Jump to content

Comparison of Original vs Restored 8mm film scans


Guy Burns

Recommended Posts

A Secret Program

As part of having a test scan done on a Laser Graphics ScanStation, I was offered a couple of minutes of restored footage. For $7 a minute, this company will restore any movie file. When I asked what software they used, I was told:

 

 

Sorry, but it's a secret; a program of our own. It took 8 years to develop.

 

I was dubious about a program that was 'secret', but since the restoration offer was free, I said to go ahead, and they uploaded several gigabytes of sample files from my films.

 

Problematic Test

Of course, they buggered up the test. During the scanning, the exposure was set to auto, and colour and contrast corrections were applied. Then during restoration, more colour and contrast corrections were applied, the image was stabilised, and frame-blended from 18 to 25 fps. All of which, several times, I specifically requested not to be done – that I wanted the scan and restoration as raw as possible to enable a proper comparison.

 

I was amazed, actually, that a company would outlay on a ScanStation, then treat it like a RetroScan.

 

Restoration Results

Anyway, I really was amazed when I saw the results. Now, let me qualify that statement: this is about 8mm amateur films, in poor condition, 40-80 years old – inherently shaky, poorly focused, faded colour, and filthy. I'm not talking feature film.

 

Dust and grime were, to me, in most cases impressively removed, at the expense of a slight softening. However, I have no experience with restoration, so I may be missing something, and would appreciate comments from others.

 

Download

A side-by-side sample can be downloaded here:

 

http://www.mediafire.com/download/so9g97btews9jvo/Restoration_Comparison.mp4

 

It is not meant to be played as a movie. It is a series of short segments, with typed comments here and there. It should be stepped through frame-by-frame.

 

If anyone wants to see a longer version to enable a motion-comparison, let me know and I'll upload it.

Edited by Guy Burns
Link to comment
Share on other sites

Owner of Film 2 has his say

A comment (by email) from the owner of Film 2. He's a retired electrical/instrument engineer, with an very good understanding of technicalities, but fairly new to video. I'm using his film as a reference because it has a history of being scanned, allowing comparison of scan quality:

  • Nikon scanner of some sort, in 2010, in Sydney
  • Epson V700 (flatbed)
  • RetroScan Universal
  • ScanStation

Here's what he said:

 

I have had a look at the video frame by frame and I am quite impressed.


Dirt and scratch removal along with filtering and sharpening is pretty good. However it’s a pity about the first few frames after a scene change. Strikes me their software needs a few frames to establish the difference between image & dirt. It seems a simple solution, run the algorithm again on the film running in reverse. Ask them they may plan it in their next software version.

They also seem to be using image stabilisation to good effect. This invariably lead to some "zooming in" & cropping to give them an edge to trim. Have a look at the cross on the top of the church in the film of the girl running.


I see image stabilisation as important for our film which is shot hand held with a lesser quality camera.

I thought the frame blending was good on the girl running, when you consider the original frames were pretty blurred due to the high subject movement & low shutter speed. Did you check it out on slower moving subjects?

Overall I think it's pretty good. We would be happy to pay to have our film restored by them if their charges are around the prices you originally indicated.

I have my say

Frame-by-frame lets you see where the problems lie – and there are several – but the real aim of this exercise is to find out whether auto-dirt removal can work effectively, and whether the projected image is improved. What is more distracting when viewed as a home movie: serious amounts of dirt, or slight degradation of the image?

 

The next step is to project the images one at a time, then side-by-side, on a 3-metre screen, telling the small audience exactly what to look for, and then ask which image they prefer. Luckily, I have regular access at home to a small audience who come along for our Blu-ray movie nights, so it will be interesting to hear what they think.

Edited by Guy Burns
Link to comment
Share on other sites

ICE came out of Applied Science Fiction – and that's just what it is. To be able to remove, almost perfectly, thousands of dust spots in a few seconds is akin to magic for me. You can download an example here: http://www.mediafire.com/download/kec2iw6jstsydoa/GM126.tif

 

It's a large file (100 MB) scanned at maximum resolution on my Coolscan. Taken in November 1964, probably on Agfa. Two layers, one with ICE, one without. Jump between the two in Photoshop and you'll see magic in action.

 

I generally use two layers so that I can mask areas of the image where ICE is not required and in which it may cause problems; or areas where I want to manually clean.

 

If movie-film scanners had ICE, it'd just about put restorers out of business.

Link to comment
Share on other sites

If movie-film scanners had ICE, it'd just about put restorers out of business.

 

Some do, and it doesn't.

 

It's not a panacea, and it only works for specific problems - namely physical gunk on the film itself (vs baked in dirt in prints). IR light is used to determine where the stuff on the surface of the film is, and then clean only that stuff up. It's a neat idea and it does speed up some aspects of restoration (and most professional restoration software can take dust maps made by scanners with IR cameras, to only apply automated effects to those areas. But it's not perfect, and it only gets at some of the problems). You also can't compare single-frame cleanup to motion picture cleanup. Eyes are tricky things - it's very hard to spot a defect from cleanup in a single frame, but it can be incredibly obvious once there's motion.

 

There simply is no software that can do this automatically for motion pictures without major compromises. We've been doing digital restoration work for almost 15 years and have tried most of what's out there from high end professional systems to homegrown setups using open source software. None of it works without causing problems, which is why restoration is primarily done manually, or you wind up with footage that looks like what you posted: With reduced resolution, reduced grain, an overall waxy appearance, and artifacts.

 

The owner of your film is correct: automated restoration software all works by looking at surrounding frames to estimate the motion and figure out what things are transient defects. In this case, the software is not looking both forward and backward, at least at the end of the scene, and they're just stopping when they hit a scene break. "All" they need to do is find a scene change to know they're at the end of the cut, and then on those last few frames start looking at previous frames. I put "all" in quotes because it's more complex than that. That will kind of work, but it will still result in artifacts because it's an algorithm, not a person, doing the decision making.

 

The problem with using surrounding frames in an automated way is that the amount of motion in those frames may mean there's nothing to pull from when either detecting the defect or concealing it if it was properly detected. Even using these same techniques in manual cleanup, the end of shots are hardest because of this very problem. In most manual tools you can specify the direction to do motion estimation in (forward or backward), as well as the number of lookahead (or behind) frames. You can also specify what type of motion there is (fast motion, slow motion, etc) to help the algorithm along. It's a painstaking process because it's so subjective. The algorithm may find something that's a defect, but it may not do a good job in the cleanup.

 

"Police" is clearer because they're smoothing out the image first then applying artificial sharpening.

 

At the end of the around frame 130, you can clearly see the fake grain they're applying. it's like a screen over the image that doesn't move correctly.

Edited by Perry Paolantonio
  • Upvote 1
Link to comment
Share on other sites

  • Premium Member

I've use ICE on Nikon Coolscans with 35mm still film for many years with generally decent results. Never gets everything and quite often there's a little halo around things it fixes, but as a timesaver for non-critical work it's great.

 

Perry's the expert but I would bet on motion picture film, those little halos around "fixed" elements might be just as distracting as a piece of dust.

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...