out. DSLR wear out before 250000 exposures probably.
 
Wet gate...

Wet gate...

CinemaTechnic said:
And there would be no way to integrate a wet gate.
Click to expand...


If you arange the scanner verticaly you can put the NC movement in a "tray" of perc with the light comming up from under through a window, and a window at the surface to avoid ripples.

You can add things to the perc to adjust the index to get closer to the film base, then clean the film after scanning for storage.

The Acme type movement is better since dust can collect at the edge of the NC movement aperture, a mitchell High Speed movement can go up to 135fps and if under fluid to cool it it might hold up.
 
That sound awesome Dancad3D. Would You ever consider re-writing and re-compiling any of that SW for OS-X platform? I have my own SW for the film recorder done on the OS-X and I would prefer to stay on this platform for several reasons I don't want to elaborate here.

Rado, "EARLY 2009" is still long way ahead and when we get nearer to some release date announcement I will recap all my research. It just could be that after the three redheads hit the streets there won't be enough "potential customers" shooting on film anymore to justify the R&D cost. The ones who would still shoot film will be the ones with substantial budgets who normally use large post-houses equipped with the latest pro film-scanners anyway...

Lets see...
 
DYI its there take it or...

DYI its there take it or...

Peter Majtan said:
That sound awesome Dancad3D. Would You ever consider re-writing and re-compiling any of that SW for OS-X platform? I have my own SW for the film recorder done on the OS-X and I would prefer to stay on this platform for several reasons I don't want to elaborate here.

Rado, "EARLY 2009" is still long way ahead and when we get nearer to some release date announcement I will recap all my research. It just could be that after the three redheads hit the streets there won't be enough "potential customers" shooting on film anymore to justify the R&D cost. The ones who would still shoot film will be the ones with substantial budgets who normally use large post-houses equipped with the latest pro film-scanners anyway...

Lets see...
Click to expand...

My Scanner program can run under FreeDOS (tm) probably on a junk computer like a PII, so there is no cost involved, it talks directly to the parallel port and outputs TTL control signals to run the relays that operate the projector advance motor and the camera, there is also a hold input that goes to a photo transistor on the red light on the back of a DSLR so that the software does not get ahead of the camera.

I do not know of any Pascal development tools that would make a conversion easy, maybe FreePascal (tm) might work, but I have not used a MAC for control functions, it is better to use DOS since there are not background programs to mess with the "microsecond" timing.

There is very little R&D involved with my programs, an optical printer could be converted in a few days probably.

===

About the market, there will probably be a market for film recorders for maybe the next 7 to 9 years before conversion, but for film scanning the market will get smaller, and start small since you can purchase a RED ONE (tm) for less that the price of processing for a 88 minute feature film, probably, I would think everyone who can will shoot Digital rather than film, and you can use two RED ONE (tm) to shoot HDR and maybe get a longer scale than the Epic (tm) or film?
 
You must log in or register to reply here.
Back
Top