8mm Film Damaged During Flatbed Scan

[jgroleau]

Hi. I am early in trying to convert some 8mm Film to DVD using Flatbed Scanner. As a test I tried to scan a small piece of film on my HP-Deskjet All in one Printer F380 just for a test. I set the DPI to 2400 and did the scan. It takes several minutes for the scan to complete. After the scan I noticed that the film frames I scanned were washed out of color compared to the film frames not scan. It appears like the scan has damaged the film. Has anyone encounted this problem? Is the light source of the scanner doing this? Am I doing something wrong? Thanks.

[DB83]

I suppose by having such a high dpi, which exposes the film to the laser light for an extended period, could damage the stock.

All-in-ones(which IMO are not flat-beds) are not designed for 8mm scanning. There are some flat-beds out there that can do it but I think they have a light source over the film and a special holder for it.

[loster]

To scan film using this type of scanner there must be a light source behind the film. Light will not reflect from the film to the scan head as with opaque objects. As far as I am aware flatbed scanners use a fluorescent light source not a laser based source.

[aedipuss]

yeah older scanners used cold cathode tubes, newer ones use led.

[nic2k4]

Are you certain there was color there in the first place? My old Apollo mission slides (AGFA film) have lost all their color. Yet, they were properly stored, not left exposed to heat, humidity and light. I guess it's entirely possible that exposing 8mm film to an intense light source after so many years, could burn off the color dye. You could try to reduce the brightness when you scan, then re-adjust it in photoshop; you will lose some info in the shadow. Does that scanner have a transparancy adapter? There are other methods of converting 8mm, google it.

[jgroleau]

Thanks for the responses.

I am pretty sure that the scanning caused the damage. The piece of film that was under the scanner cover is faded and the film looks fine where not scanned. My grand plan is to use a Canon 9000F which has a LED source in the lid for back lighting. I believe the light source in the all-in-one printer I used for the test is cold cathode flourescent. I do not think it is xenon or laser. I just don't understand why the flourescent would cause the damage. I do not think it is heat because the unit seems cool enough. The only other thing is that the light intensity or UV emission is damaging the color pigment on the old film. The strange thing is that the projector light source does not cause a problem. That may be because the frames do not spend a lot of time in front of the projector bulb. The film has not always been stored in ideal conditions so that may be contributing to the problem. I guess I will have to invest in the 9000f and see if LED light which does not emit UV causes the same kind of damage. I will have to either make an 8mm film holder or buy one. I also plan to build a computer controlled feed mechanism with a stepper motor and write software to control the scanner and feeder. That way I will not have to manually advance the film. I have also thought about writing software that will extract the individual frames (i.e. based on edge detection) from the scanned images and then stitch them into a DVD formated movie.

[DB83]

Well 8mm film will pass through a projector gate at 18/24 frames every second. if a frame jams in the gate there will not be much left of it.

Good luck with the project - my own method was much simpler but probably not as effective. Good enough for me tho.

BTW There are (ot atleast there were) specially adapted projectors with cameras, automated film advance and software avalable to do all this. If this is business orientated maybe these will be better for you.

[nic2k4]

Some guy used a digital SLR to do what you're trying to do, check it out here. I also remember seeing other projects; one used a modded projector that stopped at every frame and the image was captured with a web cam. Just remember that ultra high resolution is not needed, you only need to match the resolution of a DVD (720 or 704x480 in N. America).

[Des]

Some guy used a digital SLR to do what you're trying to do, check it out here. I also remember seeing other projects; one used a modded projector that stopped at every frame and the image was captured with a web cam. Just remember that ultra high resolution is not needed, you only need to match the resolution of a DVD (720 or 704x480 in N. America).

Interesting. I use a similar method for digitizing slides in bulk. Results are quite good.

I like the idea.

--dES

[2Bdecided]

If it was in focus, and the levels were right, and the dust was removed, it would be OK.

Cheers,
David.

[DB83]

Reading the write-up about the DSLR method, I would be most concerned about the w&t on the camera's shutter.

[Des]

Reading the write-up about the DSLR method, I would be most concerned about the w&t on the camera's shutter.

True. I believe there are 3600 frames to a 50' roll of Super 8mm movie film. Most consumer DLRS have actuation lives of around 100,000. So after about 25 rolls of film or 1200 feet you better start shopping for a new camera.

The HiDef Web cam may be a better method.

--dES

[diGriz]

There's an actual unit built for R-8/S-8 and 16mm film transfer involving frame-by-frame capture at High Definition, either through an RGB and Stereo input, or through HDMI. Heh... trust me, I had to learn how to use the unit (I transfer film and video sources in RL). It's called a Sniper HD, which involves an actual projector unit and 1080i/p camera unit plugged into an HDMI capture card (non HDCP). However, it's an expensive unit to build.

There is a group of people who do this kind of transfer you might want to look into, although I don't remember the name off the top of my head, who are based in Berkeley, CA.

Drop me a PM if you want to talk.