This page is titled How to Restore 8 mm Films: DIY Guide and Tips, but I strongly advise against restoring your home movies using the DIY method.

Beyond the fact that restoration software is very difficult to learn, there is an insurmountable problem for those who do not want to rely on a professional lab:

There are no high-quality 8mm and super 8 scanners for less than 20 thousand euros.

The most economical model to consider is the FilmFabriek Pictor Pro, which you can find illustrated by the manufacturer on this page.

All chepaer scanners are toys. Starting from the Reflecta Film Scanner Super 8 - Normal 8, which costs around 400 euros on Amazon and of which there are virtually identical versions from other brands (Film2Digital, Somikon, Kodak). Or the Reflecta Super 8 Scanner, which is not difficult to find in the second-hand market, precisely because many want to get rid of it after testing its performance.

Many years ago, when I set up the lab, I started by buying the basic scanner from Moviestuff, spending about 5 thousand euros. The image below is a photo of the first version of my equipment:

With the Moviestuff Retro 8, I used FinalCut software for restoration. When I switched to the professional scanner I have today, and at the same time replaced Final Cut with DaVinci Studio, I began redoing all the videos I had previously restored. Years of work, which I still have to finish, but it's worth it, considering the vast difference in quality between the two systems.

I know that the internet is full of tutorials from people who modify an old projector and turn it into a device that not even Hollywood could aspire to have. It would be very nice if things were like that. I would have much preferred to achieve the same quality with a 1970s projector found on eBay for 100 euros and a bit of work to adapt it, rather than going into debt to pay for the equipment I bought.

A scanner is not like a computer that, five years after being released, is worth a quarter of what was paid for it. The cost of a scanner is only minimally influenced by the processor and computer-related components. The price is determined by:

  1. Lens.
  2. Image sensor (CCD or CMOS).
  3. Cost of the project.
  4. Post-sales support

Therefore, scanners do not depreciate on the market like computers and smartphones, and prices won't decrease over the years at the same rate as in other technology sectors.

The photo above depicts the professional scanner FilmFabriek HDS+ that I use in my studio today, along with the workstation it is paired with.

Professional Labs and low-cost Labs

Even when relying on professional restoration services, it's not easy to choose the right lab because the professionalism with which they work makes the difference

That's why among labs that digitize films, there are:

  1. Budget services that simply use a camera (or smartphone) to capture the image on the wall illuminated by the projector (a procedure that risks irreparably damaging the film due to the projector's gears) and leave the restoration burden to the camera's automations.
  2. Professional services provided by labs equipped with modern equipment and staffed by individuals proficient in image correction software.

I explain the difference between these two options in depth in a guide I wrote for the site.

I have no problem publishing dozens of restored films on this site, showcasing what I am capable of achieving. Like I did with this 1988 super 8 film shot in Melbourne, which I am sharing with the written permission of the owner.

DIY Restoration of 8mm and Super 8 Films

As mentioned, having a high-quality scanner is essential for achieving satisfactory results. The scanner I utilize, the FilmFabriek HDS+, comes with a price tag of tens of thousands of euros. This cost not only makes it unsuitable for individuals looking to digitize their personal collection of 8mm and Super 8 films but also places it out of reach for labs handling fewer than 1000 films per year.

I am able to afford it for one straightforward reason:

The restoration system I have established caters not only to the home movie that clients send but also to a library of historical footage that I supply to documentary productions worldwide.

As demonstrated by the credits of a series airing on PBS, the American public television. Among the providers of historical material, my archive, FOOTAGE FOR PRO, is also listed.

Thanks to these two projects based on the same technology, I was able to afford the FilmFabriek scanner, which acquires:

  • 8mm
  • Super 8
  • 9.5mm
  • 16mm

films, both silent and with optical or magnetic soundtracks.

Speaking of professional equipment, let me give you a piece of advice. More than 90% of 8mm and Super 8 films were silent. If the service you're consulting says it can't capture audio even when it's present, it means they're using a very cheap scanner. Not only does this prevent sound digitization, but it also results in poor image quality.

Another clue to identify those using inadequate equipment is the processing times. When they exceed 7 days, they most likely use the 400-euro scanners available online because these are entirely plastic and very slow at digitizing.

Occasionally, I browse websites offering services claiming to be similar to mine. However, only a few of them publish excerpts from the films they've worked on, perhaps masking this absence with phrases like professional software, artificial intelligence, top quality, or leading company

The only important element in digitization is the resulting video, and if they don't show you examples of what they've already worked on, I wouldn't trust them if I were you.

Below, you can see an 8mm film shot in Naples in 1957.

The footage is indicative of the quality that can be achieved by professionally digitizing amateur films from that era. In some respects, it's even below average since the cameraman isn't technically good, and the film is older than most others, whose golden age is represented by the 1970s when cameras and films were slightly better than those of 1957.

I will use this footage to illustrate the various stages of restoration.

The first phase of restoration: digitizing the films

There are two types of systems available to amateurs for digitizing films:

  1. Recording the projector frame with a camera or smartphone.
  2. Using 8mm and Super 8 scanners costing a few hundred euros.

Working with a camera and projector has the unacceptable drawback of putting the film at risk because projectors have been out of production for over 40 years. Even when they were on sale, films could get stuck in the gears and break.

As a second flaw, they use an incandescent lamp. If the film stops even just for a second, the frame burns.

With these premises, anyone who cherishes the memories captured on their films, perhaps related to people who are no longer with us, should stop and think if it's really worth it.

On the other hand, budget scanners probably cannot damage the film. The film's movement is external to the device, so there's no risk of it getting jammed. And they use a cold LED light for illumination and have another significant advantage: they capture frame by frame.

8mm and Super 8 films were shot at 16 or 18 frames per second, while modern cameras and smartphones have 25 or 29.97 frames per second. Filming at 25 or 29.97 frames per second with today's equipment means that in each modern frame, two frames of the original footage overlap, resulting in a disaster in terms of final quality.

The problem with budget scanners is that despite their advantages over projectors, they still produce poor-quality footage due to:

  1. Low-quality lens.
  2. Sensor detecting the image like a toy camera.

Here's a comparison between amateur and professional digitization. The film was shot in Super 8 in 1980 during a vacation in Egypt.

The owner had the unfortunate idea of turning to the wrong laboratory. I want to thank this person because they authorized me to share the images to prevent others from making the same mistake.

Why using underexposition is the right setting during scanning

Both amateur digitization systems have the drawback of not allowing the operator to use manual settings during digitization. Automatic mode, if the film is perfect, may be an applicable solution. The problem is that home movies, being shot by amateur filmmakers and having decades of life, often present issues that only:

  1. Professional equipment.
  2. An experienced operator,

can resolve. Since  the acquisition phase.

Films should indeed be digitized with a slight underexposure. To achieve this and allow the restoration software to handle it best, the scanner must have a high bit depth. Bit depth indicates the amount of color gradations each color has. Only with sufficient bit depth can brightness of dark images be increased during restoration without drawbacks.

If, on the other hand, overexposure is acquired, and then brightness is decreased on the computer, the image quality would be lower, even with the same bit depth.

Of course, if the device used for acquisition, whether a budget scanner or the camera used to film the projector, does not allow the operator to manually adjust exposure, you do what you agreed to do when opting for the DIY route: you settle. Perhaps thinking of the typical excuse used by amateur labs when customers complain:

The film is old, so it's normal for the quality to be poor.

This is absolutely not true.

The best software for restoring home movies

To restore 8mm and Super 8 films, as well as 9.5mm and 16mm, I use Davinci Studio, paired with the Neat Video plug-in, specialized in removing black dots, scratches, and grain.

There's a free version of Davinci Studio: Davinci Resolve. It differs in a few aspects, so if you only need to do occasional work, perhaps the paid version isn't worth it.

Other video editing software, like FinalCut, Adobe Premiere, and Avid, are equally capable of image correction and restoration. As a professional editor who has used them all, I find them less practical than Davinci Studio. However, this is a very subjective matter because there are professional editors or colorists - those professionals who specialize in image processing alone - who may have different opinions from mine.

What I definitely advise against is using amateur software like Pinnacle Studio, Movie Maker, or Moravi Video Editor. These are fine for basic editing if you don't have specific needs, but they have basic image correction functions, often almost entirely automatic. Consequently, they struggle with complicated images like those in historical films.

Returning to the miraculous tutorials occasionally found online, there is also an open-source program called VirtualDubMod. I know it well because, before opening the lab, I thoroughly tested it. While it's very difficult to use, I can say  that things are different from what you see on YouTube.

Regarding artificial intelligence applied to restoration, there are two popular options:

  1. Topaz Video AI
  2. Runway

I continuously monitor them because to do my job well, one must stay constantly updated. All the tests I've done since their launch have been disappointing because the artifacts they generate are unacceptable.

I'm sorry to disappoint those who believe that technology in the restoration sector is already working miracles, allowing the operator to avoid effort and remain ignorant because AI does everything:

To this day, to restore 8mm and Super 8 films well, years of experience and a lot of work on each film are still required.

After years of hard work, both on my personal archive (you can also see it on one of my YouTube channels) and on clients' films who turn to me, only now am I perfectly capable of removing most of the signs of time that accumulate on films. At the beginning, despite coming from a background as a professional television editor, I wasn't up to the task, and I had to spend time learning. Imagine if I had been an amateur with no significant experience.

Restoration of 8mm and Super 8 Films in Davinci Studio

As we've mentioned, ideally, films should be acquired with professional equipment because, among other things, they have the advantage of allowing manual settings, starting from the ability to slightly underexpose the image.

Frame Resizing

After being scanned, the footage of Naples from 1957 appeared as seen above. The main frame included portions of the previous and next frames, and even the side edges included external areas, like the hole where the gears of the projector and the camera were mounted at the time.

This is necessary because, in certain situations, the image moves, and only if the captured image is wide can the movement be compensated for.

To get to the final video, the first thing to do is:

  1. Enlarge the frame to fill the screen.
  2. Crop the sides.
  3. Center the image.

Once this is done, you move on to the next phase: color correction.

The video I'm sharing below compares the Naples footage in these two stages:

  • On the left, after resizing only.
  • On the right, after resizing and color and brightness correction.

Color Correction

Before discussing color correction and brightness adjustment, there's an important premise to consider:

Professional scanners have a function that allows for automatic adjustment of the lamps, increasing intensity in dark scenes and decreasing it in bright ones.

However, the light works linearly, meaning it cannot be adjusted to differentiate between dark and bright areas. This is problematic because 8mm and Super 8 films deteriorate over the years, especially in shadowed or backlit parts. Therefore, to restore the films to their original quality, it's necessary to work with software curves, which allow for differentiation in brightness adjustment.

In the Curves function of Davinci Studio and Davinci Resolve, you must act on the left graph, whose line starts straight:

and after the correction, it looks like this:

The white line has been adjusted by inserting and moving a point upwards on the left side. This means that only the dark colors have been raised, as seen in the Scopes, the graph with the three basic colors: Red, Green, and Blue, on the right.

As an additional adjustment, the red color was enhanced because 8mm and Super 8 films tend to acquire a slight cool dominant over the years.

As clearly visible in the preview frame of the video above, working in this way, perhaps in a challenging shot where there are different light areas, brings back subjects in shadow that would otherwise be lost.

Grain Correction

After resizing the frame and correcting the colors, ideally done scene by scene, even though this takes a lot of time, the next step is containing the grain and black dots and scratches.

Grain is somewhat of a trademark of films and is further accentuated when brightness is increased with curves.

Professional editing software, including DaVinci Studio, all have specific functions for this type of correction. However, none of these are as effective as Neat Video. Making these corrections has drawbacks, and Neat Video has fewer functions integrated into the software, especially in the command that allows you to remove black dots, which sometimes is imprecise and also removes other elements of the shot.

On the left, you can see the video after resizing and color correction. On the right, the same video to which grain and black dots correction have also been applied.

As evident from the preview frame, Neat Video does its job excellently. In some very problematic passages, it may not be perfect, but one can accept its effectiveness.

Sofware Image Stabilization

8mm and super 8 film cameras did not have stabilizers like modern cameras and smartphones do. Additionally, the film running inside further favored vibrations.

So, if today, with the advancement of technology, one wants to do a good restoration job, it's necessary to stabilize the footage. This can be achieved with almost all editing software. They have integrated a system to analyze the images and apply a counter-movement to the oscillations generated by the filmmaker's hand. The precision of stabilization varies from software to software, and I consider Davinci's stabilizer to be the best available.

The settings I use depend on several factors, not least the resolution of the timeline and of the digitized file. You can see the settings I used in the screenshot below. Pay attention: if they are set incorrectly, they can cause significant damage to the footage.

Thanks to my personal archive, I often find myself speaking with executive producers or archivists, professionals who are responsible for sourcing historical footage for documentaries.

Among the many things that surprise them when I tell them that my videos are generated from amateur 8mm and super8 films, is the high quality of the footage, which is not typical of images shot by non-professionals.

This quality that astonishes industry professionals is due, at least in part, to the significant improvement in software stabilizers in recent years. So, it's essential not to underestimate this step, as demonstrated in the video below.

On the left, the footage after resizing, color correction, and removal of grain and black spots. On the right, the same footage after stabilization:

Correcting the frames per second (FPS)

I introduce the final video that illustrates the different stages of restoration with a technical explanation.

One of the aesthetic differences immediately noticeable between an amateur film from the last century and a modern video is the aspect ratio of the screen. Once it was 4:3, while today it is 16:9. The former, therefore, on a modern monitor, is displayed with vertical black bars.

However, there is another difference that non-professionals may not be aware of and, unfortunately, is much more complicated to manage: the frames per second. In the past, it was 16 (8mm film) or 18 (super 8 film), and today it's 25 or 29.97.

As mentioned earlier, this problem begins to be addressed during acquisition using equipment that digitizes frame by frame. However, the issue persists in subsequent stages and until the final video export.

Although most TV and computer monitors are capable of playing videos at 16 or 18 frames per second, converting them in real-time to 25 or 29.97 frames per second, the way this happens is not as accurate as what can be achieved by working well on the computer during the restoration process.

That's why my lab never delivers videos at 16 or 18 frames per second, which would save me some time, but rather converts them all to 25: doing so during export produces better results than what monitors achieve with real-time conversion during playback.

Adapting the frames per second with restoration software can be done in three ways:

  1. Adding frames by repeating existing ones.
  2. Adding frames by creating cross-dissolves between some adjacent frames.
  3. Using artificial intelligence to recalculate scenes and recreate 25 or 29.97 frames per second from scratch.

The last technology is groundbreaking but has its drawbacks. The first challenge is that it is very time-consuming.The second is that it requires constant oversight by an experienced operator and, even with this, in some scenes, it is not yet able to calculate frames perfectly and generates artifacts that make the video unacceptable.

For this reason, my lab does not use artificial intelligence to adapt the frames per second with the films sent by clients. Instead, it uses the method of repeating frames.

All professional editing software has a similar function, and having tried them all, I am convinced that Davinci Studio is the most precise. For illustrative purposes, I show below the results of adapting the frames per second with artificial intelligence:

Conclusions on DIY methods

After sharing my working method, let me reiterate:

If you want to achieve a good video, you need very expensive equipment and years of experience with software restoration.

I know well that many, after seeing the different stages of the work, are already thinking that some steps are unnecessary and can be skipped because the results in the video are not so obvious. It's a human mechanism to convince ourselves that reality is as we want it to be, but in any subject, we should have the humility to rely on those with more experience, not because they are better people, but because they have greater theory and practice in what they explain.

I receive hundreds of emails a year, and my inbox is full of people who mock me, saying that they will save money and achieve the same results by filming the projector screen with their smartphone. Good for them...

DIY methods have their charm. I'm the first to use them to be faster and to solve problems. However, I only do so when it comes to simple operations that, if not performed correctly, will not have serious consequences. It's one thing to assemble an Ikea shoe rack, but if there are problems with the electrical system in your home, it's better to call a technician to avoid the risk of a fire if something goes wrong.

Restoring 8mm and super 8 films is a complex process. If not done properly, it does not enhance the memories of one's family history, thus ruining them forever – as in the video of Egypt that I published at the beginning of the page: for every person who turns to a professional lab to redo the work, there are 10 who accept the explanation that the poor quality is because the film is old. So for them, the memory will always remain:

  • Blurry.
  • Jittery.
  • With mismatched colors.

And anyway:

what's the point of going through all this effort to learn and then work only on films from your own collection?

There is no reason to spend months studying and investing a lot of money because once the work is finished, you'll have to put the scanner in the attic and negate all the effort you've put in.

Now that you know everything, make your choices.

Daniele Carrer

THE LAB IS BASED IN ITALY AND OFFERS FREE RETURN DELIVERY IN ALL THE EUROPEAN UNION

The price to digitize and restore your 8 mm, super 8, 9.5 and 16 mm films in my laboratory is always 5 euros per minute of footage, regardless of the format or the fact that they are mute or sound.

If you want me to work on your home movies, please contact me with this form: