Jump to content

Scanning Super 8


xoct

Recommended Posts

In an audio studio everything finally ends in a cd at 44.1 Khz. Maybe in a dat at 48 Khz

Even if the sound is going to end at this CD quality all the digital sampling is normally made at 96 kHz in order to be more close to the real sound emited by the voice, amp,..

AS long as i know this is called oversampling

I guess the word might be used in many diferent worlds

Link to comment
Share on other sites

  • 2 weeks later...

Interesting topic. There is a program called “8mm2avi” it’s a program to convert 8mm films using a scanner to AVI file. Has anyone here ever used this program?

 

I downloaded it, but found out I don’t have the right scanner and it’s resolution is way to low. Plus you need a tremendous amount of patience, because you to can only scan a small strip of film at a time. Then the program removes the frames from the scanned image and puts it all together into a AVI file. Neat idea, but I’m not sure how well this method would work.

Link to comment
Share on other sites

  • 4 weeks later...
  • Premium Member
Interesting topic. There is a program called ?8mm2avi? it?s a program to convert 8mm films using a scanner to AVI file. Has anyone here ever used this program?

 

I downloaded it, but found out I don?t have the right scanner and it?s resolution is way to low. Plus you need a tremendous amount of patience, because you to can only scan a small strip of film at a time. Then the program removes the frames from the scanned image and puts it all together into a AVI file. Neat idea, but I?m not sure how well this method would work.

 

I could envision a "handcrank system" where a strip of film frames are scanned by the scanner. When the scanning is done you give the crank a turn and the film advances on the scanner to the next set of unscanned film frames and so on.

Link to comment
Share on other sites

As someone who is at the finishing stage of a tedious (not to mention expensive!) 2 + year long R&D on this exact thing (film scanner), I can tell you that this idea of using a scanner is not impossible.

Merely insane.

 

Even IF you could manage the registration problems associated with hand-placing film on a flatbed scanner, it would take literally months, if not years, of full-time, 40 hour work weeks to digitize any significant amount of film.

Just do an estimate yourself as to how long it would take.

You lay a strip of film down, and scan maybe 50-ish frames, that takes say 5-minutes.

(That doesn't include how long it's going to take you to strip out each frame after it's in the computer, assemble them sequentially, move the frames so they register correctly, etc.)

 

Do the math.

 

Just absolute madness, believe me.

 

Matt Pacini

Link to comment
Share on other sites

  • 2 months later...
Guest Aaron Francis
Just absolute madness, believe me.

 

Matt Pacini

 

Actually, it isnt all that mad if done properly. With a little (ok, more than a little) knowledge and a good dose of creativity, it can all be automated and you can sit back and watch. Check these links out

 

http://www.jiminger.com/s8/index.html

http://www.public.iastate.edu/~elvis/8mm/film_scanner.html

 

A good mesh of robotics, programming, and film

 

Aaron

Link to comment
Share on other sites

  • Premium Member
Actually, it isnt all that mad if done properly. With a little (ok, more than a little) knowledge and a good dose of creativity, it can all be automated and you can sit back and watch. Check these links out

 

http://www.jiminger.com/s8/index.html

http://www.public.iastate.edu/~elvis/8mm/film_scanner.html

 

A good mesh of robotics, programming, and film

 

Aaron

 

The author states "Since I had no intention of scanning all of that film by hand, I built a mechanical film-advancer controlled by the computer." which I think was Matt's point, handcranking would be crazy.

 

I was thinking more from a time-lapse point of view. I once spent fifty hours shooting one cartridge of Super-8 film. I certainly would not find it inconvenient to hand crank that one film reel through even if it took a couple of hours, but if one is shooting live action, than that could become more than tedious.

Link to comment
Share on other sites

  • 4 weeks later...
Hello,

 

Yes but as the film is digitized, Shannon Nyquist theorem would also apply for spatial resolution. So in theory 2K might be the digital resolution equivalent of top quality super 8mm.

If going to HD, should be uncompressed HD RGB, so film transfered to hard drives as there is no current uncompressed HD RGB tape capable, as far as I know.

Regards,

Daniel

 

Nyquist says you want to sample at twice the information rate to get all the information. So by sampling at 200 pixels/mm when we think there is 100 pixels/mm of information means we get all the information as per Nyquist. Don't need to double it again.

 

But what about color depth? I just had a lot of 1955-1970 home movie regular-8 and super-8 film transferred to mini-DV via a Rank Wetgate system. It looks great. I think its safe to say the film was not anywhere near the theoretical maximum in terms of image quality. How much information do you think was left on the table?

 

The mini-DV is 4:1:1. I guess when I transfer mini-DV to DVD, I am getting 4:1:0? Is that true and is there anyway to avoid that information loss when transcoding to MPEG?

 

Alex

Link to comment
Share on other sites

  • Premium Member
I think its safe to say the film was not anywhere near the theoretical maximum in terms of image quality. How much information do you think was left on the table? 

 

Alex

 

When you say "the film was not anywhere near the theoretical maximum" did you mean film transfered to video, aka the video master?

Link to comment
Share on other sites

When you say "the film was not anywhere near the theoretical maximum" did you mean film transfered to video, aka the video master?

 

I read the post earlier in this section saying that the estimate of super8 maximum theoretical resolution of super8 film recorded inside a camera would be about 1062 x 802. Actually as I understood it the resolution would be half that but that would want to sample at twice that resolution to be sure to capture all the information as per Nyquist.

 

I just transferred regular8 and super8 films from about 1955-1975 to miniDV tape at 720x480 4:1:1. So I wanted to get some feedback as to how much image content I failed to capture.

 

It was not clear to me whether the above theoretical max estimate assumed modern film stock and how the fact that this film stock used 30+ years ago and the performance of a consumer/home super8 camera of that period would affect the estimate of maximum resolution. Also what are the numbers for regular8?

 

I understand we can get away with throwing away U and V color content because of the diminishing returns in terms of human's ability to perceive color resolution. How perceptable and improvement would there be with 4:2:2 or 4:4:4 images giventhe source of the image?

 

As I understand it, when I transcode the miniDV data to MPEG I go from 4:1:1 to 4:1:0. Is that true? Is there any way to save that color information in the MPEG? I have Adobe Premiere Pro 1.5 and Encore.

 

Alex

Link to comment
Share on other sites

As I understand it, when I transcode the miniDV data to MPEG I go from 4:1:1 to 4:1:0.  Is that true? 

 

Yes that is true. If you are work in NTSC and mini-DV there is no way around. Your options are move over to PAL (where dv and dv-cam are 4:2:0), or transfer to a higher quality format like DVC-PRO 50 (or similar that records a color depth of 4:2:2)

Link to comment
Share on other sites

Yes that is true.  If you are work in NTSC and mini-DV there is no way around.  Your options are move over to PAL (where dv and dv-cam are 4:2:0), or transfer to a higher quality format like DVC-PRO 50 (or similar that records a color depth of 4:2:2)

 

Yes, I have been thinking that PAL would have been a good idea also for the fact that it captures 20% more lines of resolution. Not sure if my camera could have read PAL into my computer and whether there would be any strange video effects of converting it to NTSC?

 

I used Cinepost and it cost $990 for about 2:52 of runtime. I saw the next higher grade storage was Digibeta but that was $50 per hour more and so I expected that it would have cost about $400 more include tapes. As I understand it it would have gotten me deeper color resolution, but the capture would have been NTSC or PAL pixel resolution anyway.

 

If we gather some more film, any recommendations on another place that would be cost effective but capture all the information on regular 8 and super 8 tapes. I hear all this talk about spirit and shadow telecines? How do they compare with a Rank Turbo wetgate as Cinepost uses in terms of quality and price?

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...