Colour-management for film restoration

In this article, we will briefly examine what a colour-managed workflow is, why you should use an OpenColorIO-managed workflow in PFClean and simplify some of the surrounding terminologies to help you navigate the process confidently.

What does colour managed mean?

Colour management is the process of ensuring that colours are accurately represented and reproduced across various devices and platforms. A colour-managed pipeline is a system that incorporates colour management throughout the entire digital restoration process, from capture to display to export. It uses colour profiles/transforms to ensure that colours are accurately mapped between devices, resulting in consistent and accurate colour reproduction. 

The goal of colour management is to maintain the integrity of the original colours and ensure that the final output is consistent with the intended artistic vision. This is especially important in digital restoration projects where accuracy and consistency in colour reproduction are critical to maintaining the authenticity and historical significance of the content. The most important part of a colour-managed pipeline is understanding the source colourspace of your unrestored clips so that no colour information is lost in the process. 

What is OpenColorIO?

OpenColorIO (OCIO) is a colour management system that provides the framework in PFClean to manage colour transforms and colour spaces between different applications and platforms. It is designed to be a flexible and extensible system that can work with a wide variety of image applications such as 3D animation, visual effects, compositing, video editing and digital film restoration and is widely used by many animation and post-production studios, including Industrial Light & Magic, Pixar, and Sony Pictures Imageworks. It is widely adopted in the industry due to its flexibility and robustness. More information about OpenColorIO can be found here.

Why would you want to use OpenColorIO in PFClean?

In the example below, we can see a typical OpenColorIO colour-managed pipeline in PFClean. The source clip is a Cineon Log Film scan. If viewed on a workstation monitor or similar device without colour management, it will look washed out due to the scan being a different colourspace, gamma and dynamic range to the monitor. By transforming the film scan from its native colourspace (Cineon Log) into the display colour space, It can be presented correctly and accurately on the workstation monitor. Once work is complete, the footage is exported in a new colourspace to display correctly on the destination device, in this case, a Rec. 709 HD Television.

What makes this process so flexible is that you can quickly change your destination colourspace at a click of a button. Clips can also be colour managed on export to an archival/intermediate colourspace without losing information which can be read in and displayed correctly by another mastering system using OpenColorIO that guarantees accuracy and consistency. We can see this in the example below, where the clip has been transformed into the ACES 2065-1 colourspace on export. This colourspace can be interpreted and displayed correctly by the grading system, guaranteeing what the digital restoration artist is seeing will be what the colourist is seeing.

One of the key benefits of a colour-managed process is its ability to handle clips in various colour spaces. This enables digital restoration artists to effectively manage them and ensure they are all unified into a single, consistent colourspace while also implementing any required restoration corrections. As a result, the workload on the backend is reduced. In the workflow example below, three clips of different colorspaces and file types have been restored, and colour managed into a unified colourspace then exported as ACES 2065-1, EXR files ready to be handled by a grading system later on. Using a colour-managed method to bring all clips into a single colour space can prevent the inconsistencies that might have occurred when using an unmanaged workflow and manual colour correction.

When it comes to archival and preservation, employing a colour-managed pipeline in PFClean holds significant importance. It’s crucial to ensure the longevity of restored footage, especially if it’s historically significant, by saving it in a way that guarantees its preservation. Converting the restored material into a colour space like ACES 2065-1 retains all the colour information available in the source material. Employing an OpenColorIO workflow based on industry standards allows for easy unpacking of data at a later time and fast conversion to any current or future colour space for presentation. This would not have been possible without settling for compromises had the archival master been created using an unmanaged workflow and a lesser colour space.

Surrounding Terminology?

Understanding these key terms will help you navigate the complex world of colour management. If you would like to take a deeper look at colour theory, ACES has a number of helpful articles found here.

Colourspace

Colourspace specifies how pixel values in an image should be interpreted to produce colour on a display device. Using the wrong colourspace means your colours don’t look right. Different colour spaces have different properties and limitations, and understanding the colourspace of the original footage is essential for accurately reproducing and preserving the colours in the final output. Two common colour spaces used in Film and Television are Rec. 709 and DCI-P3.

Gamut 

Gamut is a subset of all visible colours that can be represented by the colourspace. Different colour spaces have different gamuts, meaning they can represent different ranges of colours.

For example, the sRGB colour space, which is commonly used in digital imaging, has a relatively small gamut compared to other colour spaces, such as Rec. 2020, which has a much wider gamut meaning it is able to represent a larger subset of all possible colours. We can see an example of this below.

Gamma

Gamma is a parameter used in colour space to describe the non-linear relationship between the input signal and the displayed brightness of an image or video. Usually expressed as a numerical value, typically between 1.8 and 2.5, and is used to adjust the brightness of an image or video to match the characteristics of the display device being used. For example, a gamma value of 2.2 is commonly used in the sRGB, which is used as a standard for many workstation displays.

Dynamic Range

Dynamic range refers to the range of brightness levels captured or displayed by a camera, film stock, or video monitor. In film, dynamic range is determined by the film stock’s ability to capture a wide range of tones, from the darkest shadows to the brightest highlights. 

Bit Depth 

Bit depth refers to the number of bits used to represent the colour information of each pixel in a digital image. It determines the number of colours represented in an image, with higher bit depths allowing for a greater range of colours / tonal values and more detail in the image.

White Point

The white point represents the colour temperature of the white colour in a particular device or system. It is usually measured in Kelvin. D65 is a commonly used white point in the Film and Television industry. An incorrect white point can affect the colour accuracy of an image.

Calibration

The process of adjusting a device or system, such as a workstation display, to ensure accurate colour reproduction. It involves measuring the colour output of a device, comparing it to a standard, and making adjustments as necessary.

ICC Profile 

An International Color Consortium (ICC) profile is a standardized format for describing the colour space and gamut of a device or system. It helps ensure colour accuracy and consistency across different devices. PFClean can load ICC profiles for calibrated monitors.

---

Conclusion

Whether exporting your restoration for presentation, archiving or passing on to another application, with PFClean’s colour-managed pipeline, users can be assured of precise colour management without the headache of overly complicated workflows, ensuring that the colourspace information is preserved throughout the restoration process. 

If you want to see this in action we have a short quick tips video that will take you through the steps in this process here.

---

Identifying Common Tape Defects

It’s everyone’s worst nightmare; you put the tape into the deck knowing you have one or two chances of capturing a fragile tape.

You cross your fingers, hoping it’s going to look OK… it doesn’t! Don’t worry – all is not lost!

Even some of the worst tape faults can be fixed and below I identify some of the more common issues you will come across and provide some hints and tips as to tools to try inside PFClean’s Telerack and Workbench to help make your video artefacts disappear post-capture. So let’s fast forward…

Fixing tape defects with Telerack

One of the features of PFClean is the powerful Telerack video restoration engine. With an emphasis on speed and a focus on the most common defects, this is an ideal way to restore tape-based media, especially in cases where a fast turnaround and large volume of media are involved. Below is a list of common tape faults with examples, along with tips on how to identify them and how they can be easily fixed in Telerack.

Tape Dropouts

This example shows the repair results using the Telerack Fix Streaks tool

Tape dropouts mainly present themselves as a horizontal line, sometimes teardrop-shaped and usually bright, and appear on screen for one or two frames with staggered intensity. Occasionally these can be persistent through an entire tape.

Off-Lock Errors

Here we can see the repair using the Fix Streaks and Fix Bands tool

Off-lock errors are like a dropout but larger and with greater frequency, with a band of colour or misregistered image usually lasting no more than a frame. They are a common sight on old and worn analogue tapes.

Timebase Corrector Dumping

To repair this issue image stabilisation was used

Timebase corrector dumping is sometimes referred to as a bump. This problem presents itself as a brief shift in image position,  normally vertically, and can range from very mild to very severe.

Excessive Tape Noise

The De-noise effect is used to strip back multiple generations of analogue noise

Excessive tape noise will appear in tapes that are multiple generations away from the original source and/or have started to deteriorate with age. Additionally, material that originated on legacy camera systems can be susceptible to excessive noise due to its low sensitivity.

Chroma Subsampling

The Chroma Resample effect in the Telerack toolset provides a quick high quality fix to this footage

High chrominance areas can lack fidelity, especially where the source is 4:1:1 or 4:2:0. It is most noticeable in areas of red along diagonal surfaces. Older professional and commercial analogue formats and lower-quality digital tape formats all suffer greatly from a lack of chroma information.

Chroma Fringing

The coloured shimmering is dramatically reduced using the Chroma Cleanup effect.

Chroma fringing can appear in high-value chrominance and specular detail. This is caused by crosstalk in the luma and chroma signals. Next time you watch an old television series, watch areas of high detail and you might notice a coloured shimmering. This is chroma fringing.

Tape Banding

Hard to remove manually, the Fix Bands tool makes short work of these dark luma bands.

Tape banding appears as horizontal light or dark bands across the image lasting over a number of frames. Rapid changes in luminance within a scene can sometimes exacerbate this problem.

Flickering, Scratches and Dirt

These artefacts can be removed easily using the Dustbust, De-Flicker and Fix Scratch effects.

Tape material that has been telecined from 16mm and 35mm can suffer all the artefacts that originate in the original film elements. Typically, archive material that’s been transferred to tape suffers from excessive dirt and scratches.

Scanline Flicker

This footage from the 1970s shows the results of a successful restoration using the Scanline De-Flicker effect.

Scanline flicker is minor variances in luminance values between scanlines. It is sometimes caused by variances in field luminance. Not to be confused with rolling bands.

---

Fixing tape defects with the Workbench

The Telerack is focused on high-performance video restoration. Ideal for the hours and hours of tape that can be found in archives. However, some fixes are so severe they require the precision that can be found in PFClean’s powerful Workbench. In the table below we will help you identify these severe faults and suggest Workbench tools to help you restore your tape.

Momentary Head Clog

The Fix Frame effect was used to remove the misaligned image area and the shift in luminance.

Momentary head clog is a horizontal band or sometimes an entire frame of misaligned and warped image, usually for a single frame. The distortion can be complex and requires the rebuilding of a portion of the frame.

Scratched Tape

A combination of the Fix Frame and Fix Bands effects were used to remove the scratches.

Scratched tape is common in formats such as 2″ Quad and TypeC where the physical tape is exposed to the environment. This artefact presents itself as a thin horizontal line of misregistered image that remains static and constant for the duration of the scratch. In a way, it is very similar to a film scratch but horizontal.

Transverse Tape Damage

Using a combination of the Fix Frame and Paint tools a successful fix was made in this example.

Transverse tape damage is a horizontal band of misregistered image that rolls up the screen, usually from bottom to top. This is a fairly common fault in old analogue tapes. Next time you watch an old VHS, look out for this fault.

Capstan Servo Off-Locks

Using the Stabilise, Fix Frame and Pan & Scan effects has fixed this tape error.

Capstan servo off-lock errors are moments of picture instability and sometimes picture breakup, along with wow distortion in audio. Normally this is seen at the top or bottom of the screen.

Mild Tape Mistracking

Panning & Scanning the image removes the undesired area at the top of the image.

Mild tape mistracking appears as a thin band at the very top of the screen with segmented or misregistered images. It can be fairly constant if the tracking was not adjusted correctly during the capture.

Severe Tape Mistracking

While not easy, a combination of Painting and using the Fix Frame were used to rebuild the entire image over several frames.

Severe tape mistracking is the breakup of the entire image resulting in multiple dark and light lines with bands of misregistered image, flickering, and loss of colour. It is possibly the most complex error you will encounter and the most difficult to fix. Sometimes this is why it is useful to have a dub of the tape even if it is of lesser quality so that it can be used to rebuild the images.

Lifted Blacks

Lifted blacks can easily be corrected by using the video grade effect and observing the scopes.

Lifted blacks are caused by transfer errors in the dubbing process. Sometimes this can occur when standards converting from one region to another. NTSC-originated material can look milky on PAL systems if not properly converted.

Chroma Phase Convergence Error

A clip like this can be salvaged by applying the grading tools.

Chroma phase convergence errors can be observed via a vectorscope and waveform where chroma phase is out of alignment. In the example above, this was caused by material from one tape being spliced into the master without correct calibration.

Blocking and Compression Artefacts

The Blocking Reduction effect smooths out any undesired areas where image breakup has occurred.

Highly compressed formats such as mini DV can suffer macro blocking and image breakup during high dynamic and kinetic shots resulting in squares and mosquito noise around detail.

Persistence Trails

The Paint effect was used to paint out the trails in this particular example.

Persistence trails are luminance/chroma trails that appear in bright highlights and chroma. They appear when the luminance value has not had time to reset to zero causing a ghosting trail or comet. They are common in material recorded using cathode ray tube cameras from the 1930s to the 1980s.

Horizontal and Vertical Sync Pulse Loss

Paint, Fix Frame and Pan & Scan effects were all used to fix this error.

Sync pulse loss errors show up as a merging of the adjoining frame with a vertical or a horizontal breakup in the image. If this error occurs, it’s normally accompanied by one or more of the errors described above on the surrounding frames.

Links

Sony has a great page here showing their milestones in broadcast history from the early 1950s through to the modern-day.

The clip examples featured in the article are from the Bolshoi Ballet restored for SKY ARTS.

Identifying Film Formats

When restoring film footage digitally, it can be advantageous to understand the film’s physical qualities before being digitised. Film gauge is a quality that defines the film’s physical width and provides hints to the image quality, such as grain size and resolution.

In this article, we explore the smallest film gauges and work our way up the common larger formats that you may come across. Throughout the article, I will be using 35mm film size (Academy 4 perf frame) for size comparison. In each case, I will be referring to the camera’s original neg / reversal areas, which is the ideal starting point for a digital restoration, rather than process or projection areas.

In the following sections we will be covering film sizes in mm.

Standard / Regular 8mm / Double 8

Primary usage: amateur/distribution.

This motion picture format, measuring a tiny 8mm wide, was first released by Eastman Kodak in 1932. It is sometimes referred to as Standard 8mm, Regular 8mm, or Double 8. The main purpose of this format was to provide a cost-effective alternative to 16mm for amateur filmmaking and distribution during the Great Depression.

Interestingly, this film gauge is actually 16mm film with double the perforations down each side, allowing the user to expose half the frame, then flip the spool and expose the other half. In practice flipping, the spool needed a bit of experience so as not to expose the film. However, it was certainly an interesting way to extend the run time from a single 16mm roll by exposing four 8mm frames in the space of one 16mm frame.

It’s unusual to see the film in its native state because when the film was sent to the lab, it was processed, slit down the middle and spliced together to form a single reel of standard 8mm. There was another less common format called straight 8, which was pre-cut.

With a very small camera aperture of 4.5mm x 3.3mm, this format produces very grainy and usually quite soft images due to the vintage lens designs used for capture. Optically resolving fine detail in such a small area is a challenge today. 

Thanks to the fact that it uses 16mm transport, the familiar gate weave that is synonymous with Super 8 is far more controlled than Standard 8. Due to this being an amateur format, cameras sometimes aren’t maintained to the same level as a professional camera and, as such, are prone to issues such as tramline scratches and gate hairs. Film burn is also fairly common due to how the spool is loaded.

Super 8

Primary usage: amateur/distribution.

Developed as an idea during the early 1960s and released in 1965, Super 8 was designed to provide an easier-to-use, better-quality variant to the ageing standard 8mm format. Super 8 still uses the 8mm film gauge but uses more of the film’s surface area to expose at 5.46mm x 4.01mm vs 4.5mm x 3.3mm of the standard 8mm format, which was achieved by drastically reducing the perforation size. Unlike 8mm, super 8 is cartridge-fed, making it much easier to load and more accessible to the end user. Super 8 also offered the ability to take a soundtrack between the frame's edge and the exposure area's edge.

As with 8mm, Super 8 has a very small exposure area compared to something like 35mm and, because of this, also suffers a large visible grain structure. That said, when using comparable film stock and scanning it to the exact resolution, it is a slightly finer grain compared to standard 8mm. Super 8 cameras and optics were generally mass-produced, and while some had outstanding mechanisms, a lot were very cheaply made, suffering from softness from poor optical design and incredible instability in film transport.

Because Super 8 is a point-and-shoot format, people quite often do not check the film gate for debris. Scratches are also typical in this format.

9.5mm

Primary usage: amateur/distribution.

Introduced by Pathé in the early 1920s, 9.5mm was intended primarily as a distribution format for the projection of films in the home. The format was easy to use and inexpensive, so it rapidly became popular. Not long after its release, a camera system became available allowing amateur filmmakers to shoot with the format, which was more common than you would think, especially in Europe. The camera mostly used reversal film stocks.

The format used an unusual single perforation between each frame, a departure from other amateur and professional formats of the time, which employed perforations on the side(s).

Although unusual, this design had a fairly significant benefit: It allowed for the area on the sides normally taken up by perforations to be used for the picture area. Given that the format is only slightly larger than 8mm, it manages to cram in an image of 8.2mm by 6.15mm, which, impressively, is just shy of the standard exposure area of 16mm.

The format slowly declined after the release of 8mm in the early 1930s despite being considered an inferior format. By the 1960s, it was no more, only kept alive by enthusiasts. Most examples of 9.5mm are in very good condition, which is perhaps a testament to the design, especially compared to other film gauges of a similar vintage.

When scanned at a similar resolution to 8mm, 9.5mm has greater fidelity and a tighter grain structure. However, it does suffer from the same common problems as those formats, such as gate weave and blemishes like scratches and dirt covering a significant part of the image due to their relative size to the aperture.

Standard 16mm

Primary usage: Industrial films, documentaries, news gathering, broadcast.

Eastman Kodak introduced standard 16mm in the early 1920s. Initially, the format was intended for amateur use as the film gauge was considered inferior quality and unsuitable for professional applications.

Having proven popular, the format was later developed with the addition of an optical soundtrack on the edge of the frame during the 1930s. The format went on to prove its worth extensively during World War Two and became the go-to format for many industries post-war, thanks to its low cost and high-quality alternative to traditional 35mm.

For television, the cost-benefit was particularly advantageous. From 1960 to the 1990s, 16mm was used in television, especially by the BBC in the United Kingdom, which worked extensively with Kodak to refine the format and found a great use for it, particularly for location shoots. Consequently, 16mm is a format you will likely come across in abundance in your archival restoration work.

Measuring 16mm across with two perforations on either side of each frame line when using double perf stock, the 10.26mm x 7.49mm camera aperture offered far superior quality to 8mm.

Having found its place in many professional environments, Standard 16mm was used with high-quality camera equipment and optics. This is why much of the footage shot on 16mm is sharp, well exposed and stable. The grain structure is much tighter and more refined than 8mm formats when using a comparable film stock and scanned to a similar resolution. Additionally, higher quality colour negative stocks were available for 16mm that weren’t available for 8mm. The format is also less prone to the artefacts you might see with poorly maintained amateur camera equipment.

Super 16

Main usage: modern broadcast, feature films, commercials.

In 1969, Rune Ericson developed a new variant of the 16mm gauge, designed to maximise the exposure area of the film. Standard 16mm has perforations on both sides of the film to allow registration pins in the camera to position the film correctly in the gate. While some 16mm cameras used registration pins for both sides of the film, quite a few 16mm cameras only used one, and some used no registration pins. Ultimately, it was deemed that using a single registration pin had no detrimental effect on the resulting images and that the space used for the perforations on one side of the film was a waste. Removing the soundtrack area and perforations and then widening the camera film gate allowed the user to record a larger 1.66:1 ratio image onto the same 16mm film gauge, and thus, super 16mm was born.

Initially, this 1.66:1 aspect ratio meant it was more suited to feature films where it could be blown up to 35mm and presented in its native ratio or with minor amounts of cropping at 1.85:1. In traditional broadcasts, 16mm was still favoured for a while because it matched the 1.33:1 viewing format, and super 16mm offered no distinct advantage. However, with widescreen future proofing in mind and with the advent of HDTV broadcasts on the horizon, filmmakers ultimately began to shift to using Super 16mm to take advantage of the extra area the format provides.

Measuring 12.52mm x 7.41mm, Super 16mm offered approximately 21.5% more exposure area over standard 16mm. Of course, this added exposure area only became an advantage when presenting the format in 1.66:1 or 1.85:1 aspect ratios and where an equivalent piece of standard 16mm would have had to have been cropped severely to match it.

Super 16mm offers very similar image quality to standard 16mm. The grain structure is relatively coarse compared to 35mm but has an aesthetic which has become popular with modern filmmakers, who are after the characteristic film look while maintaining a level of control over the image. Super 16mm was and is still used with well-maintained professional cameras and optics. Artefacts and damage on the film's surface will appear worse compared to 35mm scanned at a similar resolution due to the magnification. Generally, gate weave is superior to 8mm formats, but its stability is still not quite as good as 35mm.

The 35mm motion picture format

The 35mm film gauge is the most common and possibly the oldest type you will likely come across. The first form of this popular film gauge was introduced in 1892 by William Dickson and Thomas Edison and went on to become the accepted international standard for film projection in 1909 due to its size being a good trade-off between quality and cost.

The film gauge has chemically undergone many changes since then, including colour and a redesign to create a safer non-volatile film base. But, physically, it has remained the same with only the four perforations on either side having had very minor alterations for different applications.

Despite the rise of digital cinema cameras and only a single manufacturer making motion picture film stock, the 35mm film gauge is still in use today with many DOPs and cinematographers still preferring the look. Being the industry standard, countless feature films and television programmes have been shot on this film gauge and as such, it’s likely to be the most common format you work on as a digital restoration artist.

In its standard form, a 35mm film gauge has a maximum exposure area of 24.89mm by 18.67mm in a 1.33:1. However, the exact size of the exposure area varies depending on the given capture format. Below I describe the most common formats starting with the smallest 35mm variant.

35mm 2 perf / Techniscope

Primary usage: feature films, modern broadcasts, commercials.

Developed and introduced by Technicolour in 1963, this wide format uses an exposure area of 21.95mm by 9.47mm in a 2.33:1 ratio.

While looking physically small in comparison to the full aperture 35mm format, #Techniscope had some key advantages. Rather than using the standard four perforations in a negative pulldown, #Techniscope only used two perforations, which had the benefit of doubling the number of frames you could expose on a given roll of film. Additionally, standard spherical lenses could be used, which were more accessible and quite often faster, sharper and focused closer than the anamorphic equivalent. Despite four-perf with anamorphic lenses being the benchmark of the 35mm formats, two-perf allegedly had more clarity when projected, perhaps due to the simpler optics.

A number of major films used the format to great success including The Ipcress File (United Artists 1965) which had a major restoration using PFClean. Unfortunately, the format itself never quite reached the same popularity as four-perf as initially there was some criticism of the format due to additional complexities in editorial and printing, despite these generally being resolved with the advent of the digital intermediate. It was also generally considered to be more grainy when compared to four-perf due to the reduced size of the exposure area. Artefacts like gate hairs that would have been cropped out in other formats became a problem when restoring two-perf.

35mm three-perf

Primary usage: modern broadcast, feature films, commercials.

Initially devised and patented in 1975 by Miklos Lente and called Trilent 35, this format offered an exposure area of 21.95mm by 13.9mm in a near 1.78:1 ratio, using a three perforation pulldown. Initially, there was no industry interest in this obscure gauge and it lay dormant until the early 1980s when cinematographer Rune Ericson collaborated with Panavision to breathe new life into the format with the intention of initiating an industry change to convert all film equipment over to the format, including projection systems in theatres.

As with the two-perf Techniscope, there were advantages to using the now-renamed three-perf format. Due to the format using a three-perforation pulldown, the film ran through the camera 25% slower than four-perf, which resulted in a reasonable saving on film stock when shooting for non-anamorphic widescreen ratios.

Like Techniscope, 35mm three-perf simply never achieved the popularity of four-perf. Despite the savings in film stock, it came down to the fact that a huge amount of equipment such as projectors and optical printers would need to be converted, which proved to be undesirable. This didn’t stop people from using the format, though, as three-perf could be transferred optically to four-perf for release. Eventually, these issues became irrelevant as the digital intermediate process removed many of these obstacles. High-end television made good use of the format as it matched the ratio of HDTVs while saving money on film stock.

35mm four-perf Academy / 1.37:1 / 1.66:1 / 1.85:1

Main usage: modern broadcast, feature films, commercial.

Having gone through a few variations since its initial conception in the early 1920s, in 1932 the Academy of Motion Picture Arts and Sciences refined four-perf to bring standardisation to the industry. The format later became known simply as ‘Academy’.

With an aspect ratio of 1.37:1, the camera aperture measures 21.95 mm by 16mm with the soundtrack area on the left side. All films shot between 1932 and 1952 used the Academy 1.37:1.

In the early 1950s, Fox announced its upcoming widescreen format cinemascope. This led to the other studios looking for ways to achieve similar wider ratios for non-anamorphic (flat) presentations. In 1953, Paramount became the first to deviate from the standard 1.37:1 ratio with 1.66:1. This format was achieved by simply masking the top and bottom of the 1.37:1 image.

While America favoured 1.85:1, it was Europe which eventually adopted the format for presentations shooting large numbers of films in this format. These productions were often photographed conventionally using the full 1.37:1 and cropped later on rather than using any in-camera masking.

Universal Studios and Columbia Pictures were the first to utilise the 1.85:1 format and the ratio was achieved similarly to the way 1.66:1 was derived, by cropping the academy 1.37:1 frame at the top and bottom. This provided a way for the industry to provide a true widescreen ratio while keeping the costs down when modifying existing equipment. It also provided an easy way for films shot for 1.66:1 to be presented in 1.85:1 as the framing differences weren’t drastic. By the mid-1950s, 1.85:1 became the standard for non-anamorphic theatrical presentation in America.

Right up until today the vast majority of non-anamorphic features are still exhibited in the 1.85:1 ratio, with countless features being shot in this format. Recently, there has been a small resurgence in the format for acquisition, a great example being Wes Anderson’s 2014 award-winning The Grand Budapest Hotel.

Super 35mm

Primary usage: feature films, modern broadcasts, commercials.

The initial concept for Super35 was to return to using the entire width of the aperture area of the four-perf 35mm film, which, in essence, is the same as Edison’s silent original from the early 1900s.

The concept was to expose the full area of the frame and then crop down to 2.35:1 and reduction print/blowdown to a four-perf print, with the theory being that the use of a larger area of the film when compared to the academy formats would provide increased fidelity and tighter grain. Confusingly, there is also a three-perf version of Super35.

Super35 uses the full aperture area of 24.89mm by 18.67mm in four-perf and 24,9mm by 14,7 mm in three-perf. These were provisions in framing to allow for a television version to be made with the maximum amount of neg area. 

In the pre-digital intermediate era, the format was controversial among industry professionals. Many believed the steps required to print the extracted Super35 area to a four-perf print optically softened the image and negated its benefits. However, some preferred the process and the ease of working with spherical lenses. James Cameron is a big proponent, having shot a number of his films on the format.

When digital intermediate (DI) came about in the early 2000s, many issues using Super35 disappeared. With DI, Super35 could be scanned into the computer, cropped to 2.39:1 and rendered to an anamorphic four-perf frame size without any generation loss. Additionally, there were benefits to working in post-production with the entire uncropped area of Super35. Images could be easily re-racked, and the extra information outside of the crop area benefited visual effects and proved a popular modern format, with many films being shot on the format.

Anamorphic 35mm ‘Scope’

Main usage: feature films, commercials.

While it has its origins in the early 1920s and optically even before WW1, the primary use of anamorphic in the film industry began with the race to create wider aspect ratios for cinema presentation during the early 1950s to compete against the increasing popularity of television. While other studios went down the road of matting the top and bottom of the 1.37:1 academy area to achieve a widescreen ratio, it was 20th Century Fox that chose the anamorphic format naming it CinemaScope and produced many films between 1953 and 1967 using the total aperture area of the four-perf 35mm frame.

With SMPTE and Panavision making changes and refinements to the format during that time, including slight differences in aspect ratio and the inclusion of an optical soundtrack area, it later became known as ‘Scope’.

Rather than projecting a spherical image onto the film plane, anamorphic optics project a horizontally compressed oval image in a 2x ratio or 2:1. When the film is processed, a print is made for projection and another anamorphic lens with a matching 2:1 squeeze factor is used on the front of the projector to un-squeeze the image and present it in its correct ratio. For capture, anamorphic uses nearly the full height of the 35mm four-perf frame with an exposure area of 21.95mm by 18.6mm. The format was an ingenious way to maximise the negative space while delivering a very wide aspect ratio of 2.39:1 and still using the same film stock, cameras and projection.

Anamorphic is an exciting format which has seen a resurgence in recent years, with many directors such as J.J. Abrams and Christopher Nolan opting to use the format, keen to capture the classic cinematic look for a modern presentation. Because the format uses much more of the 35mm neg’s surface area, approximately 408 mm², the grain structure is much tighter when compared to other 35mm formats with the same presentation ratio. Due to complex optics, flaring is common, and often, there is fall-off and softness towards the edges of the frame, especially in older films. While these aren’t usually problems you would want to remove during the digital restoration process, they can make the process much more difficult.

VistaVision

Main usage: feature films, visual effects.

VistaVision was first created by Paramount Pictures back in 1954. Often referred to as lazy eight because the film travels horizontally through the gate, it is very similar to the 135 still photography format, which uses eight perforations top and bottom per exposed frame. The primary purpose of the design was to create a finer-grained print for widescreen 1.66:1 presentations. With an exposure area measuring a massive 37.39mm by 25.3mm, it was a significant step up in quality vs. the traditional four-perf Academy format.

Despite a large number of feature films being shot in the format, Paramount discontinued it less than a decade after they introduced it, mainly due to the fact there had been significant advances in film stocks which negated the benefits of the format. Vistavision was still used in other countries, such as Japan, up until the late 1980s. That wasn’t the end for VistaVision, though.

Cheaper than 65mm, combined with current finer grain film stocks and increased stability of the eight-perf mechanism, VistaVision became the go-to format for many visual effects productions. The format’s advantages became apparent when optical printing and VFX companies like Industrial Light and Magic revived the format, using it extensively in their motion control systems to photograph effects models for the Star Wars films. Today, Vistavision is still in use for visual effects heavy projects shot on film where fine grain and increased resolution are beneficial.  Christopher Nolan used it for his feature film Interstellar.

Ideas from VistaVision eventually evolved into the IMAX 70mm format, which uses a similar horizontal film feed. The design lives on in digital form with modern high-resolution digital cinema cameras using a VistaVision-sized sensor.

65mm, 5 perf

Main usage: feature films, visual effects.

The 65/70mm film gauge has been around since the early days of the film industry, but producer Mike Todd in the early 1950s should be credited with popularising the format again. The Todd-AO process was developed as a competing system to Cinerama, a hugely complex and expensive system at the time.

This film format used two separate film elements – 65mm to capture the images during production and 70mm print for distribution in the theatres. With perforations lining up on both gauge sizes, the extra 5mm on the 70mm is used for optical soundtracks.

Panavision adopted the format after Mike Todd’s death in 1958, creating slight variations in Super Panavision and Ultra Panavision 70. These variations are mechanically compatible with the original format, with Ultra 70 being a 1.25x anamorphic system, creating a vast aspect ratio. Quentin Tarantino used Ultra Panavision 70 for his 2015 film The Hateful Eight.

The 65mm camera neg uses a huge camera aperture of 52.48mm by 23.01mm surpassing the quality of even the most significant 35mm format. 65mm using 1207mm² of exposure area vs 35mm anamorphic 408mm². The quality on screen was unmatched and was only surpassed by the specialist 15 perf 70mm IMAX format in recent years. However, the cost of shooting in 65mm was extraordinarily high and reserved for use on only the highest-budget films.

Plenty of 70mm prints were made, especially from 35mm anamorphic blowups, which took advantage of the higher quality presentation format and the better sound. While a print is not the ideal starting point for restoration, 70mm does offer vastly superior quality compared to a conventional 35mm print. Additionally, the format was used for optical effects work to increase fidelity and reduce the amount of grain in the final images when optically printing.

Iconic films like 2001: A Space Odyssey (1968) have been shot entirely on 65mm, 1992’s Baraka being a stunning example of what the format is capable of. Christopher Nolan shot large segments of Dunkirk (2017) using 65mm, with key sequences being shot with the larger IMAX format. When scanned at a comparable resolution to 35mm, the grain structure is extremely tight and the fidelity is exceptionally high. A larger area of film can have the potential for more surface dirt and issues. Still, generally, being such a high-end format, the camera’s original 65mm film elements are exceptionally well looked after.

---